WO2024152027A1

WO2024152027A1 - Framework for modeling synthetic and experimental agronomic data

Info

Publication number: WO2024152027A1
Application number: PCT/US2024/011516
Authority: WO
Inventors: Philip Andrew MCLOUD; Timothy FRATANGELO; Samuel J. PETERS; Ben Wong SHULMAN
Original assignee: Indigo Ag Inc
Current assignee: Indigo Ag Inc
Priority date: 2023-01-13
Filing date: 2024-01-12
Publication date: 2024-07-18
Anticipated expiration: 2025-07-13

Abstract

A system for modeling synthetic and experimental agronomic data is provided. The system generates, based on accessed agronomic data, a representative data object for each management event performed on an agronomic region. The system generates, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template. The system identifies one or more candidate data objects corresponding to the target experimental farming practice for the candidate agronomic region and determines an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions. The system assigns an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

Description

FRAMEWORK FOR MODELING SYNTHETIC AND EXPERIMENTAL AGRONOMIC DATA

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Patent Application No. 63/479,928, titled “Framework for Modeling Synthetic and Experimental Agronomic Data,” filed on January 13, 2023, and U.S. Patent Application No. 63/503,874, titled “Framework for Modeling Synthetic and Experimental Agronomic Data,” filed on May 23, 2023; the disclosure of which is hereby incorporated by reference in its entirety.

BACKGROUND

[0002] Embodiments of the present disclosure relate to farming practices and associated environmental characteristics of agricultural production, and more specifically, to modeling of environmental characteristic and agronomic data.

BRIEF SUMMARY

[0003] According to embodiments of the present disclosure, methods and computer program products for determining agricultural interventions and associated environmental characteristics are provided. Agronomic data indicating a plurality of management events each having an associated time, geospatial boundary, and one or more attribute is obtained. A data object for each of the plurality of management events is generated where the data object indicates the respective associated time, geospatial boundary, and one or more attribute. An experimental template is read from a datastore. A plurality of experimental data objects of a first type is generated from at least one of the plurality of data objects by altering its one or more attributes. Each of the plurality of experimental data objects conforms to the experimental template. Each of the plurality of experimental data objects is provided as input to one or more models to generate a respective predicted ecosystem attribute. Based on the predicted ecosystem attribute, impact scores are assigned to each of the plurality of experimental data objects.

[0004] In some aspects, the techniques described herein relate to a method including: generating, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generating, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identifying one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; determining an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object; and, optionally, assigning an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

[0005] In some aspects, the techniques described herein relate to a method, wherein the baseline ecosystem attribute is predicted using the one or more second models without performing the target experimental farming practice on the candidate agronomic region. [0006] In some aspects, the techniques described herein relate to a method, wherein each management event includes a farming practice on a corresponding agronomic region. Farming practices are actions taken or avoided, and, optionally, include additional attributes such as a location (for example, a point location or an area), a descriptor (e.g. a type of tillage implement utilized, a variety of seed, one or more crop species, irrigation type), a date or time (e.g., a particular time and or date, for example a planting date) or time period (e.g., crop season or year), or amount (e.g. quantity, depth, duration, or intensity). Farming actions taken or avoided may include, without limitation, planting a crop (e.g. one or more cover crops), tillage, irrigation, an input applied (for example, a fertilizer, manure, one or more microbe, a material for direct air capture of a greenhouse gas, a silicate material (for example, crushed silicate rock such as basalt)), harvest, termination, burning, grazing, etc. In various embodiments, farming practices may apply to entire fields, to more than one field, or subregions or points within a field. Within a single crop season some farming practices may be applied to an entire field, while other farming practices may be applied to a subfield region.

[0007] In some aspects, the techniques described herein relate to a method, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

[0008] In some aspects, the techniques described herein relate to a method, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production.

[0009] In some aspects, the techniques described herein relate to a method, wherein the target experimental farming practice includes altering one or more attributes of the management event for the candidate agronomic region.

[0010] In some aspects, the techniques described herein relate to a method, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region. An experimental template may specify one or more of: one or more models to be applied, a model version for the one or more models, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use.

[0011] In some aspects, the techniques described herein relate to a method, wherein the second models include at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator.

[0012] In some aspects, the techniques described herein relate to a method, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory-based greenhouse gas emissions calculator, a statistical model.

[0013] In some aspects, the techniques described herein relate to a method, further including: generating an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts.

[0014] In some aspects, the techniques described herein relate to a method, further including: enabling a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

[0015] In some aspects, the techniques described herein relate to a method, further including: quantifying ecosystem attributes for one or more agronomic regions; and displaying the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

[0016] In some aspects, the techniques described herein relate to a method, further including: assigning an impact score based on the determined ecosystem attribute of the candidate agronomic region; and providing a recommendation based on the impact score to a user via a user interface.

[0017] In some aspects, the techniques described herein relate to a method, further including: receiving a modification to one or more parameters in the target experimental template from a user via a user interface.

[0018] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium including stored program code, the program code including instructions, the instructions when executed cause a processor system to: generate, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generate, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identify one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; determine an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object; and, optionally, assign an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

[0019] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the baseline ecosystem attribute is predicted using the second model without performing the target experimental farming practice on the candidate agronomic region.

[0020] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein each management event includes a farming practice on a corresponding agronomic region, and the farming practice includes planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, organic amendment, or combinations thereof.

[0021] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

[0022] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production

[0023] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the target experimental farming practice includes altering one or more attributes of the management event for the candidate agronomic region.

[0024] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region.

[0025] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the target experimental template specifies one or more of: the one or more second models to be applied, a model version for each second model, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use. [0026] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the second models include at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator. [0027] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory -based greenhouse gas emissions calculator, a statistical model.

[0028] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the instructions when executed further cause a processor system to: generate an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts.

[0029] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the instructions when executed further cause a processor system to: enable a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

[0030] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the instructions when executed further cause a processor system to: quantify ecosystem attributes for one or more agronomic regions; and display the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

[0031] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the instructions when executed further cause a processor system to: assign an impact score based on the determined ecosystem attribute of the candidate agronomic region; and provide a recommendation based on the impact score to a user via a user interface.

[0032] In some aspects, the techniques described herein relate to a non-transitory computer readable storage medium, wherein the instructions when executed further cause a processor system to: receive a modification to one or more parameters in the target experimental template from a user via a user interface.

[0033] In some aspects, the techniques described herein relate to a system including: one or more computer processors; and one or more computer-readable mediums storing instructions that, when executed by the one or more computer processors, cause the system to: generate, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generate, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identify one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; determine an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object; and, optionally, assign an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region. For example, one or more second models may comprise at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator. For example, an inventory-based greenhouse gas emissions calculator computes pre-field and some on-field greenhouse gas emissions for a set of farming practices (e.g., planting, harvest, tillage, etc.). In one example, an inventory-based greenhouse gas emissions calculator model ingests agronomic events, translates each event into a set of activities and their corresponding amounts (for example, using a life cycle inventory database), computes emissions (e.g. kg CCh-equivalent, “kg CC e”) for each activity (for example, using TRACI (an EP A method for equating emissions to various impact categories) impact factors, and aggregates emissions for each high-level agronomic event to provide total kg CO?.e input region (e.g. a field, set of fields, or fanning operation).

[0034] In some aspects, the techniques described herein relate to a system, wherein the baseline ecosystem attribute is predicted using the one or more second models without performing the target experimental farming practice on the candidate agronomic region. [0035] In some aspects, the techniques described herein relate to a system, wherein each management event includes a farming practice on a corresponding agronomic region, and the farming practice includes planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, or organic amendment, or combinations thereof.

[0036] In some aspects, the techniques described herein relate to a system, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

[0037] In some aspects, the techniques described herein relate to a system, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production.

[0038] In some aspects, the techniques described herein relate to a system, wherein the target experimental farming practice includes altering one or more attributes of the management event for the candidate agronomic region.

[0039] In some aspects, the techniques described herein relate to a system, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region.

[0040] In some aspects, the techniques described herein relate to a system, wherein the target experimental template specifies one or more of: the one or more second models to be applied, a model version for each second model, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use.

[0041] In some aspects, the techniques described herein relate to a system, wherein the second models include at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator.

[0042] In some aspects, the techniques described herein relate to a system, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory-based greenhouse gas emissions calculator, a statistical model.

[0043] In some aspects, the techniques described herein relate to a system, wherein the instructions when executed further cause a processor system to: generate an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts. [0044] In some aspects, the techniques described herein relate to a system, wherein the instructions when executed further cause a processor system to: enable a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

[0045] In some aspects, the techniques described herein relate to a system, wherein the instructions when executed further cause a processor system to: quantify ecosystem attributes for one or more agronomic regions; and display the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

[0046] In some aspects, the techniques described herein relate to a system, wherein the instructions when executed further cause a processor system to: assign an impact score based on the determined ecosystem attribute of the candidate agronomic region; and provide a recommendation based on the impact score to a user via a user interface.

[0047] In some aspects, the techniques described herein relate to a system, wherein the instructions when executed further cause a processor system to: receive a modification to one or more parameters in the target experimental template from a user via a user interface.

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS [0048] Figure (FIG.) l is a block diagram that illustrates an ecosystem management environment, in accordance with one or more embodiments.

[0049] FIG. 2 is a data flow diagram of an example process for generating a predicted ecosystem attribute, in accordance with one or more embodiments.

[0050] FIG. 3 provides a layout of an example data object, in accordance with one or more embodiments.

[0051] FIG. 4 A provides exemplary DSL instructions that are included in the experimental data object to add and remove new management events during simulation, in accordance with one or more embodiments.

[0052] FIG. 4B provides an example of DSL instructions of the experimental data object to alter the attributes of the management events of the candidate data object while running simulations, in accordance with one or more embodiments.

[0053] FIG. 5 provides a visual representation illustrating the permutations of the management events and the one or more attributes describing the management events, in accordance with one or more embodiments. [0054] FIG. 6 is a flow diagram that illustrates an example process of generating a predicted ecosystem attribute, in accordance with one or more embodiments.

[0055] FIG. 7 is a flow diagram that illustrates a method for generating synthetic agronomic data, in accordance with one or more embodiments.

[0056] FIGs. 8-11 are plots comparing select univariate distributions between the empirical dataset and the synthetic dataset in the above-described example, in accordance with one or more embodiments.

[0057] FIGs. 12-13 are plots comparing select bivariate distributions between the empirical dataset and the synthetic dataset in the above-described example, in accordance with one or more embodiments.

[0058] FIG. 14 illustrates an exemplary branching during the method of generating a management zone, in accordance with one or more embodiments.

[0059] FIG. 15 is a schematic of an example of a classical computing node, in accordance with one or more embodiments.

DETAILED DESCRIPTION

[0060] Farmers today are presented with a wide array of agronomic and eco-program choices that provide various economic incentives to adopt farming practices, such as environmentally friendly growing practices. Additionally, it is increasingly important for farmers to provide current reporting on the environmental attributes associated with the agricultural products they produce. Adoption of environmentally friendly farming practices can produce verifiable environmental characteristics (for example, increased soil organic carbon and or reduced greenhouse gas emissions, reduced water usage, reduced chemical contamination (e.g., reduced nitrogen run-off, reduced insecticide/pesticide/herbicide residue, increased biodiversity, and the like).

[0061] The environmental characteristics can be quantified and monetized as ecosystem credits (for example, a carbon credit equivalent to one metric ton of carbon sequestered) under a particular methodology (e.g., a set of requirements). Ecosystem credits may be fully fungible and may be retired (purchased and held without further trading) or may be traded on a secondary market. Companies may purchase and retire ecosystem credits to offset the negative impacts of their operations. For these companies, ecosystem credits are perfectly fungible. Other companies gain additional benefit from being able to associate a particular environmental practice with a particular product (for example, wheat produced using farming practices that result in sequestration of soil organic carbon). Environmentally conscious consumers value and pay premiums for retail goods that have verified and traceable connections to environmentally beneficial production practices.

[0062] Decision making for optimizing such environmental characteristics is increasingly dependent on data. Decision makers such as agricultural production managers, food supply managers, and farmers require accurate, timely, and cost-effective information to maximize their economic incentives but also to maintain a supply of food. However, collecting experimental agronomic data by conducting millions of actual experiments on agronomic region is a highly impractical, expensive, and wasteful endeavor due to a variety of reasons, including the substantial resource requirements, time constraints, space limitations, labor intensity, environmental impact, etc. Performing physical experiments in agriculture involves significant resources, including land, water, seeds, fertilizers, and labor. Agricultural experiments often require considerable time for crops to grow and mature. Running a large number of experiments simultaneously or sequentially would extend the timeline, making it impractical for obtaining timely results, especially in the context of seasonal and climate-related constraints. Moreover, managing and monitoring the experiments would demand an extensive workforce. The labor-intensive nature of agricultural tasks makes it impractical to conduct experiments on a massive scale.

[0063] Agricultural simulation models have been designed that use data such as agricultural practices, farming methodologies and ecosystem attributes such as soil, crop, tree, biological and chemical processes, climate, etc. to predict crop production and assess the viability of novel and current farming practices or combinations thereof. However, analyzing agronomic data can be very challenging due to various factors, including the complexity of agricultural systems, the dynamic nature of environmental conditions, and the diversity of data sources. Agricultural systems are highly influenced by diverse and dynamic environmental factors such as weather, soil conditions, and pest pressures. This variability makes it difficult to identify patterns and trends in the data. Agronomic data can be extensive and diverse, often involving numerous variables. Each variable may correspond to a plurality of values, the combinations of the variables can easily generate 10,000 to 1,000,000 different experimental scenarios. Simulating millions of combinations of these variables requires substantial computational resources and complex models, contributing to the overall cost and complexity. Handling and processing large volumes of data while ensuring its accuracy and relevance is challenging. Moreover, agronomic systems are influenced by a multitude of interconnected factors, making it difficult to isolate the effects of individual variables. For example, crop yield may be affected by a combination of weather conditions, soil health, and pest management practices. This makes it challenging to explore and simulate all possible combinations efficiently, leading to increased computational requirements and potential difficulties in interpreting results.

