NL2036869A

NL2036869A - Drought risk assessment system

Info

Publication number: NL2036869A
Application number: NL2036869A
Authority: NL
Inventors: Dong Zhaoyu; Weng Baisha
Original assignee: China Inst Water Resources & Hydropower Res
Priority date: 2023-04-24
Filing date: 2024-01-23
Publication date: 2024-03-29
Also published as: CN116663885A; NL2036869B1

Abstract

Disclosed is a drought risk assessment system. The system includes a hazard module, an exposure module, a vulnerability module, a toughness module, a comprehensive drought risk module and a drought risk early warning information distribution module, where the hazard module corresponds to a f1rst-grade hazard early warning; the vulnerability module corresponds to a second-grade hazard early warning; the comprehensive drought risk module corresponds to a third-grade hazard early warning; and the drought risk early warning information distribution module is configured to distribute drought early warning information of an area to a client. The drought risk assessment system in the present invention avoids over-assessment and under-assessment of a real water scarcity situation by a single drought indeX, can accurately understand a disaster-causing reason of a drought risk in a user area, is conducive to promotion of implementation of a water-saving policy in an industrial water-scarce area.

Description

DROUGHT RISK ASSESSMENT SYSTEM

TECHNICAL FIELD

[0001] The present invention belongs to the field of hydrological events and disaster risk assessment, and particularly relates to a drought risk assessment system.

BACKGROUND

[0002] Under the background of climate change, it has become a key topic to research and cope with drought disasters for sustainable development of China's economy and society.

Drought events can be mainly divided into meteorological drought, hydrological drought, agricultural drought and ecological drought under conditions of most carriers. However, a single drought assessment cannot fully reflect real severity of water scarcity, and it is still required to explore a transitive relation of drought. Long-term drought is likely to lead to land degradation and desertification to aggravate deterioration of ecological environment, which poses an enormous threat to economic and social development.

[0003] In order to systematically assess cascade effects of drought risks under conditions of different carriers, it is considered to improve a drought risk assessment system. Thus, a drought risk assessment method considering toughness of carriers is provided.

SUMMARY

[0004] In order to solve the above problems, an example of the present invention provides a drought risk assessment system, which is a progressive automatic early warning service system implementing a drought hazard early warning through Beidou and a 5G internet of things technology and serving industrial and agricultural production.

[0005] The drought risk assessment system in an example of the present invention includes: a hazard module, which assesses a hazard of a drought at a basin by means of daily precipitation data and soil moisture data, where the hazard module corresponds to a first-grade hazard early warning; an exposure module, which reflects fractional vegetation cover by means of a normalized difference vegetation index (NDVI) and an enhanced vegetation index (EVI) and computes exposure of a drought in the basin; a vulnerability module, which computes water demand of a carrier by means of a root water absorption model and then computes vulnerability of water supply and demand of the carrier, where the vulnerability module corresponds to a second-grade hazard early warning; a toughness module, which computes toughness of carrier vegetation under drought stress, that is, toughness of a carrier in the basin, by means of an autoregressive with exogenous input

(ARx) model; a comprehensive drought risk module, which decomposes and estimates contribution degrees of the hazard, the exposure, the vulnerability and the toughness by means of random forest and interpretable machine learning models, and computes a comprehensive drought risk, where the comprehensive drought risk module corresponds to a third-grade hazard early warning; and a drought risk early warning information distribution module, where the drought risk early warning information distribution module is configured to distribute drought early warning information of the area to a client.

[0006] The drought risk assessment system in an example of the present invention avoids over-assessment and under-assessment of a real water scarcity situation by a single drought index, can accurately understand a disaster-causing reason of a drought risk in a user area, is conducive to promotion of implementation of a water-saving policy in an industrial water-scarce area, and provides strong scientific and technological support for industrial and agricultural production.

[0007] In some examples, the hazard module refers to a disaster-causing factor hazard assessment module. Long-time daily relative soil wetness and number of days without precipitation are obtained first. A most unfavorable situation is obtained by combining double correction of meteorological drought and agricultural drought. A real water scarcity feature of carrier crops is accurately assessed. A first-grade hazard early warning of drought grades is issued.

