Agroclimate resource evaluation system based on climate mode driving

Abstract

The invention provides an agricultural climate resource evaluation system based on climate mode driving, and the system can generate future climate estimation data through the cooperation of a plurality of modules including data acquisition, preprocessing, statistical downscaling, mode evaluation, multi-mode collection, agricultural climate resource calculation, visual output and the like. According to the system, the prediction precision can be adjusted according to actual observation data, automatic data downloading and space-time matching are achieved through a Python script library, and a multi-level data output module provides provincial and city-county scale evaluation reports and an interactive visual platform. And real-time query and decision support is provided for the user, so that the precision and reliability of climate prediction are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural climate assessment, and particularly to an agricultural climate resource assessment system driven by climate models. Background Art

[0002] With the intensification of global climate change, agricultural production is facing unprecedented challenges. Traditional methods for assessing agricultural climate resources often rely on historical observational data and are unable to accurately predict the impact of future climate change on agricultural production. At the same time, due to the complexity and uncertainty of the climate system, the prediction results of a single climate model often have large errors and biases. Therefore, it is particularly important to develop an agricultural climate resource assessment system that can comprehensively consider multiple climate models, combine high-resolution observational data, and accurately predict future climate change. Currently, the CMIP (Coupled Model Intercomparison Project) series of data is widely used as the basis for climate prediction in the international climate prediction field. As the latest generation of climate prediction data, CMIP6 provides more abundant and detailed climate simulation results. However, how to effectively utilize these massive data and combine them with high-resolution observational data sets for accurate assessment of agricultural climate resources remains a technical problem to be solved urgently. In addition, the assessment of agricultural climate resources not only needs to consider the spatio-temporal distribution characteristics of key climate elements such as temperature and precipitation, but also needs to set reasonable thresholds for accumulated temperature and precipitation in combination with the actual needs of agricultural production to accurately reflect the dependence and utilization of agricultural production on climate resources. Therefore, an agricultural climate resource assessment system driven by climate models is proposed to solve the above problems. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an agricultural climate resource assessment system driven by climate models to solve at least the above problems.

[0004] The technical solution adopted by the present invention is as follows:

[0005] An agricultural climate resource assessment system driven by climate models, the system includes:

[0006] A data acquisition module, used to download global multi-climate model data from the Earth System Grid Federation (ESGF) platform and obtain the CN05.1 high-resolution observational data set;

[0007] A data preprocessing module, which performs spatial interpolation, time alignment, regional cropping, and format standardization on the global multi-climate model data;

[0008] A statistical downscaling module, which uses the method of Delta bias correction to downscale the global multi-climate model data to the regional resolution and generates future climate prediction data in combination with the observational data;

[0009] The pattern evaluation module quantitatively scores the simulation performance of multiple climate models based on the Taylor diagram and screens the optimal model set;

[0010] The multi-model ensemble module integrates multi-climate model data by the method of equal skill score weighting to generate ensemble prediction results;

[0011] The agricultural climate resource calculation module calculates agricultural climate resource indicators according to the preset accumulated temperature and precipitation thresholds;

[0012] The visualization output module generates spatio-temporal change trend charts and risk assessment reports.

[0013] Further, the statistical downscaling module specifically includes:

[0014] Based on the observed data and climate model data of the historical reference period, the Delta value of future climate change signal is calculated by linear change and proportional change;

[0015] The Delta value is superimposed on the observed data to generate a high-resolution future climate data set for the 0.25°×0.25° target grid.

[0016] Further, the pattern evaluation module further includes:

[0017] Construct a Taylor diagram using the standard deviation, correlation coefficient, and root mean square error;

[0018] Quantify the simulation ability of each climate model through the skill score S, and screen the models with a correlation coefficient R>0.8 and a root mean square error E'<1.5. The specific calculation formula is:

[0019] S = aR - bE'

[0020] Where S is the skill score, R is the correlation coefficient, E' is the root mean square error, and a and b are weight coefficients.

[0021] Further, the agricultural climate resource calculation module supports custom thresholds, including:

[0022] Calculate the corresponding accumulated temperature with thresholds taking 0°C, 5°C, 10°C, 15°C as the lower limits and 35°C, 40°C as the upper limits;

[0023] Generate monthly, quarterly, and annual precipitation indicators in combination with daily precipitation.

[0024] Further, it further includes a dynamic rebalancing module for:

[0025] Correct the deviation of the climate model ensemble results every six months;

[0026] Adjust the weight distribution according to the actual observed data to optimize the prediction accuracy.