[0064] Therefore, developing accurate predictive models for agronomic outcomes can be challenging. It requires sophisticated modeling techniques, consideration of multiple variables, and an awareness of the limitations associated with historical data. This specification provides methods and computer program products for generating a recommendation for agricultural practices and/or farming methodologies. The methods include performing experiments based on multiple counter-factual farming practices using multiple agricultural simulation models and predicting ecosystem attributes such as greenhouse gas emissions, crop production, etc. for purposes including long-term revenue planning and evaluating new and emerging sustainability practice changes. The predicted ecosystem attributes are then used to recommend agricultural practices and/or farming methodologies.

[0065] In particular, the disclosed system may obtain agronomic data including historical information of management events and generate a data object for each management event. The system receives a request to perform a target experimental farming practice on a candidate location. When receiving a request to perform a target experimental farming practice on a candidate agronomic region, the system may generate a target experimental data object using a first model by retrieving, from a datastore, an experimental template corresponding to the target experimental farming practice. In some embodiments, the first model may be trained using historical agronomic data.

[0066] Simulations and models play a crucial role in identifying optimal farming practices by providing a virtual environment to test different experimental scenarios and understand the complex interactions within agricultural systems. Using the experimental template, the system may clearly define the objectives of the simulation study and identify the key parameters and variables that influence the outcomes of interest, such as crop yield, resource use efficiency, or environmental impact. Virtual experiments eliminate the need for large amounts of physical resources, resulting in significant cost savings, as simulations can be performed on powerful computing systems without the need for extensive real-world resources. Virtual experiments may easily scale to simulate millions of scenarios, and the scalability allows exploration on a wide range of conditions and variables efficiently. Simulations may provide precise control over experimental conditions and can be performed iteratively so that the parameters be modified, and the models can be refined in a rapid cycle. Additionally, running simulations and models with virtual experiment data do not contribute to environmental degradation associated with large-scale physical experiments. There is no need for excessive use of water, fertilizers, or pesticides, reducing the environmental footprint of agricultural research.

[0067] The disclosed system may identify one or more data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects, provide the generated target experimental data object and the one or more candidate data objects to a second model as input. The second model is a model that predicts an ecosystem attribute of the candidate agronomic region. The system receives, from the second model, a result including a predicted ecosystem attribute of the candidate location corresponding to the target experimental farming practice.

[0068] In some embodiments, the disclosed system may choose from a plurality of models as the second mode. Models can range from simple spreadsheet-based models to more complex, dynamic simulation tools. The disclosed system may choose an appropriate simulation model based on the specific aspects of experimental farming practice and create different scenarios to represent various farming practices or management strategies. This could include changes in crop rotation, tillage practices, irrigation, use of fertilizers, and pest control methods. Develop a range of scenarios that cover potential variations in the system. By executing the simulations using the selected scenarios (e.g., models), the system may predict various ecosystem attributes over multiple seasons or time steps to observe how the ecosystem responds to different inputs and management practices. This helps identify trends, patterns, and potential trade-offs.

[0069] The disclosed system may determine a difference between the predicted ecosystem attribute in the received result and a corresponding ecosystem attribute in a baseline result and assign an impact score to the target experimental data object associated with the target experimental farming practice based on the determined difference. In this way, the system may understand which parameters have the most significant impact on the outcomes. This helps prioritize key factors and focus on the most influential aspects of the farming system. Additionally, the system may use optimization algorithms to find the combination of input variables that lead to optimal outcomes. This may involve maximizing crop yield, achieve a balance between crop quality and ecosystem attributes, maximizing a beneficial ecosystem attribute, minimizing a negative ecosystem attribute, minimizing resource use, or achieving a balance between economic and environmental factors.

[0070] The disclosed system may determine a difference (e.g., a quantification, uncertainty, etc.) in an estimated ecosystem attribute generated between one or more models and/or one or more model versions or model parameters. In some embodiments, the outputs of one or more models may be aggregated. Additionally, the system may use optimization algorithms to find the combination of models that lead to optimal outcomes, for example, reduction of uncertainty, minimization of bias, etc.

[0071] The figures and the following description relate to certain embodiments by way of illustration only. It should be noted that from the following discussion, alternative embodiments of the structures and methods disclosed herein will be readily recognized as viable alternatives that may be employed without departing from the principles of what is claimed.

[0072] Reference will now be made in detail to several embodiments, examples of which are illustrated in the accompanying figures. It is noted that wherever practicable, similar or like reference numbers may be used in the figures and may indicate similar or like functionality. The figures depict embodiments of the disclosed system (or method) for purposes of illustration only. One skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles described herein.

[0073] As used herein an “ecosystem attribute” or “ecosystem benefits” each refer to an environmental characteristic (e.g., a result of agricultural production and/or farming) that may be quantified and valued (for example, as an ecosystem credit or sustainability claim). Examples of ecosystem attributes include without limitation reduced water use, reduced nitrogen use, increased soil carbon sequestration, greenhouse gas emission avoidance, increased yield, reduced nitrous oxide production etc. An example of a mandatory program requiring accounting of ecosystem attributes is California’s Low Carbon Fuel Standard (LCFS). Field-based agricultural management practices can be a means for reducing the carbon intensity of biofuels (e.g., biodiesel from soybeans).

[0074] An “ecosystem impact” is a change in an ecosystem attribute relative to a baseline. In some embodiments, baselines may reflect a set of regional standard practices of production (a comparative baseline), prior production practices and outcomes for a field or farming operation (a temporal baseline), or a hypothetical counter-factual set of production practices (a counterfactual baseline). For example, a temporal baseline for determination of an ecosystem impact may be the difference between a safrinha crop production period and the safrinha crop production period of the prior year. In some embodiments, an ecosystem impact can be generated from the difference between an ecosystem attribute for the latest crop production period and a baseline ecosystem attribute averaged over a number (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10) of prior production periods.

[0075] An “ecosystem credit” is a unit of value corresponding to an ecosystem benefit or ecosystem impact, where the ecosystem attribute or ecosystem impact is measured, verified, and or registered according to a methodology. In some embodiments, an ecosystem credit may be a report of the inventory of ecosystem attributes (for example, an inventory of ecosystem attributes of a management zone, an inventory of ecosystem attributes of a farming operation, an inventory of ecosystem attributes of a supply shed, an inventory of ecosystem attributes of a supply chain, an inventory of a processed agricultural product, etc.).

[0076] In some embodiments, an ecosystem credit is a life-cycle assessment. In some embodiments, an ecosystem credit may be a registry issued credit. Optionally, an ecosystem credit is generated according to a methodology approved by an issuer. An ecosystem credit may represent a reduction or offset of an ecologically significant compound (e.g., carbon credits, water credits, nitrogen credits). In some embodiments, a reduction or offset is compared to a baseline of ‘business as usual’ if the ecosystem crediting or sustainability program did not exist (e.g., if one or more practice change made because of the program had not been made).

[0077] In some embodiments, a reduction or offset is compared to a baseline of one or more ecosystem attributes (e.g., ecosystem attributes of one or more: field, sub-field region, county, state, region of similar environment, supply shed geographic region, a supply shed, etc.) during one or more prior production period. For example, ecosystem attributes of a field in 2022 may be compared to a baseline of ecosystem attributes of the field in 2021. In some embodiments, a reduction or offset is compared to a baseline of one or more ecosystem attributes (e.g., ecosystem attributes of one or more: field, sub-field region, county, state, region of similar environment, supply shed geographic region, a supply shed, etc.) during the same production period. For example, ecosystem attributes of a field may be compared to a baseline of ecosystem attributes of a supply shed comprising the field. An ecosystem credit may represent a permit to reverse an ecosystem benefit, for example, a license to emit one metric ton of carbon dioxide. A carbon credit represents a measure (e.g., one metric ton) of carbon dioxide or other greenhouse gas emissions reduced, avoided, or removed from the atmosphere. A nutrient credit, for example a water quality credit, represents pounds of a chemical removed from an environment (e.g., by installing or restoring nutrient-removal wetlands) or reduced emissions (e.g., by reducing rates of application of chemical fertilizers, managing the timing or method of chemical fertilizer application, changing type of fertilizer, etc.). Examples of nutrient credits include nitrogen credits and phosphorous credits. A water credit represents a volume (e.g., 1000 gallons) of water usage that is reduced or avoided, for example by reducing irrigation rates, managing the timing or method of irrigation, employing water conservation measures such as reducing evaporation or transpiration.

[0078] Offsets” are credits generated by third parties outside the value chain of the party with the underlying carbon liability (e.g., oil company that generates greenhouse gases from combusting hydrocarbons purchases carbon credit from a farmer).

[0079] “Insets” are ecosystem resource (e.g., carbon dioxide) reductions within the value chain of the party with the underlying carbon liability (e.g., oil company who makes biodiesel reduces carbon intensity of biodiesel by encouraging farmers to produce the underlying soybean feedstock using sustainable farming practices). Insets are considered Scope 1 reductions.

[0080] Emissions of greenhouse gases are often categorized as Scope 1, Scope 2, or Scope 3. Scope 1 emissions are direct greenhouse gas emissions that occur from sources that are controlled or owned by an organization. Scope 2 emissions are indirect greenhouse gas emissions associated with purchase of electricity, steam, heating, or cooling. Scope 3 emissions are the result of activities from assets not owned or controlled by the reporting organization, but that the organization indirectly impacts in its value chain. Scope 3 emissions represent all emissions associated with an organization’s value chain that are not included in that organization’s Scope 1 or Scope 2 emissions. Scope 3 emissions include activities upstream of the reporting organization or downstream of the reporting organization. Upstream activities include, for example, purchased goods and services (e.g., agricultural production such as wheat, soybeans, or corn may be purchased inputs for production of animal feed), upstream capital goods, upstream fuel and energy, upstream transportation, and distribution (e.g., transportation of raw agricultural products such as grain from the field to a grain elevator), waste generated in upstream operations, business travel, employee commuting, or leased assets. Downstream activities include, for example, transportation and distribution other than with the vehicles of the reporting organization, processing of goods sold, use of goods sold, end of life treatment of goods sold, leased assets, franchises, or investments.

[0081] An ecosystem credit may generally be categorized as either an inset (when associated with the value chain of production of a particular agricultural product), or an offset, but not both concurrently.

[0082] As used herein, a “crop-growing season” may refer to fundamental unit of grouping crop events by non-overlapping periods of time. In some embodiments, harvest events are used where possible.

[0083] An “issuer” is an issuer of ecosystem credits, which may be a regulatory authority or another trusted provider of ecosystem credits. An issuer may alternatively be referred to as a “registry”.

[0084] A “token” (alternatively, an “ecosystem credit token”) is a digital representation of an ecosystem benefit, ecosystem impact, or ecosystem credit. The token may include a unique identifier representing one or more ecosystem credit, ecosystem attribute, or ecosystem impact, or, in some embodiments a putative ecosystem credit, putative ecosystem attribute, or putative ecosystem impact, associated with a particular product, production location (e.g., a field), production period (e.g., crop production season), and/or production zone cycle (e.g., a single management zone defined by events that occur over the duration of a single crop production season).

[0085] “Ecosystem credit metadata” is at least information sufficient to identify an ecosystem credit issued by an issuer of ecosystem credits. For example, the metadata may include one or more of a unique identifier of the credit, an issuer identifier, a date of issuance, identification of the algorithm used to issue the credit, or information regarding the processes or products giving rise to the credit. In some embodiments, the credit metadata may include a product identifier as defined herein. In other embodiments, the credit is not tied to a product at generation, and so there is no product identifier included in the credit metadata.

[0086] A “product” is any item of agricultural production, including crops and other agricultural products, in their raw, as-produced state (e.g., wheat grains), or as processed (e.g., oils, flours, polymers, consumer goods (e.g., crackers, cakes, plant-based meats, animal-based meats (for example, beef from cattle fed a product such as corn grown from a particular field), bioplastic containers, etc.)). In addition to harvested physical products, a product may also include a benefit or service provided via use of the associated land (for example, for recreational purposes such as a golf course), pastureland for grazing wild or domesticated animals (where domesticated animals may be raised for food or recreation). [0087] “Product metadata” are any information regarding an underlying product, its production, and/or its transaction which may be verified by a third party and may form the basis for issuance of an ecosystem credit and/or sustainability claim. Product metadata may include at least a product identifier, as well as a record of entities involved in transactions. [0088] As used herein, “quality” or a “quality metric” may refer to any aspect of an agricultural product that adds value. In some embodiments, quality is a physical or chemical attribute of the crop product. For example, a quality may include, for a crop product type, one or more of: a variety; a genetic trait or lack thereof; genetic modification of lack thereof; genomic edit or lack thereof; epigenetic signature or lack thereof; moisture content; protein content; carbohydrate content; ash content; fiber content; fiber quality; fat content; oil content; color; whiteness; weight; transparency; hardness; percent chalky grains; proportion of corneous endosperm; presence of foreign matter; number or percentage of broken kernels; number or percentage of kernels with stress cracks; falling number; farinograph; adsorption of water; milling degree; immature grains; kernel size distribution; average grain length; average grain breadth; kernel volume; density; L/B ratio; wet gluten; sodium dodecyl sulfate sedimentation; toxin levels (for example, mycotoxin levels, including vomitoxin, fumonisin, ochratoxin, or aflatoxin levels); and damage levels (for example, mold, insect, heat, cold, frost, or other material damage).