[0008] In some examples, the exposure module extracts NDVI and EVI data from a remote sensing image, converts the NDVI and EVI data into fractional vegetation cover data, carries out weighted fusion on the fractional vegetation cover data, and obtains carrier exposure data.

[0009] In some examples, the vulnerability module carries out computation through a balance of water supply and demand for growth and development of the carrier. Water demand for growth of the carrier and available water supply of a plurality of layers of soil are quantitatively computed by means of the root water absorption model. In a case that the water demand is greater than the water supply, water scarcity is computed and a second- grade hazard early warning of vulnerability of the carrier is carried out according to a rolling frequency of historical-present-future data.

[0010] In some examples, computation of toughness of carrier vegetation under drought stress by means of an ARx model is to assess a response of the vegetation to a short-term climate anomaly by means of the ARx model and quantify toughness of short-term carries around the world.

[0011] In some examples, a computation formula of the comprehensive drought risk is as follows:

[0012] R = Helix E2 x V& x (1-T)*

[0013] Ris a comprehensive drought risk, H is hazard, E is exposure, V is vulnerability,

Tis toughness, al is a hazard weight coefficient, a2 is an exposure weight coefficient, a3 is a vulnerability weight coefficient and a4 1s a toughness weight coefficient.

[0014] In some examples, the third-grade hazard early warning includes a third-grade red early warning, a third-grade orange early warning, a third-grade yellow early warning and a third-grade blue early warning.

[0015] In some examples, the drought risk early warning information distribution module uses a progressive early warning mode in which early warning information is distributed on the basis of an internet protocol (IP) address. Various scenarios of combinations of soil wetness and number of days without precipitation in the future is preset first. Whether an area of the IP address will face drought risk stress is determined. Accurate positioning is carried out by means of a satellite, a user is matched, and drought early warning information is distributed.

[0016] In some examples, the drought early warning information includes a first-grade hazard early warning, a second-grade hazard early warning and a third-grade hazard early warning.

[0017] A computer device in an example of the present invention includes a processor and a storage medium. The storage medium stores a computer instruction. The computer instruction implements the drought risk assessment system described in any one of examples of the present invention when run by the processor.

[0018] The computer device in an example of the present invention avoids over- assessment and under-assessment of a real water scarcity situation by a single drought index, can accurately understand a disaster-causing reason of a drought risk in a user area, is conducive to promotion of implementation of a water-saving policy in an industrial water-scarce area, and provides strong scientific and technological support for industrial and agricultural production.

BRIEF DESCRIPTION OF DRAWINGS

[0019] FIG. 1 is a schematic diagram of index construction of a hazard module according to an example of the present invention.

[0020] FIG. 2 is a schematic diagram of construction of a comprehensive drought risk module according to an example of the present invention.

[0021] FIG. 3 is a flow diagram of an example of a comprehensive drought risk module according to an example of the present invention.

[0022] FIG. 4 1s a schematic structural diagram of a drought risk early warning information distribution module according to an example of the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] The examples of the present invention will be described in detail below, instances of the examples are shown in the accompanying drawings. The following examples described with reference to the accompanying drawings are exemplary and merely used to explain the present invention, but cannot be construed as limitations on the present invention.

[0024] As shown in FIGs. 1-4, a drought risk assessment system in an example of the present invention includes a hazard module, an exposure module, a vulnerability module, a toughness module, a comprehensive drought risk module and a drought risk early warning information distribution module.

[0025] A drought risk of a basin 1s assessed by means of daily precipitation data and soil moisture data. The hazard module corresponds to a first-grade hazard early warning.

[0026] Fractional vegetation cover is reflected by means of a normalized difference vegetation index (NDVI) and an enhanced vegetation index (EVI) and exposure of a drought in the basin is computed.

[0027] Water demand of a carrier 1s computed by means of a root water absorption model and then vulnerability of water supply and demand of the carrier is computed. The vulnerability module corresponds to a second-grade hazard early warning. Vulnerability of water demand of evapotranspiration and available water supply of each layer of soil wetness may be computed by means of the root water absorption model.