[0027] Furthermore, the visualization output module further includes:

[0028] Generate spatio-temporal distribution maps of agricultural climate resources under multiple climate scenarios of SSP126, SSP245, and SSP585;

[0029] Output the migration path of the suitable planting area and the early warning report of extreme climate risks.

[0030] Furthermore, it also includes an adaptation strategy generation module, specifically:

[0031] Recommend crop variety adjustment plans according to the changing trend of climate resources;

[0032] Generate optimization suggestions for planting layout based on accumulated temperature balance.

[0033] Furthermore, the data preprocessing module uses the bilinear interpolation algorithm to unify the climate model data to a 0.25°×0.25° grid through the CDO tool.

[0034] Furthermore, it also includes a Python script library, and the Python script library is used for:

[0035] Automatically download CMIP6 data;

[0036] Batch process the spatio-temporal matching of climate models and observational data.

[0037] Furthermore, it also includes a multi-level data output module, and the multi-level data output module is used for

[0038] Output the provincial agricultural climate resource assessment report;

[0039] Output the refined planting zoning map at the city and county scales;

[0040] A Web-based interactive visualization platform for users to query and obtain decision support in real time.

[0041] Compared with the prior art, the beneficial effects of the present invention are:

[0042] 1. Through the synergistic effect of the statistical downscaling module and the model evaluation module, the system can screen out the optimal climate model ensemble and generate high-resolution future climate prediction data, thus significantly improving the accuracy and reliability of climate prediction.

[0043] 2. The agricultural climate resource calculation module supports users to set different thresholds for accumulated temperature, precipitation, etc. according to actual needs, making the evaluation results more in line with the actual situation and needs of agricultural production.

[0044] 3. The system provides multi-level data output forms including provincial agricultural climate resource assessment reports, refined planting zoning maps at the city and county scales, and a Web-based interactive visualization platform, facilitating users to query and use according to their actual needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] To clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for the description of the embodiments. Obviously, the following drawings are only the preferred embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0046] Figure 1 It is a schematic diagram of the overall structure of the system in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The following describes the principles and features of the present invention with reference to the drawings. The listed embodiments are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0048] Refer to Figure 1 , the present invention provides an agricultural climate resource assessment system driven by a climate model. The system includes:

[0049] A data acquisition module for downloading global multi-climate model data from the Earth System Grid Federation (ESGF) platform and obtaining the CN05.1 high-resolution observation dataset;

[0050] A data preprocessing module for performing spatial interpolation, time alignment, regional cropping, and format standardization on the global multi-climate model data;

[0051] A statistical downscaling module for downscaling the global multi-climate model data to the regional resolution using the Delta bias correction method and generating future climate prediction data by combining the observation data;

[0052] A model evaluation module for quantitatively scoring the simulation performance of multiple climate models based on a Taylor diagram and screening the optimal model set;

[0053] A multi-model ensemble module for integrating multi-climate model data by the method of skill score weighting to generate an ensemble prediction result;

[0054] An agricultural climate resource calculation module for calculating agricultural climate resource indicators according to preset accumulated temperature and precipitation thresholds;

[0055] A visualization output module for generating spatio-temporal change trend charts and risk assessment reports.

[0056] Exemplarily, the data acquisition module downloads data covering multiple climate scenarios from the ESGF platform and simultaneously obtains the high-resolution observational dataset CN05.1. The data preprocessing module then performs necessary preprocessing on this data, such as spatial interpolation, time alignment, regional cropping, and format standardization, to ensure the consistency and accuracy of the data. The statistical downscaling module uses the Delta method to reduce global climate model data to a specific regional resolution and combines observational data to predict future climate; the model evaluation module quantifies the simulation performance of each climate model through Taylor diagrams and selects the optimal model ensemble from them; the multi-model ensemble module integrates the data of these optimal models through a skill score weighting method to form an ensemble prediction result; the agricultural climate resource calculation module calculates various agricultural climate resource indicators according to preset thresholds such as accumulated temperature and precipitation; the visualization output module generates intuitive spatio-temporal change trend diagrams and risk assessment reports to provide strong support for decision-makers.

[0057] The statistical downscaling module specifically includes:

[0058] Based on the observational data and climate model data of the historical reference period, calculate the future climate change signal Delta value using linear change and proportional change;

[0059] Overlay the Delta value on the observational data to generate a high-resolution future climate dataset for the 0.25°×0.25° target grid.