[0089] In some embodiments, quality is an attribute of a production method or environment. For example, quality may include, for a crop product, one or more of: soil type; soil chemistry; climate; weather; magnitude or frequency of weather events; soil or air temperature; soil or air moisture; degree days; rain fed; irrigated or not; type of irrigation; tillage frequency; cover crop (present or historical); fallow seasons (present or historical); crop rotation; organic; shade grown; greenhouse; level and types of fertilizer use; levels and type of chemical use; levels and types of herbicide use; pesticide-free; levels and types of pesticide use; no-till; use of organic manure and byproducts; minority produced; fair-wage; geography of production (e.g., country of origin, American Viti cultural Area, mountain grown); pollution-free production; reduced pollution production; levels and types of greenhouse gas production; carbon neutral production; levels and duration of soil carbon sequestration; and others. In some embodiments, quality is affected by, or may be inferred from, the timing of one or more production practices. For example, food grade quality for crop products may be inferred from the variety of plant, damage levels, and one or more production practices used to grow the crop. In another example, one or more qualities may be inferred from the maturity or growth stage of an agricultural product such as a plant or animal. In some embodiments, a crop product is an agricultural product.

[0090] In some embodiments, quality is an attribute of a method of storing an agricultural good (e.g., the type of storage: bin, bag, pile, in-field, box, tank, or other containerization), the environmental conditions (e.g., temperature, light, moisture, relative humidity, presence of pests, CO2 levels) during storage of the crop product, method of preserving the crop product (e.g., freezing, drying, chemically treating), or a function of the length of time of storage. In some embodiments, quality may be calculated, derived, inferred, or subjectively classified based on one or more measured or observed physical or chemical attributes of a crop product, its production, or its storage method. In some embodiments, a quality metric is a grading or certification by an organization or agency. For example, grading by the USDA, organic certification, or non-GMO certification may be associated with a crop product. In some embodiments, a quality metric is inferred from one or more measurements made of plants during growing season. For example, wheat grain protein content may be inferred from measurement of crop canopies using hyperspectral sensors and/or near infrared (NIR) or visible spectroscopy of whole wheat grains. In some embodiments, one or more quality metric is collected, measured, or observed during harvest. For example, dry matter content of corn may be measured using near-infrared spectroscopy on a combine. In some embodiments, the observed or measured value of a quality metric is compared to a reference value for the metric. In some embodiments, a reference value for a metric (for example, a quality metric or a quantity metric) is an industry standard or grade value for a quality metric of a particular agricultural good (for example, U.S. No. 3 Yellow Corn, Flint), optionally as measured in a particular tissue (for example, grain) and optionally at a particular stage of development (for example, silking). In some embodiments, a reference value is determined based on a supplier’s historical production record or the historical production record of present and/or prior marketplace participants.

[0091] A “field” is the area where agricultural production practices are being used (for example, to produce a transacted agricultural product) and/or ecosystem credits and/or sustainability claims.

[0092] As used herein, a “field boundary” may refer to a geospatial boundary of an individual field. [0093] As used herein, an “enrolled field boundary” may refer to the geospatial boundary of an individual field enrolled in at least one ecosystem credit or sustainability claim program on a specific date.

[0094] In some embodiments, a field is a unique object that has temporal and spatial dimensions. In some embodiments, the field is enrolled in one or more programs, where each program corresponds to a methodology. As used herein a “methodology” (equivalently “program eligibility requirements” or “program requirements”) is a set of requirements associated with a program, and may include, for example, eligibility requirements for the program (for example, eligible regions, permitted practices, eligible participants (for example, size of farms, types of product permitted, types of production facilities permitted, etc.) and or environmental effects of activities of program participants, reporting or oversight requirements, required characteristics of technologies (including modeling technologies, statistical methods, etc.) permitted to be used for prediction, quantification, verification of results by program participants, etc. Examples of methodologies include protocols administered by Climate Action Reserve (CAR) (climateactionreserve.org), such as the Soil Enrichment Protocol; methodologies administered by Verra (verra.org), such as the Methodology for Improved Agricultural Land Management, farming sustainability certifications, life cycle assessment, and other similar programs. In some embodiments, the field data object includes field metadata. “One or more methodologies” refers to a data structure comprising program eligibility requirements for a plurality of programs. More briefly, a methodology may be a set of rules set by a registry or other third party, while a program implements the rules set in the methodology.

[0095] In various embodiments, the field metadata includes a field identifier that identifies a farm (e.g., a business) and a farmer who manages the farm (e.g., a user). In various embodiments, the field metadata includes field boundaries that are a collection of one or more polygons describing geospatial boundaries of the field. In some embodiments, polygons representing fields or regions within fields (e.g., management event boundaries, etc.) may be detected from remote sensing data using computer vision methods (for example, edge detection, image segmentation, and combinations thereof) or machine learning algorithms (for example, maximum likelihood classification, random forest classification, support vector machine classification, ensemble learning algorithms, convolutional neural network, etc.). [0096] In various embodiments, the field metadata includes farming practices that are a set of farming practices on the field. In various embodiments, farming practices are a collection of practices across multiple years. For example, farming practices include crop types, tillage method, fertilizers and other inputs, etc. as well as temporal information related to each practice which is used to establish crop growing seasons and ultimately to attribute outcomes to practices. In various embodiments, the field metadata includes outcomes. In various embodiments, the outcomes include at least an effect size of the farming practices and an uncertainty of the outcome. In various embodiments, an outcome is a recorded result of a practice, notably: harvest yields, sequestration of greenhouse gases, and/or reduction of emissions of one or more greenhouse gases.

[0097] In various embodiments, the field metadata includes agronomic information, such as soil type, climate type, etc. In various embodiments, the field metadata includes evidence of practices and outcomes provided by the grower or other sources. For example, a scale ticket from a grain elevator, an invoice for cover crop seed from a distributor, farm machine data, remote sensing data, a time stamped image or recording, etc. In various embodiments, the field metadata includes product tracing information such as storage locations, intermediaries, final buyer, and tracking identifiers.

[0098] In various embodiments, the field object is populated by data entry from the growers directly. In various embodiments, the field object is populated using data from remote sensing (satellite, sensors, drones, etc.). In various embodiments, the field object is populated using data from agronomic data platforms such as John Deere and Granular, and/or data supplied by agronomists, and/or data generated by remote sensors (such as aerial imagery, satellite derived data, farm machine data, soil sensors, etc.). In various embodiments, at least some of the field metadata within the field object is hypothetical for simulating and estimating the potential effect of applying one or more practices (or changing one or more practices) to help growers make decisions as to which practices to implement for optimal economic benefit.

[0099] In various embodiments, the system may access one or more model capable of processing the field object, processing the field object (e.g., process the field object based on one or more model), and returning an output based on the metadata contained within the field object. In various embodiments, a collection of models that can be applied to a field object to estimate, simulate, and/or quantify the outcome (e.g., the effect on the environment) of the practices implemented on a given field. In various embodiments, the models may include process-based biogeochemical models. In various embodiments, the models may include machine learning models. In various embodiments, the models may include rule-based models. In various embodiments, the models may include a combination of models (e.g., ensemble models).

[0100] As used herein, a “management event” may refer to a grouping of data about one or more farming practices (such as tillage, harvest, etc.) that occur within a field boundary or an enrolled field boundary. A “management event” contains information about the time when the event occurred and has a geospatial boundary defining where within the field boundary the agronomic data about the event applies. Management events are used for modeling and credit quantification, designed to facilitate grower data entry and assessment of data requirements. Each management event may have a defined management event boundary that can be all or part of the field area defined by the field boundary. A “management event boundary” (equivalently a “farming practice boundary”) is the geospatial boundary of an area over which farming practice action is taken or avoided. In some embodiments, if a farming practice action is an action taken or avoided at a single point, the management event boundary is a point location. As used herein, a farming practice and agronomic practice are of equivalent meaning.

[0101] As used herein, a “management zone” may refer to an area within an individual field boundary defined by the combination of management event boundaries that describe the presence or absence of management events at any particular time or time window, as well as attributes of the management events (if any event occurred). A management zone may be a contiguous region or a non-contiguous region. A “management zone boundary” may refer to a geospatial boundary of a management zone. In some embodiments, a management zone is an area coextensive with a spatially and temporally unique set of one or more farming practices. In some embodiments, an initial management zone includes historic management events from one or more prior cultivation cycles (for example, at least 2, at least 3, at least 4, at least 5, or a number of prior cultivation cycles required by a methodology). In some embodiments, a management zone generated for the year following the year for which an initial management zone was created will be a combination of the initial management zone and one or more management event boundaries of the next year. A management zone can be a data-rich geospatial object created for each field using an algorithm that crawls through management events (e.g., all management events) and groups the management events into discrete zonal areas based on features associated with the management event(s) and/or features associated with the portion of the field in which the management event(s) occur. The creation of management zones enables the prorating of credit quantification for the area within the field boundary based on the geospatial boundaries of management events.

[0102] In some embodiments, a management zone is created by sequentially intersecting a geospatial boundary defining a region wherein management zones are being determined (for example, a field boundary), with each management event boundary occurring within, or overlapping at least partially with, that region at any particular time or time window, wherein each of the sequential intersection operations creates two branches - one with the intersection of the geometries and one with the difference. This is illustrated further in Fig. 14 - it will be appreciated that a third branch is implicit in the tree structure because all management event geometries we are using to subdivide the management zones at each level are contained within the root geometry (the field boundary). The new branches are then processed with the next management event boundary in the sequence, bifurcating whenever there is an area of intersection and an area of difference. This process is repeated for all management event boundaries that occurred in the geospatial boundary defining the region. The final set of leaf nodes in this branching process define the geospatial extent of the set of management zones within the region, wherein each management zone is non-overlapping and each individual management zone contains a unique set of management events relative to any other management zone defined by this process.

[0103] As used herein, a “zone-cycle” may refer to a single cultivation cycle on a single management zone within a single field, considered collectively as a pair that define a foundational unit (e.g., also referred to as an “atomic unit”) of quantification for a given field in a given reporting period. A field can be a zone cycle if there is only one management zone making up said field.

[0104] As used herein, a “baseline simulation” may refer to a point-level or polygonbased simulation of constructed baselines for the duration of the reported project period, using initial soil sampling at that point or polygon (following SEP requirements for soil sampling and model initialization) and management zone-level grower data (that meets SEP data requirements).

[0105] As used herein, a “with-project simulation” may refer to a point-level simulation of adopted practice changes at the management zone level that meet a methodology’s requirements for credit quantification. [0106] As used herein, a “field-level project start date” may refer to the start of the earliest cultivation cycle, where a practice change or already implemented practice was detected and attested by a grower depending on program requirements.

[0107] As used herein, a “required historic baseline period” may refer to years (in cultivations cycles approximating 365 day periods, not calendar years) of required historic information prior to the field-level project start date that must fit requirements of the data hierarchy in order to be modeled for credits. A number of required years is specified by the methodology, based on crop rotation and management.

[0108] As used herein, a “cultivation cycle” (equivalently a “crop production period” or “production period”) may refer to the period between the first day after harvest or cutting of a prior crop on a field or the first day after the last grazing on a field, and the last day of harvest or cutting of the subsequent crop on a field or the last day of last grazing on a field. For example, a cultivation cycle may be: a period starting with the planting date of current crop and ending with the harvest of the current crop, a period starting with the date of last field prep event in the previous year and ending with the harvest of the current crop, a period starting with the last day of crop growth in the previous year and ending with the harvest or mowing of the current crop, a period starting the first day after the harvest in the prior year and the last day of harvest of the current crop, etc. In some embodiments, cultivation cycles are approximately 365 day periods from the field-level project start date that contain completed crop growing seasons (planting to harvest/mowing, or growth start to growth stop). In some embodiments, cultivation cycles extend beyond a single 365 day period and cultivation cycles are divided into one or more cultivation cycles of approximately 365 days, optionally where each division of time includes one planting event and one harvest or mowing event.

[0109] As used herein, “historic cultivation cycles” may refer to the same definition as cultivation cycles, but for the period of time in the required historic baseline period.

[0110] As used herein, a “historic segments” may refer to individual historic cultivation cycles, separated from each other in order to use to construct baseline simulations.

[OHl] As used herein, “historic crop practices” may refer to crop events occurring within historic cultivation cycles.

[0112] As used herein, “baseline thread/parallel baseline threads” may refer to each baseline thread as a repeating cycle of the required historic baseline period, that begin at the management zone level project start date. The number of baseline threads equals the number of unique historic segments (e.g., one baseline thread per each year of the required historic baseline period). Each baseline thread begins with a unique historic segment and runs in parallel to all other baseline threads to generate baseline simulations for a with-project cultivation cycle.

[0113] As used herein, an “overlap in practices” may refer to unrealistic agronomic combinations that arise at the start of baseline threads, when dates of agronomic events in the concluding cultivation cycle overlap with dates of agronomic events in the historic segment that is starting the baseline thread. In this case, logic is in place based on planting dates and harvest dates to make adjustments based on the type of overlap that is occurring.

[0114] An “indication of a geographic region” is a latitude and longitude, an address or parcel id, a geopolitical region (for example, a city, county, state), a region of similar environment (e.g., a similar soil type or similar weather), a supply shed, a boundary file, a shape drawn on a map presented within a GUI of a user device, image of a region, an image of a region displayed on a map presented within a GUI of a user device, a user id where the user id is associated with one or more production locations (for example, one or more fields). [0115] For example, polygons representing fields may be detected from remote sensing data using computer vision methods (for example, edge detection, image segmentation, and combinations thereof) or machine learning algorithms (for example, maximum likelihood classification, random forest classification, support vector machine classification, ensemble learning algorithms, convolutional neural network, etc.).

[0116] “Ecosystem observation data” are observed or measured data describing an ecosystem, for example weather data, soil data, remote sensing data, emissions data (for example, emissions data measured by an eddy covariance flux tower), populations of organisms, plant tissue data, and genetic data. In some embodiments, ecosystem observation data are used to connect agricultural activities with ecosystem variables. Ecosystem observation data may include survey data, such as soil survey data (e.g., Soil Survey Geographic Database (SSURGO)). In various embodiments, the system performs scenario exploration and model forecasting, using the modeling described herein. In various embodiments, the system proposes climate-smart crop fuel feedstock carbon intensity (CI) integration with an existing model, such as the Greenhouse gases, Regulated Emissions, and Energy use in Technologies Model (GREET), which can be found online at https://greet.es.anl.gov/ (the GREET models are incorporated by reference herein). [0117] A “crop type data layer” is a data layer containing a prediction of crop type, for example USDA Cropland Data Layer (CDL) provides annual predictions of crop type, and a 30m resolution land cover map is available from MapBiomas (https://mapbiomas.org/en). A crop mask may also be built from satellite-based crop type determination methods, ground observations including survey data or data collected by farm equipment, or combinations of two or more of: an agency or commercially reported crop data layer (e.g., CDL), ground observations, and satellite-based crop type determination methods.