[0028] Toughness of carrier vegetation under drought stress, that is, toughness of a carrier in the basin, is computed by means of an autoregressive with exogenous input (ARx) model.

[0029] Contribution degrees of hazard, exposure, vulnerability and toughness are decomposed and estimated by means of random forest and interpretable machine learning models, and a comprehensive drought risk is computed. The comprehensive drought risk module corresponds to a third-grade hazard early warning.

[0030] The drought risk early warning information distribution module is configured to distribute drought early warning information of the area to a client.

[0031] The hazard module refers to a disaster-causing factor hazard assessment module.

Long-time precipitation data and soil moisture data are obtained under conditions of 5 different carriers. Large-range long-time daily relative soil wetness and number of days without precipitation are computed by using a python programming language. A most unfavorable situation is obtained by combining double correction of meteorological drought and agricultural drought. A real water scarcity feature of carrier crops is accurately assessed. A first-grade hazard early warning of drought grades is issued.

[0032] High-precision daily precipitation grid data and multi-layer soil moisture data of the basin are obtained, and all grid data values and corresponding latitudes and longitudes are exported as data frame files by using a geospatial data abstraction library (GDAL) program package of python. Hazard assessment and a first-grade early warning are carried out on a drought degree of the basin by combining the number of days without precipitation and relative soil wetness.

[0033] Example: As shown in FIG. 1, relative soil wetness (W) and a number of days without precipitation (China natural runoff dataset (CnRD)) in a basin are computed. With precipitation of 3 mm as a demarcation point, a number of days with precipitation of 3 mm or above is regarded as a number of days with precipitation, and a number of days with precipitation of 3 mm or blow is regarded as a number of days without precipitation. A distribution diagram of daily numbers of days without precipitation of all grid points is computed. Through grain composition and volume density of gridded soil, a field water- holding capacity of a gridded basin is computed according to formula (2), and then daily relative soil wetness of the gridded basin is computed according to formula (1). few U0SES HUNT X clay--0.00003 X silt + 0.0088 som OAT XEEN

[0034] ie

[0035] In the formulas, W is relative soil wetness (%), 8 is a water content (%) of soil, fc is a field water-holding capacity of soil (%), clay 1s a clay content, silt is a silt content, som is an organic matter content of soil, and bd is a volume weight of soil.

[0036] Drought grade assessment takes into account climate characteristics and development degrees of carriers in different seasons and regions. By combing a number of days without precipitation and relative soil wetness, classification criteria for identifying light drought, moderate drought and heavy drought are obtained. After daily drought changes in the basin are identified, a drought frequency is computed to reflect occurrence hazard of a drought event in the basin. According to different drought degrees and frequency distribution obtained through assessment, the drought risk early warning information distribution module is configured to distribute first-grade drought early warning information in the area to a client.

[0037] The exposure module extracts NDVI and EVI data from a remote sensing image by combining an ArcGIS software platform, converts the NDVI and EVI data into fractional vegetation cover data by means of an empirical relation model between NDVI and EVI and carrier coverage, carries out weighted fusion on the fractional vegetation cover data, and obtains exposure data of a carrier.

[0038] The vulnerability module computes water demand of carrier vegetation in different soil layers by means of a root water absorption model, converts the water demand into surface density of water demand of a soil carrier in each layer, compares the surface density with surface density of soil wetness, and computes a vulnerability probability of a balance of supply and demand of the carrier.

[0039] Vulnerability of the carrier at the basin is computed through a balance of water supply and demand for growth and development of the carrier. Water demand for growth of the carrier and available water supply of a plurality of layers of soil are quantitatively computed by means of the root water absorption model. In a case that the water supply is greater than the water demand, a value of V is 1. In a case that the water demand is greater than the water supply, water scarcity is computed and a second-grade early warning of vulnerability of the carrier is carried out according to a rolling frequency of historical- present-future data. The drought risk early warning information distribution module is configured to distribute second-grade drought early warning information of the area to a client.