[0118] A “vegetative index” (“VI”) is a value related to vegetation as computed from one or more spectral bands or channels of remote sensing data. Examples include simple ratio vegetation index (“RVI”), perpendicular vegetation index (“PVI”), soil adjusted vegetation index (“SAVI”), atmospherically resistant vegetation index (“AR VI”), soil adjusted atmospherically resistant VI (“SARVI”), difference vegetation index (“DVI”), normalized difference vegetation index (“ND VI”). ND VI is a measure of vegetation greenness which is particularly sensitive to minor increases in surface cover associated with cover crops.

[0119] SEP” stands for soil enrichment protocol. The SEP version 1.0 and supporting documents, including requirements and guidance, (incorporated by reference herein) can be found online at https://www.climateactionreserve.org/how/protocols/soil-enrichment/. As is known in the art, SEP is an example of a carbon registry methodology, but it will be appreciated that other registries having other registry methodologies (e.g., carbon, water usage, etc.) may be used, such as the Verified Carbon Standard VM0042 Methodology for Improved Agricultural Land Management, vl.O (incorporated by reference herein), which can be found online at https://verra.org/methodology/vm0042-methodology-for-improved- agri cultural -land-management-v 1-0/. The Verified Carbon Standard methodology quantifies the greenhouse gas (GHG) emission reductions and soil organic carbon (SOC) removals resulting from the adoption of improved agricultural land management (ALM) practices.

Such practices include, but are not limited to, reductions in fertilizer application and tillage, and improvements in water management, residue management, cash crop and cover crop planting and harvest, and grazing practices.

[0120] “LRR” refers to a Land Resource Region, which is a geographical area made up of an aggregation of Major Land Resource Areas (MLRA) with similar characteristics.

[0121] DayCent is a daily time series biogeochemical model that simulates fluxes of carbon and nitrogen between the atmosphere, vegetation, and soil. It is a daily version of the CENTURY biogeochemical model. Model inputs include daily maximum/minimum air temperature and precipitation, surface soil texture class, and land cover/use data. Model outputs include daily fluxes of various N-gas species (e.g., N2O, N0_x, N2); daily CO2 flux from heterotrophic soil respiration; soil organic C and N; net primary productivity; daily water and nitrate (NO3) leaching, and other ecosystem parameters.

SYSTEM OVERVIEW

[0122] Figure (FIG.) l is a block diagram that illustrates an ecosystem management environment 100, in accordance with one or more embodiments. The ecosystem management environment 100 includes a computing device 102, a data store 104, a sensing device 106, a user device 108, and a network 109. The entities and components in the ecosystem management environment 100 may communicate with each other through a network 109. In various embodiments, the ecosystem management environment 100 includes fewer or additional components. In some embodiments, the ecosystem management environment 100 also includes different components. While each of the components in the ecosystem management environment 100 is described in a singular form, the ecosystem management environment 100 may include one or more of each of the components. For example, in many situations, the computing device 102 may receive information form a plurality of sensing devices 106. Different user devices 108 may also access the computing device 102 simultaneously.

[0123] The computing device 102 may include one or more processors/computers that perform various tasks related to ecosystem management. In some embodiments, the computing device 102 includes a data object creator 120, an experiment module 140, a database 150, and a simulator 170. The computing device 102 may receive data from one or more data sources through the data store 104, the sensing device 106, user device 108, and/or the network 109.

[0124] The data object creator 120 generates data objects using the received data. A data object is associated with a management event and describes a management zone that includes geospatial boundary defining an agronomic region. The agronomic region may be, e.g., a region of a farm, a field, one or more farms, one or more fields, etc.

[0125] The experiment module 140 creates an experimental data object for a candidate agronomic region. A candidate agronomic region is an agronomic region for which a user instructs to simulate farming practices and determine their results. An experimental data object is used to specify a target experiment included in the user’s instruction. The experimental data object may be applied with a data object to modify one or more attributes that identify the corresponding management event. The experiment module 140 may access an experimental template which may be stored in the database 150. The experiment template provides a construct for modifying the management event and attributes of the candidate agronomic region.

[0126] The simulator 170 receives the experimental data object and a candidate data object (i.e., the data object of the candidate agronomic region) as input to one or more models to generate a respective predicted ecosystem attribute. The simulator 170 may estimate the potential effect of changing one or more management events and/or attributes defining the management events (or changing one or more practices) to help users make decisions as to which practices to implement for improving ecosystem benefit. For example, the simulator 170 may run a first simulation on a data object and run a second simulation on the data object with a corresponding experimental data object. The experimental data object specifies a target experiment to modify one or more attributes of a farming practice included in the data object. The simulator 170 may obtain and compare the results from the first and second simulations. Based on the comparison, the simulator 170 may determine the impact of the experimental data object, e.g., changes in the result due to the target experiment. In some embodiments, the simulator 170 may assign an impact score to an experimental data object, indicating a measure of magnitude of impact of farming activity on the environment.

[0127] The ecosystem impact of farming practices on the candidate locations may be determined by performing simulations of different farming practices using one or more models. The simulations and modeling provide a virtual environment to test different experimental scenarios and understand the complex interactions within an agricultural ecosystem. Using the experimental template, the system may clearly define the objectives of the simulation study and identify the key parameters and variables that influence the outcomes of interest, such as crop yield, resource use efficiency, or environmental impact. The simulations and modeling may predict various ecosystem attributes over multiple seasons or time steps to observe how the ecosystem responds to different inputs and management practices. In this way, input variables in experimental farming practices that lead to desirable outcomes of an ecosystem may be identified. The desirable outcomes, whether for the purposes of ecosystem credits or other environmental considerations may include, e.g., increased soil organic carbon, or reduced greenhouse gas emissions, reduced water usage, reduced chemical contamination (e.g., reduced nitrogen run-off, reduced insecticide/pesticide/herbicide residue), increased biodiversity). [0128] In some embodiments, the received data may be agronomic data (also referred to as agricultural data) that describes information of a particular agronomic region at a specific time. The agronomic data may include information, such as, attributes of the farm field, environmental conditions, farming practice performed on the farm field, and the like. In some embodiments, the agronomic data may be made available in different formats (or methods). For example, the agronomic data can be obtained in file formats such as JavaScript Object Notation (JSON), yet another markup language (YAML), agronomic data store (ADS), comma-separated values (CSV) via one or more applications programming interfaces (API) (as shown in FIG. 2).

[0129] In some embodiments, the computing device 102 may receive the agronomic data from the user device 108, for example, input by growers of the fields. In some embodiments, the computing device 102 may receive the agronomic data from the sensing device 106, for example, via remote sensing (satellite, sensors, drones, etc.). In some embodiments, the computing device 102 may receive the agronomic data from a data store 104, for example, agronomic data platforms such as John Deere and Granular, and/or data supplied by agronomists, and/or data generated by remote sensors (such as aerial imagery, satellite derived data, farm machine data, soil sensors, etc.). Exemplary remote sensing algorithms are provided in Publication Nos. WO 2021/007352, WO 2021/041666, WO 2021/062147, and WO 2022/020448, which are hereby incorporated by reference.

[0130] The data object creator 120 creates one or more data objects using the received agronomic data. In some embodiments, a data object may be used to describe a management zone that includes geospatial boundary defining an agronomic region. For example, the management events may include the planting of different types of crops, tillage, chemical application such as use of fertilizers and other inputs, harvest yields, crop termination, sequestration of greenhouse gases, and/or reduction of emissions of one or more greenhouse gases. In some embodiments, a data object describes a particular agronomic region at a time or within a time window. In some embodiments, the data object is associated with a farming event on a field at a given time. For example, the data object may include a field identifier that identifies a particular agronomic region, a date (or a time period), a farming event that was conducted on a field identified by the field identifier and one or more attributes describing the farming event. A farming event may refer to farming activities/practices such as planting, harvesting, termination, and the like. In some embodiments, the data object may include the field identifier of the agronomic region, dates and multiple farming practices that were conducted on the agronomic region in a given period of time.

[0131] In some embodiments, the management event may include a farming practice on a agronomic region, and the one or more attributes of the management event correspond to one or more parameters of the farming practice. Each attribute may include one or more values describing the corresponding parameters of the farming practice. For example, for a planting practice, the corresponding parameters may include the kind of crop that is planted, whether the crop is perennial, and whether the crop is covered, and the like. For a harvesting practice, the corresponding parameters may include the yield of the crop, the harvesting method, whether the farm field is burnt, and the like.

[0132] The experiment module 140 receives a data object and generates an experimental data object. The experimental module 140 may receive a request from a user to conduct experiments on a candidate location. The request may include instructions on performing simulations using one or more models. In some embodiments, the model(s) may be a machine learning model, but may also be a process based biogeochemical model, an inventor -based greenhouse gas emissions calculator, a statistical model or some other type of model. The experiment module 140 may create one or more experimental data objects to specify one or more experiments as per the user instructions. An experiment may refer to one or more target farming activities. In some embodiments, an experiment may add, remove, and/or modify a management event on an agronomic region. In some embodiments, each experiment predicts an ecosystem attribute; alternatively, multiple ecosystem attributes may be predicted with one experiment.

[0133] In some embodiments, the experiment module 140 may access an experimental template that provides a construct for altering the management events (e.g., adding, removing, modifying a farming practice) on a candidate location and one or more attributes describing the management events on the candidate location during simulations, and one or more models to be applied, a model version for the one or more models, one or more sets of model parameters, a life cycle inventory' database to use, and one or more default equations to use.

[0134] In other words, the experimental template provides a construct for running virtual experiments on agronomic regions using data objects within the management environment. The experimental data object may modify one or more attributes that identify the corresponding management event. In some embodiments, the instructions may be stored in the experimental data object along with the field identifier, field metadata and the information regarding one or more simulations that is needed to be performed. In some embodiments, the experimental template 130 may be determined using one or more models which are trained using historical agronomic data. In some embodiments, the instructions and/or experimental templates may be stored in the database 150.

[0135] The simulator 170 inputs the experimental data object and the candidate data objects to one or more models to generate a respective predicted ecosystem attribute, such as greenhouse gas emissions, crop production, etc. In some examples, the ecosystem attributes may include information related to water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production, etc. In some embodiments, the model may be a machine learning model. In some embodiments, the models may include process-based biogeochemical models. In some embodiments, the models may include rule-based models. In some embodiments, the models may include a combination of models (e.g., ensemble models). The software system can compare the predicted ecosystem attributes with the baseline results to assign an impact score to the respective experimental data object. In some embodiments, an impact score is a measure of magnitude of the impact of farming activity on the environment. In various embodiments, the impact score can be a negative or positive indication of the impact of farming activity on the environment.

[0136] In some embodiments, The training of the machine-learned models described herein (such as neural networks and other models referenced herein) include the performance of one or more non-mathematical operations or implementation of non-mathematical functions at least in part by a machine or computing system, examples of which include but are not limited to data loading operations, data storage operations, data toggling or modification operations, non-transitory computer-readable storage medium modification operations, metadata removal or data cleansing operations, data compression operations, image modification operations, noise application operations, noise removal operations, and the like. Accordingly, the training of the machine-learned models described herein may be based on or may involve mathematical concepts, but is not simply limited to the performance of a mathematical calculation, a mathematical operation, or an act of calculating a variable or number using mathematical methods.

[0137] Likewise, it should be noted that the training of the models described herein cannot be practically performed in the human mind alone. The models are innately complex including vast amounts of weights and parameters associated through one or more complex functions. Training and/or deployment of such models involves so great a number of operations that it is not feasibly performable by the human mind alone, nor with the assistance of pen and paper. In such embodiments, the operations may number in the hundreds, thousands, tens of thousands, hundreds of thousands, millions, billions, or trillions. Moreover, the training data may include hundreds, thousands, tens of thousands, hundreds of thousands, or millions of temperature measurements. Accordingly, such models are necessarily rooted in computer-technology for their implementation and use.

[0138] In some embodiments, the computing device 102 may include a database 150 which includes agronomic data, data objects, experimental templates, experimental data objects, models, etc. Examples of components and functionalities of the computing device 102 are discussed in further detail below with reference to FIG. 2.

[0139] The data store 104 includes memory or other storage media for storing various files and data which are accessible to the computing device 102. The data stored in the data store 104 includes agronomic data, data objects, experimental templates, experimental data objects, models, etc. In various embodiments, the data store 104 may take different forms. In one embodiment, the data store 104 is part of the computing device 102. For example, the data store 104 is part of the local storage (e.g., hard drive, memory card, data server room) of the computing device 102. In some embodiments, the data store 104 is a network-based storage server (e.g., a cloud server). The data store 104 may be a third-party storage system/platform.

[0140] The sensing device 106 collects/observes agronomic data for a particular agronomic region at a time or within a time window. In some embodiments, the sensing device 106 may include one or more sensors, such as, soil probes, land-based vehicles (e.g., tractors, planters, trucks, robots), hand-held devices (e.g., a cell phones, cameras, spectrophotometers), drones, airplanes, and satellites. In some embodiments, the sensing device 106 includes a “field sensor” operated within a field boundary, for example, a soil moisture sensor, a flux tower (for example, a micrometeorological tower to measure the exchanges of carbon dioxide, water vapor, and energy between the biosphere and atmosphere), a soil temperature sensor, an air temperature sensor, a pH sensor, a nitrogen sensor, an irrigation system, a tractor, a robot, a vehicle, etc. In some embodiments, preliminary field data are automatically populated based on average practices and average practice dates within a region (for example as detected based on current season or historical remote sensing data analysis).

[0141] The user device 108 is a computing device that is used by a user. A user may use the user device 108 to communicate with the computing device 102 and performs ecosystem management related operations. In some embodiments, a user device 108 may include one or more applications and interfaces that may display visual elements of the applications. In some embodiments, preliminary data may be verified by input received from a farmer’s user device 108. For example, preliminary data may be presented and verified within a graphical user interface of a farmer’s user device 108. In some implementations, preliminary data may be verified by location and or accelerometer data or other data collected from a user device 108. For example, a harvest practice identified by remote sensing data may be confirmed where machine data corresponding the typical engine speed of a harvester is recorded between the periodic images within a remote sensing time series collected from a satellite, where the first of that time series period does not indicate harvest has occurred and the next image indicates that harvest has occurred or is in progress. The user device 108 may be any computing device. Examples of such user device 108 include personal computers (PC), desktop computers, laptop computers, tablets (e.g., iPADs), smartphones, wearable electronic devices such as smartwatches, farm equipment such as a drone or tractor, or any other suitable electronic devices. Other data collected from a user device may include a machine data (such as engine rpm, fuel level, location, machine hours, and changes in the same), input usage (for example, amounts and types of seeds, fertilizers, chemicals, water, applied), imagery and sensor data (for example, photographs, videos, LiDAR, infrared).

[0142] The network 109 provides connections to the components of the ecosystem management environment 100 through one or more sub-networks, which may include any combination of local area and/or wide area networks, using both wired and/or wireless communication systems. In one embodiment, a network 109 uses standard communications technologies and/or protocols. For example, a network 109 may include communication links using technologies such as Ethernet, 802.11, worldwide interoperability for microwave access (WiMAX), 3G, 4G, Long Term Evolution (LTE), 5G, code division multiple access (CDMA), digital subscriber line (DSL), etc. Examples of network protocols used for communicating via the network 109 include multiprotocol label switching (MPLS), transmission control protocol/Internet protocol (TCP/IP), hypertext transport protocol (HTTP), simple mail transfer protocol (SMTP), and file transfer protocol (FTP). Data exchanged over a network 109 may be represented using any suitable format, such as hypertext markup language (HTML), extensible markup language (XML), JavaScript object notation (JSON), structured query language (SQL). In some embodiments, some of the communication links of a network 109 may be encrypted using any suitable technique or techniques such as secure sockets layer (SSL), transport layer security (TLS), virtual private networks (VPNs), Internet Protocol security (IPsec), etc. The network 109 also includes links and packet switching networks such as the Internet. In some embodiments, a data store belongs to part of the internal computing system of a server (e.g., the data store 104 may be part of the computing device 102). In such cases, the network 109 may be a local network that enables the server to communicate with the rest of the components.

PREDICTING ECOSYSTEM ATTRIBUTE

[0143] FIG. 2 is a data flow diagram of an example process 200 for generating a predicted ecosystem attribute, in accordance with one or more embodiments . The process 200 described in the present disclosure can be implemented as a software system by one or more computers located in one or more geographical locations. The process 200 can include multiple sub-processes implemented by various components of the software system. The process 200 can be implemented as an online service wherein the service can be accessed using a user computing device such as a smartphone, personal computer (PC), laptop, etc. The software implementing the process 200 can include a user interface (UI) that allows interaction between the user and the process 200.

[0144] Various embodiments described herein use machine learned models. In some embodiments, the process may include using artificial neural networks. Suitable artificial neural networks include but are not limited to a feedforward neural network, a radial basis function network, a self-organizing map, learning vector quantization, a recurrent neural network, a Hopfield network, a Boltzmann machine, an echo state network, long short term memory, a bi-directional recurrent neural network, a hierarchical recurrent neural network, a stochastic neural network, a modular neural network, an associative neural network, a deep neural network, a deep belief network, a convolutional neural networks, a convolutional deep belief network, a large memory storage and retrieval neural network, a deep Boltzmann machine, a deep stacking network, a tensor deep stacking network, a spike and slab restricted Boltzmann machine, a compound hierarchical-deep model, a deep coding network, a multilayer kernel machine, or a deep Q-network. [0145] As shown in FIG. 2, the computing device 102 receives input data 110. In some embodiments, the input data 110 includes agronomic data which may be in various file formats such as JSON 110-1, YAML 110-2, ADS 110-3, CSV 110-4. In some embodiments, the input data 110 may include synthetic data. In some embodiments, the synthetic data of soil measurements and management practices may be provided, in combination with observed data or in place of observed data.

[0146] In some embodiments, the input data 110 may include a field identifier that identifies a farm (e.g., a business) and/or a farmer who manages the farm (e.g., a user). In some embodiments, the input data 110 includes field boundaries that are a collection of one or more polygons describing geospatial boundaries of the field. In some embodiments, polygons representing fields may be detected from remote sensing data using computer vision methods (for example, edge detection, image segmentation, and combinations thereof) or machine learning algorithms (for example, maximum likelihood classification, random forest classification, support vector machine classification, ensemble learning algorithms, convolutional neural network, etc.).

[0147] In some embodiments, the input data 110 includes management events that describe farming practices on the field. In some embodiments, the management events include farming practices across multiple years. For example, management events include the planting of different types of crops, tillage, chemical application such as use of fertilizers and other inputs, harvest yields, crop termination, sequestration of greenhouse gases, and/or reduction of emissions of one or more greenhouse gases. These management events may be described using temporal information related to each event which is used to establish crop growing seasons and ultimately to attribute outcomes of the management events.

[0148] For example, the farming event of planting a given crop on a field can be described using a type of crop that is planted on the field, a date when the crop was planted on the field, and/or whether the crop that was planted is a cover crop or a crop for harvesting. As another example, the farming event of harvesting may be described using the type of crop that is going to be harvested, a date when the harvesting is performed, whether the crop is a cover-crop, a method of harvesting used, a total yield of harvesting the crop, and/or an indication of whether the leftover straws after harvesting require burning. As another example, the farming event of termination may be described using the type of crop, a date of crop termination, whether the crop that is being terminated is a cover crop, and/or methods used for the termination of crop. [0149] In one example, the farming event of tillage can be described using a date when tillage is performed on the field, one or more methods of tillage, and/or equipment used to perform tillage. In another example, the farming event of chemical application (e.g., a fertilizer) may be described using a date when chemicals are applied on the fields, an application rate of the chemicals, a nitrate fraction of the chemicals, an urea fraction of the chemicals, an anhydrous ammonia fraction of the chemicals, an ammonium fraction of the chemicals, a nitrogen percentage in the chemicals, a phosphorus pentoxide percentage in the chemicals, a potassium oxide percentage in the chemicals, a chemical identifier, a rate of chemical release of the applied chemical, an indication of nitrification inhibition, and/or an indication of urea inhibition. In still another example, the farming event of soil amendments (e.g., organic amendment) can be described using a date of applying organic material, one or more methods of applying organic material, and/or a rate of applying organic material. In some embodiments, soil amendments can also involve inorganic or mineral amendments. Such amendments may be described using the minerals used for the soil amendment and the rate at which the amendment was performed.

[0150] In some embodiments, the input data 110 may also include soil attributes, such as soil composition. Soil composition may be described using SOC stock, an indication of hydricity, soil bulk density, coarse fragments, sand, clay, organic fraction, and/or pH. In some embodiments, the soil attributes may describe different layers of soil where each layer of the soil may be described using the depth of the layer in the soil. In some embodiments, the input data 110 includes information, such as soil type, climate type, etc. In some embodiments, the agronomic data includes evidence of practices and outcomes provided by the grower or other sources. For example, a scale ticket from a grain elevator, an invoice for cover crop seed from a distributor, farm machine data, remote sensing data, a time stamped image or recording, etc. In some embodiments, the field metadata includes product tracing information such as storage locations, intermediaries, final buyer, and tracking identifiers. [0151] The data object creator 120 receives the input data 110 and create data objects 122 based on the agronomic data from the input data 110. In some embodiments, the created data objects 125 may be one or more management zones. In some embodiments, a sub-process to create data objects 122 may include sequentially intersecting a geospatial boundary defining a region wherein management zones are being determined, with one or more geospatial management event boundary occurring within or at least partially overlapping with that region at a time or within a time window. In an embodiment, the data objects 125 are immutable, and do not change once they have been created. In some embodiments, the data objects 125 describe a farming event on a field at a given time. For example, a data object can include a field identifier, a date (or a time period), a farming event that was conducted on a field identified by the field identifier and one or more attributes describing the farming event.

[0152] To further illustrate a data object, consider an example of a field that has undergone three management events - planting, harvesting and termination. In such an example, the data object creator 120 may create three data objects 125 where the first data object may include a field identifier of the field that uniquely identifies the field among multiple other fields, the farming activity of planting, a date when planting was performed on the field and one or more attributes describing the farming event of planting. Similarly, the second data object may include the field identifier of the field, the farming activity of harvesting, a date when harvesting was performed on the field and one or more attributes describing the farming event of harvesting. Likewise, the third data object can include the field identifier of the field, the farming activity of termination, a date when termination was performed on the field and one or more attributes describing the farming event of termination. In some embodiments, the data object 125 may include the field identifier, the date and multiple management events that were conducted on the field in a given period of time. For example, the data object creator 120 may create a single data object 125 rather than three data objects. In such embodiments, the data object 125 may list all three management events - planting, harvesting and termination along with the one or more attributes describing the management events (various embodiments may combine or separate events based on constraints imposed by a given downstream model). This is further explained with reference to Fig. 3.

[0153] In some embodiments, the data objects 125 are provided to a model as described below with regard to Fig. 7 for the generation of synthetic agronomic data. Such synthetic datasets may in turn be stored for future retrieval and processing by as described above with regard to data object creator 120. Such synthetic datasets may also be directly provided to data object creator 120 for creation of additional data objects 125 corresponding to the generated synthetic data for further processing as set forth below.

[0154] While various examples provided herein focus on soil samples, it will be appreciated that the techniques provided are applicable to additional measurements associated with a geo-spatial boundary (point or polygon). Examples include measurements of soil, canopy temperature, water level, or an average of some measurement across a field or management zone.

[0155] The experiment module 140 receives a request from a user device of the user. The request may include instructions to conduct experiments by performing simulations using one or more models. As shown in FIG. 2, the experiment module 140 may include an experiment scheduler 135 and an experiment creator 142. The experiment scheduler 135 receives instructions from a user input. The instruction may specify a candidate agronomic region (e.g., candidate location) on which a user instructs to perform simulation with the experiment. The experiment may refer to one or more target farming activities. The experiment scheduler 135 instructs the experiment creator 142 to create a plurality of experimental data objects 155 for the candidate locations. The experimental data object 155 specify one or more experiments (e.g., target farming activities) as per the user instructions. The data object of the candidate locations may be referred to as candidate data objects 160. In some embodiments, each experiment predicts an ecosystem attribute. For example, the experiment scheduler 135 specifies four experiments 140-1 to 140-4. In this example, the experiment scheduler 135 submits a request for the first experiment that predicts a 1^st ecosystem attribute. Similarly, the experiment scheduler 135 submits three more requests for experiments that would predict the 2^nd, 3^rd and 4^th ecosystem attributes respectively.

[0156] The experiment creator 142 may execute a process (e.g., create experiment 145) to generate experimental data objects 155. In some embodiments, the experiment creator 142 may access to one or more experimental templates 130 for generating the experimental data object 155. The experimental template 130 may be stored in the database 150. In some embodiments, the experimental template 130 may include a construct for altering (e.g., adding, removing, modifying, etc.) the management events and one or more attributes describing the management events of the candidate data objects 160. In some implementations, the experimental template 130 may specify the target farming activity to performed at the candidate location, and one or more attributes that describe the target activity and/or the candidate location.

[0157] In some embodiments, the experimental template 130 may be determined using one or more models which are trained using historical agronomic data. The models may identify the attributes/parameters that are associated with the experiment and include the identified attributes/parameters in the experimental template 130. For example, for a harvesting practice experiment, the model may determine attributes such as, yield of the crop, the harvesting method, etc., are associated with the harvesting practice and include these attributes in the experimental template 130. In another example, the model may determine a correlation between one or more attributes. For example, a solar condition may be correlated with the time of year, the latitude of the farm field, etc. The model may add these parameters in the experimental template 130 so that the user may modify the parameters and/or the simulator 170 may use these parameters as input to the models.

[0158] In one example, three management events of planting, chemical application and harvesting have been implemented on a candidate location. The candidate data object 160 for the candidate location may include a field identifier, along with the three management events of planting, chemical application and harvesting in the year 2022 along with one or more attributes describing each of the management events. For instance, the computing device 102 may receive a user instruction to predict the ecosystem attribute of the quantity of nitrous oxide produced, if a different chemical (e.g., fertilizer) was applied instead of the one that was used. The experiment scheduler 135 may generate a corresponding experimental data object 155 that alters the attribute associated with the chemical previously used with the different chemical.

[0159] In some embodiments, the experimental data objects 155 can include instructions using a domain specific language (DSL) to describe the altering of the management events and one or more attributes describing the management events. In some embodiments, the DSL instructions can be provided by the user. In other embodiments, the experiment creator 142 can generate the DSL instructions based on inputs received from the user. For example, the user can use a user interface (UI) provided by the software system to select management events and the attributes describing the management events that needs to be altered in the simulations. Likewise, the user can also provide instructions to alter the management events and attributes using the UI. Details of using DSL to create experimental data objects 155 are further explained with reference to Figs. 4 A and 4B.

[0160] The simulator 170 receives the experimental data object 155 and the candidate data object 160 as an input to one or more models to generate respective predicted ecosystem attributes. In some embodiments, the models may include process-based biogeochemical models. In some embodiments, the models may include machine learning models. In some embodiments, the models may include rule-based models. In some embodiments, the models may include a combination of models (e.g., ensemble models). It will be appreciated that a given model may retrieve weather conditions from various data providers based on the location associated with the provided data objects.