[0040] In a case of the toughness module, vegetation toughness refers to an ability of vegetation to recover functions thereof or adapt to environmental changes after the vegetation faces impacts of drought and other unfavorable factors. A response of the vegetation to a short-term climate anomaly is assessed by means of the ARx model, and recovery and adaptation abilities of short-term carriers under drought stress around the world can be quantified. The ARx model quantifies an ability of the carrier to recover functions thereof or adapt to environmental changes by computing a recovery degree of the carrier after an adverse impact of meteorological drought stress.

[0041] In a case of the comprehensive drought risk module, as shown in FIGs. 2 and 3, a comprehensive drought risk (R) of the basin takes a withering coefficient of soil as a quantitative index of the comprehensive drought risk of the carrier. By means of random forest and interpretable machine learning model shapley additive explanations (SHAP), hazard (H) of a disaster-causing factor in the basin, exposure (E) of a carrier, vulnerability (V) of the carrier and toughness (T) of the carrier are determined. Contribution degrees of the hazard, the exposure, the vulnerability and the toughness are decomposed and estimated. Weight coefficients al, a2, a3 and a4 of four indexes in comprehensive drought risk assessment are obtained.

[0042] A computation formula of comprehensive drought coefficient R is as follows:

[0043] R = H¥ x E& x V& x (1-T)*

[0044] R is a comprehensive drought risk, H is hazard, E is exposure, V is vulnerability,

T is toughness, al is a hazard weight coefficient, a2 is an exposure weight coefficient, a3 is a vulnerability weight coefficient and a4 is a toughness weight coefficient.

[0045] As shown in FIG. 4, a flow diagram of the drought risk early warning information distribution module in an example of the present invention is shown, which distributes early warning information to a client on the basis of a computer device and an IP address.

The computer device includes a cloud database, a cloud cache and a scenario early warning computation module.

[0046] The computer device may obtain daily soil wetness and precipitation remote sensing data to automatically compute relative soil wetness of a day, computes a number of days without precipitation of the day according to a past number of days without precipitation in the cloud cache, and computes hazard of a disaster-causing factor, exposure of a carrier, vulnerability of the carrier and toughness of the carrier of the day. In a case of a computed comprehensive risk index, through a historical-present-future rolling frequency method, a third-grade red early warning is issued for top 1% of rolling risk levels, a third- grade orange early warning is issued for top 5% of the rolling risk levels, and top 10% and top 30% correspond yellow and blue early warnings.

[0047] A scenario early warning computation module in the computer device provides two scenarios of numbers of days without precipitation in a next day and a next week, and computes comprehensive drought risks of no precipitation in the next day and the next week again by means of relative soil wetness of a number of days corresponding to a multi- year average. Computation results will be stored in an information management system.

The information management system may provide a short message reminding service for a drought risk early warning through positioning of an IP address of a subscriber. The information management system further supports a user to carry out self-service inquiry on comprehensive drought risk information according to an own IP address, which provides scientific and technological support for accurate policy-making of industrial and agricultural water usage.

[0048] According to the drought risk early warning information distribution module in an example of the present invention, by automatically computing the comprehensive drought risk, a plurality of scenarios of combinations of soil wetness and numbers of days without precipitation in the next day, the next week and the next month are preset, and whether an area of a user IP address will face drought risk stress is automatically determined. If there is less precipitation in an early stage in a user area and relative soil wetness remains within a low-level value range for a long time, a possibility that the area faces drought stress will be greatly increased. Through precise positioning of Beidou, a subscriber having a potential drought risk is matched. Drought early warning information in the user area will be sent to a client in a distributed subscription mode.

[0049] The present invention further provides a computer device. The computer device includes a processor and a storage medium. The storage medium stores a computer instruction. The computer instruction implements hazard, exposure, vulnerability and toughness assessment and comprehensive drought risk assessment method described in an example of the present invention when run by the processor, provides early warnings of different comprehensive drought risk scenarios, matches information of a user who is about to face risk stress by means of an IP address, and automatically distributes early warning information.

[0050] In the present invention, the term “an example”, “some examples”, “instance”, “particular instance” or “some instances” means that a particular feature, structure, material or characteristic described in combination with the example or instance is included in at least one example or instance of the present invention. In the description, the schematic expression of the above term is not necessarily directed to the same example or instance.