[0161] In some embodiments, one or more models may be applied to a candidate data object to estimate, simulate, and/or quantify the outcome (e.g., the effect on the environment) of the practices implemented on a candidate location. In some embodiments, the simulator 170 may select the models for predicting ecosystem attributes based on the experiments specified by the experiment scheduler 135. For example, the simulator 170 may include ten models where each model predicts a respective ecosystem attribute. In some embodiments, the experiment scheduler 135 specifies four experiments 140-1 to 140-4 and each of the experiments 140-1 to 140-4 generates a predicted ecosystem attribute (e.g., 170-1 to 170-4) that is specified by the user. The simulator 170 outputs the result data 190 that includes one or more predicted ecosystem attributes, e.g., 190-1 to 190-4. In some embodiments, the ecosystem attributes may include reduced water use, reduced nitrogen use, increased soil carbon sequestration, greenhouse gas emission avoidance, increased yield, reduced nitrous oxide production, etc.

[0162] In some embodiments, the simulator 170 generates the predicted ecosystem attributes for the candidate location specified by the experimental data object 155 (i.e., the field specified by the user) along with the predicted ecosystem attributes of the nearby fields specified by the candidate data objects 160. In such embodiments, one or more data objects 125 of the fields that are geographically located near the candidate location is provided as input to the simulator 170. The simulator 170 may average the predicted ecosystem attributes of the candidate location and the respective predicted ecosystem attributes of the nearby fields to as to generate a statistically consistent result.

[0163] In some embodiments, the computing device 102 may, based on the predicted ecosystem attributes, assign an impact score to the respective experimental data object 155. In on implementation, the computing device 102 may compare the predicted ecosystem attributes with baseline results 180 that were collected from the field to assign an impact score to the respective experimental data object 155. In some embodiments, an impact score is a measure of magnitude of the negative impact of farming activity on the environment. In other embodiments, an impact score can be a negative or positive indication of the impact of farming activity on the environment. For example, the impact score may be a relative amount of ecosystem credits relative to a baseline. The baseline results 180 includes baseline results 180-1 to 180-4 of the four ecosystem attributes that represent the true results of the management events conducted on the field. In some embodiments, the impact score is a difference in ecosystem credits yielded by a given program. For example, the credits for a baseline and counterfactual scenario may be determined from an issuer and compared to determine the impact score.

[0164] In some embodiments, the difference between the baseline results 180 and the predicted ecosystem attributes determines the ecosystem impact of altered management events on a candidate field. For example, assume that the user wants to predict the ecosystem attribute of the nitrous oxide production if a new type of agrichemical is applied to a field. The experiment creator 142 can generate an experimental data object 155 for the candidate location and the simulator 170 can use the experimental data object 155 and the candidate data object 160 to execute simulations with the new type of agrichemical. The simulator 170 selects the model that predicts nitrous oxide production and provides the experimental data object 155 of the first type and the candidate data object 160 as input to the model. The model generates a predicted quantity of nitrous oxide production which is compared to the baseline result 180 of nitrous oxide production. If the predicted quantity of nitrous oxide production is less than the baseline result 180, the computing device 102 may determine that the new type of chemical has a lower environmental impact than the chemical that was previously used. If the predicted quantity of nitrous oxide production is more than the baseline result 180, the computing device 102 may determine that the new type of chemical has a greater environmental impact than the agrichemical that was previously used.

[0165] The simulator 170 may process each of the plurality of experimental data objects 155 using one or more models to generate a respective predicted ecosystem attribute. To assign impact scores to each of the plurality of experimental data objects 155, each respective predicted ecosystem attribute is compared to the respective ecosystem attribute of the baseline result 180. For example, if a permutation of management events and one or more attributes specified by a first experimental data object produces more nitrous oxide than the baseline results 180 and the second experimental data object produces less nitrous oxide than the baseline results 180, the first experimental data object is assigned a higher impact score than the second experimental data object.

[0166] In some embodiments, the impact scores can be assigned to experimental data objects 155 based on multiple predicted ecosystem attributes. For example, the simulator 170 can use the experimental data object 155 to predict the quantity of nitrous oxide generated, a predicted quantity of water consumption and a predicted quantity of crop yield for a candidate location. If the predicted quantity of nitrous oxide generated is less than the quantity of nitrous oxide in the baseline result 180, the experimental data object 155 can be assigned a lower score signifying a lower impact on the environment. However, if the predicted quantity of water consumption is more than the water consumption of the baseline result 180, the experimental data object 155 can be assigned a higher score signifying a higher impact on the environment. In such situations, an overall impact score based on multiple predicted ecosystem attributes can be assigned to the experimental data objects 155. For example, if the multiple models form an ensemble model, the outputs of the individual models can be combined into a single predicted value. In such embodiments, the baseline result 180 can also be a single value based on the one or more ecosystem attributes collected from the fields. In such embodiments, impact scores are assigned to the experimental data objects 155 based on the difference between the predicted value of the ensemble and the single value of the baseline 180.

[0167] In some embodiments, the computing device 102 may assign the impact scores based on an average of the multiple predicted ecosystem attributes. For example, the simulator 170 can use multiple models and generate a respective predicted ecosystem attribute. The simulator 170 can then compute an average of the multiple predicted ecosystem attributes and use the average to assign the impact scores. In such embodiments, the baseline result 180 is also an average value of the one or more ecosystem attributes collected from the fields.

[0168] In some embodiments, the software system creates an immutable association between one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts, associated with an experimental template. Such an immutable association allows confirmation and reproduction of predicted ecosystem attributes and or ecosystem impacts.

[0169] In some embodiments, the software system can recommend environmentally friendly farming practices to the user. For example, the software system can select an experimental data object 155 of the first type that has been assigned the lowest impact score. The selected experimental object 155 is also referred to as a first experimental object. A low impact score of the first experimental score means that the management events and one or more attributes describing the management events specified by the first experimental object have a low impact on the environment according to the models. The software system can therefore make recommendations to the user about management events and the attributes describing the management events of the first experimental object. For example, assume that the models predicts that quantity of nitrous oxide generated because of a new type of chemical is less than the baseline result 180, the corresponding experimental data object 155 will be assigned a lower score signifying a lower impact on the environment. In such a case, the software system can recommend the use of the new chemical on the field.

[0170] In some embodiments, the software system can select a subset of experimental data object 155 of the first type based on the impact scores. For example, assume that the simulator 170 processes multiple experimental data objects 155 using one or more models to generate multiple ecosystem attributes. In such embodiments, the software system can select more than one experimental data objects 155 based on the respective impact sores and recommend the user with farming practices based on the selected experimental data objects. For example, the software system can select two experimental data objects 155 that were assigned the least impact scores. After selecting the two experimental data objects 155, the software system can recommend the user with farming practices of the two selected experimental data objects 155 that were altered in simulations. It is understood that a recommendation may also be generated based on an increased value for a beneficial ecosystem system attribute, for example an increase in soil carbon sequestration base on two or more selected experimental data objects 155 that were altered in simulations.

[0171] The user can accept the recommendation provided by the software system to produce verifiable environmental characteristics (for example, increasing the soil organic carbon and or reduce greenhouse gas emissions, reduced water usage, reduced chemical contamination (e.g., reduced nitrogen run-off, reduced insecticide/pesticide/herbicide residue, increased biodiversity, and the like)).

[0172] In some embodiments, the user may be interested in predicting the one or more ecosystem attribute based on the current management events and one or more attributes describing the management events. In such embodiments, the user can leverage the existing system to predict the ecosystem attributes. In such embodiments, the user can use the experimental template 130 to generate an experimental data object of a second type. The experimental data object of the second type includes management events and attributes that have been implemented on the field or will be implemented on the field in future. In other words, the experimental data object of a second type does not represent a counterfactual scenario. Instead, the experimental data object of the second type represents the management events and attributes that has been implemented on the field or will be implemented in future. Accordingly, DSL need not be employed in this scenario for varying the attributes.

[0173] In some embodiments, the simulator 170 can use one or more models to process the experimental data object of the second type to predict one or more respective predicted ecosystem attributes. In some embodiments, the experimental template specifies the one or more models, one or more model versions, and/or one or more parameters sets available to process the experimental data object of the second type (for example, based on one or more methodologies of one or more programs which one or more geographic regions of the experimental data object is or has been enrolled). Optionally, an experimental template may comprise a fully customizable model parameter set and/or one or more pre-determined sets of model parameters. After generating the predicted ecosystem attributes, the software system can notify the user with the predicted ecosystem attributes.

[0174] In some embodiments, the computing device 102 may provide a user interface that enables the display of the output of the simulation. In some implementations, the computing device may quantify the ecosystem attributes for the one or more regions and display the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region. In some implementations, the computing device 102 may assign an impact score based on the predicted ecosystem attribute and provide a recommendation to the user via the user interface in the user device. In some embodiments, the user interface may include user interface elements allowing modifications to one or more parameters in an experimental template as described here.

[0175] In some embodiments, the simulations estimate for an experimental data object of the second type the potential effect of changing: one or more models, one or more methodologies, and/or one or more programs. Simulations estimating the effect of application of one or more models, one or more methodologies, and/or one or more programs, over one or more years, can generate a recommendation of a schedule of program participations to implement for optimal ecosystem benefit.

[0176] In some embodiments, an impact score is assigned to each model applied to data objects of the first type. For example, to assign impact scores to each of the plurality of models or set of models associated with a schedule of program participations, each respective predicted is compared to the respective ecosystem attribute of a reference model or reference set of models (for example, a set of models associated with a schedule of program participations). For example, if a first model or first set of models produces a more optimal ecosystem attribute profile than a second model or second set of models, the first model or first set of models is assigned a higher impact score than the second model or second set of models.

EXEMPLARY DATA OBJECT

[0177] FIG. 3 provides a layout of an example data object 300, in accordance with one or more embodiments. Data object 300 includes soil attributes and one or more management events conducted on a field. As seen in FIG. 3, data object 225 includes soil attributes collected on 2021-09-01 followed by the latitude and longitude of the field from which the soil sample was analyzed. The soil attributes further include bulk-density, coarse-fragments sand fraction, clay fraction, organic fraction and pH level. The data object 300 further includes five agronomic events that occurred on the field. For example, the first farming event is planting of corn that was performed on 2020-04-29. The second farming event is harvesting that was performed on 2020-08-22. The third farming event is planting of barley that was performed on 2020-09-01. The fourth farming event is termination that was conducted on 2020-10-25 and finally the fifth farming event is tillage that was performed on 2020-11-03. Note that each farming event is described using one or more attributes. In some embodiments, the data objects are stored in a database 150. The database 150 can be located within the software system or it can be executed by a remote server that is connected to the software system via a network.

EXEMPLARY DOMAIN- SPECIFIC LANGUAGE (DSL) INSTRUCTIONS

[0178] FIG. 4 A provides exemplary DSL instructions that can be included in the experimental data object 155 to add and remove new management events during simulation, in accordance with one or more embodiments. In some embodiments, the DSL provides methods for adding and removing management events while performing simulations. For example, assume that a candidate data object 160 includes multiple management events along with one or more attributes describing each of the management events. Further assume, that the user wants to run simulations to predict one or more ecosystem attributes using an alternate (or counter factual) scenario of management events. For example, the user wants to add two more management events of planting bentgrass and harvesting in addition to the existing management events of the candidate data object 155. Also assume that the user wants to remove three management events of tillage, planting and termination. To add and remove management events, the user can provide the DSL instructions. [0179] As seen in FIG. 4A, the first set of instructions add the management events planting and harvest. The second set of instructions remove the management events tillage, planting, and termination. Note that each farming event is attributed by a date since the date and the type of farming event uniquely identifies a farming event in a candidate data object 160.

[0180] In some embodiments, the DSL provides methods for altering the one or more attributes of the management events of a candidate data object 160 to generate an alternate scenario during simulation. For example, the experiment creator 142 can generate DSL instructions to specify discrete values for one or more attributes to be included in the experimental data object 155. Assume that the user wants to predict the ecosystem attribute of the nitrous oxide production when two different quantities of chemicals are applied to a candidate location. The experiment creator 142 can generate an experimental data object 155 of the first type by reading DSL instructions that specify the two values of the quantities of chemicals. The simulator 170 can use the experimental data object 155 and the candidate data object 160 to run two simulations with the two different values of chemical quantities. [0181] As another example, assume that the user wants to predict the ecosystem attribute of the nitrous oxide production if different agrichemicals are applied in different quantities on a candidate location. The experiment creator 142 can generate an experimental data object 155 of the first type by reading DSL instructions that specify the two quantities of chemicals and the identifiers of the two chemicals. The simulator 170 can use the experimental data object 155 and the candidate data object 160 to run four simulations with different permutations of the two quantities and types of chemicals.

[0182] In some embodiments, the DSL also provides methods for specifying a continuous range of values. Assume that the user wanted to predict the ecosystem attribute of the nitrous oxide production when different quantities of agrichemicals are applied to the field. The user can provide a range of values indicating different quantities of chemicals via the UI. The instructions can be passed on to the experiment creator 142 via the experiment scheduler 135. The experiment creator 142 can generate an experimental data object 155 of the first type based on DSL instructions that alter the attribute indicating the rate of chemical application of the candidate data object 160 to different values in the given range while running simulations. [0183] FIG. 4B provides an example of DSL instructions of the experimental data object 155 to alter the attributes of the management events of the candidate data object 160 while running simulations, in accordance with one or more embodiments. As seen in FIG. 4B, the first DSL instruction set provides three discrete values of 1 kg, 5 kg, and 10 kg for the attribute “harvest.yield kg” of the farming event “harvest”. The experiment creator 142 can use the first DSL instruction set to generate experimental data objects 155 where the attribute “harvest.yield kg” is altered to 1 kg, 5 kg, and 10 kg to generate one or more predicted ecosystem attributes. Similarly, the second DSL instruction set provides a range of values for the attribute “harvest.yield kg” of the farming event “harvest”. That range starts from 0 to 10 with a step size of 3. The experiment creator 142 can use the second DSL instruction set to generate experimental data objects 155 where the attribute “harvest.yield kg” is altered to 0 kg, 3 kg, 6 kg, and 9 kg to generate one or more predicted ecosystem attributes. Note that each farming event is uniquely identified by the date and type in a candidate data object 160. For example, the farming event “harvest” is identified by the “date” that indicates the date when harvesting was implemented on the candidate location.

[0184] In addition to the above examples of altering the attributes of the management events, the geospatial boundary of a given input may also be varied. In such embodiments, the attributes of the management event include a geospatial boundary. For example, the boundary may correspond to a sub-field, which is expanded to a full field boundary; a given management event may be expanded to more than one field; or a boundary may be modified to account for different boundary mapping techniques.

PERMUTATIONS OF MANAGEMENT EVENTS

[0185] FIG. 5 provides a visual representation illustrating the permutations of the management events and the one or more attributes describing the management events, in accordance with one or more embodiments. As seen in FIG. 5, three management events 502, 504 and 506 represent farming practices, tillage, planting and harvest performed on a field in the year 2018. A baseline scenario for 2019 may be determined by creating an experimental data object of the second type that includes the same management events 502, 504 and 506. Note that in 2018 tillage method of “moldboard” was used. Since there is no change in management events and attributes in 2019, the same tillage 502 and harvest 506 method will be used for the baseline simulation in 2019.

[0186] In some embodiments, the computing device 102 can be used to determine the best combination of management events and attributes for a candidate location. For example, the user can instruct the experiment scheduler 135 to determine the best permutation of management events and attributes for the field. In such embodiments, the experiment scheduler 135 can specify a plurality of experiments based on different permutations of management events. The experiment scheduler 135 can instruct the experiment creator 142 to create a plurality of experimental data objects 155 of the first type for the candidate location where each experimental data object 155 is based on DSL instructions that specify altered management events and attributes. The experiment creator 142 can further specify a range of values for the attributes describing the management events of each of the plurality of experimental data objects 155.

[0187] Three management events 508, 510 and 512 represent farming practices, tillage, planting and harvest that are performed on a field in the year 2018. However, the user may simulate different permutations of the one or more attributes of the management events to select the optimal attributes for the management events tillage, planting and harvest that could be implemented in the year 2019. The computing device 102 creates an experimental data object 155 of the first type based on DSL instructions to simulate a range of values for the one or more attributes. The simulator 170 can use experimental data object 155 of the first type to perform multiple simulations to select the most optimum attributes. For example, the 1st permutation of the one or more attributes includes tillage 308-1 that uses a different tillage method “Chisel”. Similarly, in the 2nd permutation the method of tillage 308-2 is changed to “Disk”. Similarly, in the nth permutation the method of tillage 308-3 is changed to “Finisher” and the method of harvest 312-1 is changed to “Digging.” The simulator 170 uses the one or more models to simulate each of the permutations to predict one or more ecosystem attributes as directed by the user. The software system selects a permutation of the one or more attributes of the management events that have the impact scores.

[0188] It will be appreciated that a given management event (e.g., tillage, fertilization) may have associated attributes (e.g., none, 0 kg) indicating that it does not occur in a given scenario. Accordingly, agricultural events may be added or removed entirely. For example, removing all tillage events may be used to determine the effects of a no-till scenario. In another example, a cover crop may be analyzed by adding the corresponding events. In various embodiments, events are added/removed or parameters are changed within a single experimental file.

[0189] Various non-limiting examples of relevant management events and associated attributes are provided herein. Exemplary management events and associated attributes include: planting a crop and a particular variety of seed planted (e.g., a non-GMO seed); planting a cover-crop and one or more cover crop species planted; tillage and the particular tillage technique (including not tilling); irrigating and irrigation type; water conservation, and a specific technique; pesticide application and pesticide type/amount; insecticide application and insecticide type/amount; application of a product and type of input applied (for example, a fertilizer, manure, one or more microbe, a material for direct air capture of a greenhouse gas, a silicate material, crushed silicate rock such as basalt, or a material for passive direct air capture of a greenhouse gas); harvesting and a harvesting technique; and a field residue burning event and a type and or amount of field residue.

[0190] As noted above, in various embodiments, location attributes are varied. In such embodiments, alternate locations may be specified, for example using alternate latitude and longitude coordinates, incremental adjustments to coordinates, or other transformations to be applied to a geospatial boundary.

PROCESS OF PREDICTING ECOSYSTEM ATTRIBUTE

[0191] FIG. 6 is a flow diagram that illustrates an example process 600 of generating a predicted ecosystem attribute, in accordance with one or more embodiments. Operations of the process 600 can be implemented, for example, by the components of the ecosystem management environment 100 includes the computing device 102, the data store 104, the sensing device 106, the user device 108, and the network 109.

[0192] The computing device 102 may access 610 agronomic data from a data source. The agronomic data may include historical information of one or more management events. Each management event is associated with a time, a geospatial boundary, and one or more attributes of the management event for an agronomic region. In some embodiments, each management event may include a farming practice on a corresponding agronomic region, and the farming practice may be planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, or organic amendment. In some embodiments, the one or more attributes of the management event may correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

[0193] The computing device 102 generates 620 based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions. The data object may indicate the respective time, geospatial boundary, and one or more attributes of the management event for the respective agronomic region. [0194] The computing device 102 receives 630 a request to perform a target experimental farming practice on a candidate agronomic region from a user device. The target experimental farming practice may include altering one or more attributes of the management event for the candidate agronomic region.

[0195] The computing device 102 generates 640, a target experimental data object for the candidate agronomic region, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region. In some embodiments, generating the target experimental data may include using a target experimental template generated by a first model trained to generate experimental templates for farming practices. In some embodiments, the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the candidate agronomic region.

[0196] The computing device 102 identifies 650, one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects. In some embodiments, the computing device 102 may identifiy the representative data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects.

[0197] The computing device 102 provides 660 the generated target experimental data object and the one or more candidate data objects to a second model as input. The second model may be an model that predicts an ecosystem attribute of the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object. In some embodiments, an ecosystem attribute may include information related to water use, nitrogen use, soil carbon sequestration, greenhouse gas emission avoidance, yield, or nitrous oxide production.

[0198] The computing device 102 receives 670, determining an ecosystem attribute of the candidate agronomic region using a second model trained to predict ecosystem attributes of agronomic regions.

[0199] The computing device 102 determines 680 determining an ecosystem attribute of the candidate agronomic region using the second model trained to predict ecosystem attributes of agronomic regions.

[0200] Optionally, the computing device 102 may, based on the predicted ecosystem attributes, assign an impact score to the target framing practice. For example, the computing device 102 assigns 690 an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region. In some embodiments, the baseline ecosystem attribute is predicted using the second model without performing the target experimental farming practice on the candidate agronomic region. In some embodiments, the computing device 102 may generate a recommendation on an agricultural intervention based on the predicted ecosystem attribute of the candidate location corresponding to the target experimental farming practice. In some embodiments, the computing device 102 may generate a recommendation on an agricultural intervention based on the determined impact score to the target experimental data object associated with the target experimental farming practice. [0201] FIG. 7 is a flow diagram that illustrates a method for generating synthetic agronomic data, in accordance with one or more embodiments. In various embodiments, synthetic data in both space and time are generated for use in further simulation, including to analyze and predict changes to soil organic carbon (SOC) stocks. Synthetic agronomic data can be used to train one or more models and to perform simulations to predict ecosystem attributes. The synthetic data is particularly useful in situations where there exists limited training data, in particular paired management practice and soil lab measurements. In some embodiments, data types require significant manual effort to obtain and may have constraints due to grower privacy. Manually-generated or parameter-grid scenarios may be used for experimenting with biogeochemical models, but is often time-consuming and error-prone. Accordingly, more realistic synthetic data streamline and increase the quality of these experiments.

[0202] In some embodiments, the synthetic agronomic data are generated using Gaussian Mixture Models (GMMs). GMMs are a class of unsupervised models that attempt to cluster data in a multidimensional space using a mixture of multivariate Gaussian distributions. This approach to clustering approximates complex joint densities of soil variables using a combination of relatively simple probability densities. Furthermore, GMMs have an advantage over a completely non-parametric approach to modeling joint probability densities such as kernel density estimation since it allows regularization of the final clusters using prior expectations of what the relationships between different soil variables should look like. The GMMs accurately represent the correlations between input variables, based on current experimental data. Such methods can then generate sample data sets based on particular filters (e.g., latitude or clay content). This enables analysis of different regions based on their aggregated characteristics. Faster and more realistic end-to-end testing of various methodologies is achieved using a set of input data that more closely mimics actual soil measurements and management practices. More realistic scenarios allow model exploration (e.g., sensitivity analysis) and comparison. Synthetic data of management practices and soil measurements further allow realistic data to be shared with scientific collaborators while protecting grower privacy. Synthetic data may include representative agricultural data that is based on domain knowledge and public data.

[0203] In various embodiments, a Variational Bayesian Gaussian Mixture Model (VBGMM) is employed. VBGMMs combine two features that allow information from both scientific literature and observed fields to be effectively leveraged in simulating new data. First, VBGMMs allow for the flexible modeling of the joint distribution of a set of continuous variables, i.e., they enable modeling how soil parameters such as texture, bulk density (BD), SOC, and latitude covary with changes in SOC and/or BD due to practice changes. This is valuable, since it enables greater precision in modeling and understanding of how practice changes influence changes in SOC and/or BD depending on the environmental conditions of the particular fields in which they occur. Second, VBGMMs are a tool for Bayesian analysis, meaning that this distribution can be modeled as a posterior distribution proportional to the product of a prior distribution of the parameters and a likelihood function of the data observed given the parameters. This allows prior knowledge about the effects a practice change should have on SOC and/or BD from the scientific literature to be jointly leveraged with the data observed from fields and experiments.

[0204] Furthermore, this approach can be extended to generate synthetic soil data even under practice changes for which there is a well-informed effect size distribution, but no observed data. Once a VBGMM has been fit to a set of empirical data and a posterior joint distribution has been modeled, that joint distribution can be sampled infinitely to generate new soil samples in space or time, with values that are representative of the original distribution.

[0205] As time progresses and more samples are collected, this process can iterate, and the synthetic samples will become more and more reflective of what one can actually expect to observe.

[0206] At 701, a prior distribution of parameters is determined, including both initial soil conditions and their one-year changes due to a practice change condition. At 702, using values from the scientific literature and global averages, a vector of mean values and a diagonal matrix of covariances that maps onto the means and variances of each of the parameters is generated. A prior can be generated automatically for the joint distribution using the mean values and covariances of the data; however, this approach does not take advantage of prior knowledge collected from the scientific literature.

[0207] At 703, a dataset is constructed to serve as the likelihood using soil sample data (e.g., data objects as discussed above with regard to Fig. 1. In various embodiments, missing values are imputed and all SOC, BD, and pH measurement values are harmonized to the same depth or filtered to a specific depth range.

[0208] At 704, a VBGMM is fit to the static soil data collected in a baseline soil sampling year after scaling. At 705, each soil sample is assigned in the baseline sampling year to the cluster in which it has the highest probability of occurring according to the model. At 706, clusters are assigned to each of the samples in the next year’s dataset using the VBGMM fit on the baseline static soil data. At 707, within each cluster, pairwise subtraction of static soil data is performed between the second and baseline years’ values to yield delta values. At 708, a new VBGMM is fit on the static soil data from the baseline year as well as the calculated delta values after scaling, using the values from step 702 as the parameters of the prior distribution.

[0209] At 708, samples are taken from the new VBGMM and reverse scale to yield new emissions and baseline soil measurements in space.

[0210] In an exemplary embodiment of the VBGMM described above, a small subset of data with minimal quality checks, data preprocessing, or model parameter tuning was selected. Four fields were included with soil samples from two years. Of those four fields, three reported a practice change of “Tillage change/reduction” (the most representative practice change). The data from those three fields (n = 593) were run through the algorithm above to generate a completely synthetic dataset (n = 100,000).

[0211] The VBGMM for this practice change was built using 11 different soil variables: Sample latitude, Sample longitude, Baseline sample SOC (%), Baseline sample bulk density (g cm-3), Baseline sample pH, Clay content (%), Silt content (%), One-year change in sample SOC, One-year change in sample BD, One-year change in sample pH, Sample top depth (cm), Sample bottom depth (cm), etc.

ADDITIONAL CONSIDERATIONS [0212] FIGs. 8-11 are plots comparing select univariate distributions between the empirical dataset and the synthetic dataset in the above-described example, in accordance with one or more embodiments.

[0213] FIGs. 12-13 are plots comparing select bivariate distributions between the empirical dataset and the synthetic dataset in the above-described example, in accordance with one or more embodiments.

[0214] In various embodiments, various approaches for generating synthetic sequestration/emissions data are employed. These include the VBGMM framework described above, Bayesian hierarchical linear models, and combinations thereof.

[0215] In an exemplary embodiment, if no change-over-time data is available for a particular practice change, static soil sample data for a subset of fields is selected (e.g., where the practice change of interest has been indicated for future evaluation, or by a geography of interest, or indiscriminately). A random variable is created for the effect of the selected practice change on SOC and BD. Optionally, if given a mean or median effect size has a 95% confidence interval, it may be modeled as a normal or skew-normal random variable. For each observation in the static soil sample dataset, the SOC or BD value is multiplied by a sample from the effect size random variable and stored in a new column as the delta value simulating a one-year-change in the respective variable for that location.

[0216] In an exemplary embodiment, if change-over-time data are available for a particular practice change, the data are clustered according to the VBGMM framework described above. For each soil sample, the marginal posterior distribution of the delta value is sampled for the variable of interest within the cluster to which the soil sample belongs. This sampled value represents a simulated change in SOC or BD due to the selected practice change for another year, assuming that the effect size is constant over time (not a strong assumption).