Furthermore, the particular feature, structure, material or characteristic described can be combined in any suitable manner in any one or more examples or instances. In addition, without any contradiction, a person skilled in the art can combine different examples or instances and features of the different examples or instances described in the description of the present invention.

[0051] Although the above examples are shown and described, it can be understood that the above examples are exemplary and cannot be construed as limitations on the present invention.

Changes, modifications, substitutions and variations made to the above examples by those of ordinary skill in the art all fall within the scope of protection of the present invention.

Claims

CONCLUSIONS

1. Drought risk assessment system, including: Hazard module, to use daily precipitation data and soil moisture data to evaluate the drought risk of the watershed, in which the hazard module corresponds to the first level hazard warning; Exposure module, to use NDVI and EVI to represent vegetation cover and calculate drought exposure in the corresponding watershed; Vulnerability module, to use the root water absorption model to calculate the water demand of the bearing body, and then calculate the vulnerability of the water supply and demand of the bearing body, in which the vulnerability module corresponds to the second-level hazard warning; Resilience module, to use the ARx model to calculate the resilience of vegetation in the carrier body under drought stress, that is, the resilience of the carrier body in the corresponding watershed; Integrated drought risk module, to use random forests and interpretable machine learning models to decompose and estimate the contribution levels of hazard, exposure, vulnerability and resilience, to calculate the integrated drought risk, in which the integrated drought risk module corresponds to the third-level hazard warning ; and Drought Risk Warning Information Distribution Module, to distribute drought risk warning information of the corresponding area to the user end.

2. The drought risk assessment system according to claim 1, characterized in that the hazard module refers to a hazard assessment module of a disaster-causing factor, that is, it first obtains a long-term daily relative soil humidity and the number of days without rain, and then combines meteorological drought and agricultural drought for double correction to accurately evaluate actual water deficit characteristics of carrier crops.

The drought risk assessment system according to claim 1, characterized in that the exposure module extracts NDVI and EVI data from remote sensing images, converts NDVI and EVl data into vegetation data, weight and fuses vegetation cover data to obtain exposure data of the supporting body.

4. The drought risk assessment system according to claim 1, characterized in that the vulnerability module calculates on the basis of a balance of water supply and demand for growth and development of the supporting body, that is, using a root water absorption model to quantitatively determine the water demand for supporting body growth and an available calculate water supply for multiple layers of soil.

5. The drought risk assessment system according to claim 1, characterized in that using the ARx model to calculate the resilience of vegetation in the supporting body under drought stress refers to using the ARx model to evaluate the response of vegetation to short periods of time. to evaluate -term climate anomalies and quantify the resilience of short-term load-bearing bodies on a global scale.

6. The drought risk assessment system according to claim 1, characterized in that the calculation formula for the integrated drought risk is: R=H al x<E a2 xV a3 *<(1-T) a4

Where R is the integrated drought risk, H is the hazard, E is the exposure, V is the vulnerability, T is the resilience, al is the hazard weight coefficient, a2 is the exposure weight coefficient, a3 is the vulnerability weight coefficient and a4 is the resilience weight coefficient.

The drought risk assessment system according to claim 1, characterized in that the third-level hazard warning includes a red warning, an orange third-level warning, a yellow third-level warning and a blue third-level warning.

8. The drought risk assessment system according to claim 1, characterized in that the distribution module of drought risk warning information uses a progressive warning mode based on IP address distribution of warning information, i.e. it first sets multiple scenarios of soil moisture and non-rainy days in the future for, determines whether the IP address area is facing drought risk stress, and then uses satellite-accurate positioning to match users and distribute drought warning information.

The drought risk assessment system according to claim 1, characterized in that the drought warning information includes a first-level hazard warning, a second-level hazard warning and a third-level hazard warning.

A computing device comprising a processor and a storage medium, wherein computer instructions are stored on the storage medium and the computer instructions are implemented by the processor during execution, as described in any one of claims 1 to 9, of the drought risk assessment system.