[0217] In additional embodiments, synthetic SOC sequestration/emissions data are generated in time using hierarchical Bayesian linear regression and Markov Chain-Monte Carlo simulation. Hierarchical Bayesian linear regression is used to determine a posterior effect size estimate for change in SOC due to a practice change, where the posterior effect size slope and intercept vary by soil sample cluster.

[0218] In an exemplary embodiment, if no change-over-time data is available for a particular practice change, static soil sample data are selected for a subset of available fields (e.g., where the practice change of interest has been indicated for future evaluation, or by a geography of interest, or indiscriminately). A random variable is created for the effect of the selected practice change on SOC and BD. Optionally, if given a mean or median effect size with a 95% confidence interval, it may be modeled as a normal or skew-normal random variable. For each observation in the static soil sample dataset, the SOC or BD value is multiplied by a sample from the effect size random variable and stored as the delta value simulating a one-year-change in the respective variable for that location.

[0219] In an exemplary embodiment, if change-over-time data are available for a particular practice change, the data are clustered according to the algorithm discussed above with regard to the VBGMM framework. An unpooled Bayesian linear regression model is built. A hyperprior may be selected to reflect a-priori knowledge of the effect size from the scientific literature. Fitting an individual slope to each cluster will introduce a lot of variance into the models/posterior predictive distributions, which creates certain tradeoffs. A benefit to this approach is that it allows simulation of changes to SOC and BD heterogeneously in ways that are represented in the data, without having to explicitly model the joint distribution of all of the variables used to assign samples to clusters in the first place. One of the negative consequences of this is that the posterior distributions may not be stable. The posterior distribution is sampled for each sample in the original dataset. This sampled value of represents a simulated change in SOC or BD for that sample due to the practice change.

[0220] FIG. 14 illustrates an exemplary branching during the method of generating a management zone, in accordance with one or more embodiments. In addition to the use cases noted above, it will be appreciated that the present disclosure is suitable for use in a variety of additional modeling scenarios.

[0221] In various embodiments, sensitivity analyses are conducted. In an exemplary sensitivity analysis, inputs to a biogeochemical (or other type) model are perturbed while holding all other inputs constant. This allows one to understand how beneficial practices or practice changes should be defined. Additionally, this allows one to understand what data the models are particularly sensitive too, helping to define program data collection requirements. [0222] In various embodiments, the frameworks described herein are used to: 1) empirically select a conservative emissions value when management data are missing; and 2) determine the uncertainty around emissions estimates that results from the missing data. [0223] In various embodiments, when management data are missing, they are gap-filled with values that are known conservative estimates based on domain expertise. For example, where information on tillage method is missing, intensive tillage may be used as a conservative assumption, because it will lead to the under- statement of the carbon credits produced.

[0224] However, there may be scenarios in which there is not strong domain knowledge to inform expectations about which practices will be conservative. In addition, in a multimodel approach, which practices are conservative may differ by model.

[0225] The frameworks provided herein can be used to determine what constitutes a conservative emissions value by assessing the range of emissions values that are possible depending on what values are assumed in place of missing data. For categorical attributes, all possible management values can be exhaustively tried. For continuous variables, or simply for performance reasons, the space of possible management values may be sampled. With these different management values, these frameworks can be used to determine what the corresponding model-estimated emissions would be. By running these scenarios through the models, a distribution of possible emissions values can be determined that are consistent with the available and missing data. A conservative emissions estimate can then be chosen by slicing the distribution at a given percentile, e.g., the 90th percentile. The resulting distribution can also be used to empirically quantify the uncertainty that results from missing data, by looking at the spread of the resulting emissions distribution; e.g. by calculating the variance of the resulting emissions values.

[0226] FIG. 15 is a schematic of an example of a classical computing node 400, in accordance with one or more embodiments. Computing node 400 is only one example of a suitable computing node and is not intended to suggest any limitation as to the scope of use or functionality of embodiments described herein. Regardless, computing node 10 is capable of being implemented and/or performing any of the functionality set forth hereinabove.

[0227] In computing node 400 there is a computer system/server 12, which is operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well-known computing systems, environments, and/or configurations that may be suitable for use with computer system/server 12 include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputer systems, mainframe computer systems, and distributed cloud computing environments that include any of the above systems or devices, and the like. [0228] Computer system/server 12 may be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, and so on that perform particular tasks or implement particular abstract data types. Computer system/server 12 may be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computer system storage media including memory storage devices.

[0229] As shown in FIG. 15, computer system/server 12 in computing node 400 is shown in the form of a general-purpose computing device. The components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including system memory 28 to processor 16.

[0230] Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronics Standards Association (VESA) local bus, Peripheral Component Interconnect (PCI) bus, Peripheral Component Interconnect Express (PCIe), and Advanced Microcontroller Bus Architecture (AMBA).

[0231] Computer system/server 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by computer system/server 12, and it includes both volatile and non-volatile media, removable and nonremovable media.

[0232] System memory 28 can include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and/or cache memory 32. Computer system/server 12 may further include other removable/non-removable, volatile/non-volatile computer system storage media. By way of example only, storage system 34 can be provided for reading from and writing to a non-removable, non-volatile magnetic media (not shown and typically called a "hard drive"). Although not shown, a magnetic disk drive for reading from and writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable, non-volatile optical disk such as a CD-ROM, DVD-ROM or other optical media can be provided. In such instances, each can be connected to bus 18 by one or more data media interfaces. As will be further depicted and described below, memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the disclosure.

[0233] Program/utility 40, having a set (at least one) of program modules 42, may be stored in memory 28 by way of example, and not limitation, as well as an operating system, one or more application programs, other program modules, and program data. Each of the operating system, one or more application programs, other program modules, and program data or some combination thereof, may include an implementation of a networking environment. Program modules 42 generally carry out the functions and/or methodologies of embodiments as described herein.

[0234] Computer system/server 12 may also communicate with one or more external devices 14 such as a keyboard, a pointing device, a display 24, etc.; one or more devices that enable a user to interact with computer system/server 12; and/or any devices (e.g., network card, modem, etc.) that enable computer system/server 12 to communicate with one or more other computing devices. Such communication can occur via Input/Output (VO) interfaces 22. Still yet, computer system/server 12 can communicate with one or more networks such as a local area network (LAN), a general wide area network (WAN), and/or a public network (e.g., the Internet) via network adapter 20. As depicted, network adapter 20 communicates with the other components of computer system/server 12 via bus 18. It should be understood that although not shown, other hardware and/or software components could be used in conjunction with computer system/server 12. Examples, include, but are not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0235] The present disclosure may be embodied as a system, a method, and/or a computer program product. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present disclosure.

[0236] The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non- exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.

[0237] Computer readable program instructions described herein can be downloaded to respective computing/processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and/or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing/processing device.

[0238] Computer readable program instructions for carrying out operations of the present disclosure may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++ or the like, and conventional procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions may execute entirely on the user’s computer, partly on the user’s computer, as a stand-alone software package, partly on the user’s computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user’s computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present disclosure.

[0239] Aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.

[0240] These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function/act specified in the flowchart and/or block diagram block or blocks.

[0241] The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions/acts specified in the flowchart and/or block diagram block or blocks.

[0242] The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.

[0243] The descriptions of the various embodiments of the present disclosure have been presented for purposes of illustration but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

CLAIMS What is claimed is:

1. A method comprising: generating, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generating, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identifying one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; and determining an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object.

2. The method of claim 1, further comprising: assigning an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

3. The method of claim 2, wherein the baseline ecosystem attribute is predicted using the second model without performing the target experimental farming practice on the candidate agronomic region.

4. The method of claim 1, wherein each management event comprises a farming practice on a corresponding agronomic region, and the farming practice comprises planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, or organic amendment.

5. The method of claim 4, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

6. The method of claim 1, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production

7. The method of claim 1, wherein the target experimental farming practice comprises altering one or more attributes of the management event for the candidate agronomic region.

8. The method of claim 7, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region.

9. The method of claim 1, wherein the target experimental template specifies one or more of: the one or more second models to be applied, a model version for each second model, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use.

10. The method of claim 1, wherein the second models comprise at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator.

11. The method of claim 1, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory-based greenhouse gas emissions calculator, a statistical model.

12. The method of claim 1, further comprising: generating an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts.

13. The method of claim 1, further comprising: enabling a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

14. The method of claim 1, further comprising: quantifying ecosystem attributes for one or more agronomic regions; and displaying the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

15. The method of claim 1, further comprising: assigning an impact score based on the determined ecosystem attribute of the candidate agronomic region; and providing a recommendation based on the impact score to a user via a user interface.

16. The method of claim 1, further comprising: receiving a modification to one or more parameters in the target experimental template from a user via a user interface.

17. A non-transitory computer readable storage medium comprising stored program code, the program code comprising instructions, the instructions when executed cause a processor system to: generate, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generate, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identify one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; and determine an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object.

18. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: assign an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

19. The non-transitory computer readable storage medium of claim 18, wherein the baseline ecosystem attribute is predicted using the second model without performing the target experimental farming practice on the candidate agronomic region.

20. The non-transitory computer readable storage medium of claim 17, wherein each management event comprises a farming practice on a corresponding agronomic region, and the farming practice comprises planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, or organic amendment.

21. The non-transitory computer readable storage medium of claim 20, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

22. The non-transitory computer readable storage medium of claim 17, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production.

23. The non-transitory computer readable storage medium of claim 17, wherein the target experimental farming practice comprises altering one or more attributes of the management event for the candidate agronomic region.

24. The non-transitory computer readable storage medium of claim 23, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region.

25. The non-transitory computer readable storage medium of claim 17, wherein the target experimental template specifies one or more of: the one or more second models to be applied, a model version for each second model, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use.

26. The non-transitory computer readable storage medium of claim 17, wherein the second models comprise at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator.

27. The non-transitory computer readable storage medium of claim 17, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory-based greenhouse gas emissions calculator, a statistical model.

28. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: generate an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts.

29. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: enable a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

30. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: quantify ecosystem attributes for one or more agronomic regions; and display the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

31. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: assign an impact score based on the determined ecosystem attribute of the candidate agronomic region; and provide a recommendation based on the impact score to a user via a user interface.

32. The non-transitory computer readable storage medium of claim 17, wherein the instructions when executed further cause a processor system to: receive a modification to one or more parameters in the target experimental template from a user via a user interface.

33. A system comprising: one or more computer processors; and one or more computer-readable mediums storing instructions that, when executed by the one or more computer processors, cause the system to: generate, based on accessed agronomic data, a representative data object for each of one or more management events performed on an agronomic region of a plurality of agronomic regions, each data object indicating a time, a geospatial boundary, and attributes of the management event; generate, in response to receiving a request to perform a target experimental farming practice on a candidate agronomic region, a target experimental data object for the candidate agronomic region using a target experimental template generated by a first model trained to generate experimental templates for farming practices; identify one or more representative data objects corresponding to the target experimental farming practice for the candidate agronomic region as one or more candidate data objects based on the time, geospatial boundary, and attributes of the management event of the one or more representative data objects; and determine an ecosystem attribute of the candidate agronomic region using one or more second models trained to predict ecosystem attributes of agronomic regions, the one or more second models determining the ecosystem attribute for the candidate agronomic region by simulating the target experimental farming practice in the candidate agronomic region using the one or more candidate data objects and the target experimental data object.

34. The system of claim 33, wherein the instructions when executed further cause a processor system to: assign an impact score to the target experimental farming practice based on a difference between the determined ecosystem attribute and a baseline ecosystem attribute for the candidate agronomic region.

35. The system of claim 34, wherein the baseline ecosystem attribute is predicted using the second model without performing the target experimental farming practice on the candidate agronomic region.

36. The system of claim 34, wherein each management event comprises a farming practice on a corresponding agronomic region, and the farming practice comprises planting, water conservation, irrigation, pesticide application, insecticide application, grazing, harvesting, termination, tillage, input application, residue cover, burning, or organic amendment.

37. The system of claim 36, wherein the one or more attributes of the management event correspond to one or more parameters of the corresponding farming practice, and each attribute includes one or more values describing the corresponding parameters of the farming practice.

38. The system of claim 34, wherein the ecosystem attribute includes water use, biodiversity, nitrogen or other chemical input use or run-off, soil carbon sequestration, greenhouse gas emissions, greenhouse gas emission avoidance, yield, or nitrous oxide production.

39. The system of claim 34, wherein the target experimental farming practice comprises altering one or more attributes of the management event for the candidate agronomic region.

40. The system of claim 39, wherein the experimental template specifies the target experimental farming practice to be performed on the candidate agronomic region, and the one or more attributes of the management event to be altered for the agronomic region.

41. The system of claim 34, wherein the target experimental template specifies one or more of: the one or more second models to be applied, a model version for each second model, one or more sets of model parameters, a life cycle inventory database to use, and one or more default equations to use.

42. The system of claim 34, wherein the second models comprise at least one biogeochemical model and one or more inventory-based greenhouse gas emissions calculator.

43. The system of claim 34, wherein the second models include one or more of a machine learning model, a process based biogeochemical model, an inventory-based greenhouse gas emissions calculator, a statistical model.

44. The system of claim 34, wherein the instructions when executed further cause a processor system to: generate an immutable association between an experimental template and one or more of experimental data objects, candidate data objects, predicted ecosystem attributes, and ecosystem impacts.

45. The system of claim 34, wherein the instructions when executed further cause a processor system to: enable a display of the determined ecosystem attribute of the candidate agronomic region via a user interface from a user device.

46. The system of claim 34, wherein the instructions when executed further cause a processor system to: quantify ecosystem attributes for one or more agronomic regions; and display the quantified ecosystem attributes with a map showing each quantified ecosystem attribute with the respective agronomic region.

47. The system of claim 34, wherein the instructions when executed further cause a processor system to: assign an impact score based on the determined ecosystem attribute of the candidate agronomic region; and provide a recommendation based on the impact score to a user via a user interface.

48. The system of claim 34, wherein the instructions when executed further cause a processor system to: receive a modification to one or more parameters in the target experimental template from a user via a user interface.