[0060] Exemplarily, in the statistical downscaling module, the system calculates the future climate change signal Delta value through linear change and proportional change based on the observational data and climate model data of the historical reference period. Then, these Delta values are overlaid on the observational data to generate a high-precision future climate dataset with a resolution of 0.25°×0.25°, which will provide strong support for subsequent climate resource assessment and adaptation strategy formulation.

[0061] The model evaluation module further includes:

[0062] Construct a Taylor diagram using standard deviation, correlation coefficient, and root mean square error;

[0063] Quantify the simulation ability of each climate model through the skill score S, and select models with a correlation coefficient R>0.8 and a root mean square error E'<1.5. The specific calculation formula is:

[0064] S = aR - bE'

[0065] where S is the skill score, R is the correlation coefficient, E' is the root mean square error, and a and b are weight coefficients.

[0066] Exemplarily, in the model evaluation module, the system constructs a Taylor diagram by using three key indicators: standard deviation, correlation coefficient, and root mean square error, so as to quantitatively evaluate the simulation performance of each climate model. Then, the simulation ability of each model is further quantified through the skill score S, and high-quality models with a correlation coefficient greater than 0.8 and a root mean square error less than 1.5 are selected as the basis for subsequent analysis;

[0067] The following is an application example of the model evaluation module and its calculation process. This module constructs a Taylor diagram by using standard deviation, correlation coefficient, and root mean square error, and quantifies the simulation ability of each climate model through the skill score S, and selects models with a correlation coefficient > 0.8 and a root mean square error < 1.5.

[0068] Suppose we have simulation data of three climate models (Model A, Model B, and Model C) and corresponding observed data. Our goal is to evaluate the simulation performance of these three climate models and select the optimal model. Observed data: Obtained from the CN05.1 high-resolution observation dataset. Simulation data: Download the simulation data of Model A, Model B, and Model C from the CMIP6 platform. Calculate the standard deviation, correlation coefficient, and root mean square error: Standard deviation: Calculate the standard deviation of the observed data and the simulation data of each climate model to evaluate the dispersion degree of the data. Correlation coefficient: Calculate the correlation coefficient between the observed data and the simulation data of each climate model to evaluate the linear relationship between the simulation data and the observed data. Root mean square error: Calculate the root mean square error between the observed data and the simulation data of each climate model to evaluate the error size between the simulation data and the observed data. Construct a Taylor diagram by integrating the three indicators of standard deviation, correlation coefficient, and root mean square error on a polar coordinate diagram, that is, the Taylor diagram. The Taylor diagram can intuitively show the performance of each climate model in terms of simulation standard deviation, correlation coefficient, and root mean square error. Calculate the skill score S through the formula S = aR - bE', where R is the correlation coefficient, E' is the root mean square error, and a and b are weight coefficients that can be adjusted according to the actual situation. Select the optimal model according to the screening conditions of the skill score S and the correlation coefficient and root mean square error (correlation coefficient > 0.8 and root mean square error < 1.5) to obtain the example results:

[0069]

[0070] According to the screening conditions, the correlation coefficient of Model A is greater than 0.8 and the root mean square error is less than 1.5, so it is selected as the optimal model. While the correlation coefficient of Model B is lower than 0.8, and the root mean square error of Model C is higher than 1.5, so neither of them passes the screening.

[0071] The agricultural climate resource calculation module supports customizing thresholds, including:

[0072] Calculate the corresponding accumulated temperature using thresholds with lower limits of 0°C, 5°C, 10°C, 15°C and upper limits of 35°C, 40°C;

[0073] Generate monthly, quarterly and annual precipitation indicators by combining daily precipitation.

[0074] Exemplarily, the agro-climatic resource calculation module provides users with the function of customizing thresholds. Users can set different temperature thresholds (such as lower limits of 0°C, 5°C, 10°C, 15°C and upper limits of 35°C, 40°C) according to actual needs to calculate the corresponding accumulated temperature. At the same time, this module will also combine daily precipitation data to generate key indicators of monthly, quarterly and annual precipitation. These indicators will provide users with detailed information about agro-climatic resources.

[0075] This system also includes a dynamic rebalancing module for:

[0076] Perform bias correction on the results of the climate model ensemble every six months;

[0077] Adjust the weight distribution according to actual observed data to optimize the prediction accuracy.

[0078] Exemplarily, in order to maintain the accuracy and stability of the prediction results, the system also includes a dynamic rebalancing module. This module will perform bias correction on the results of the climate model ensemble every six months and adjust the weights of each model according to the latest observed data, so that the system can better adapt to the uncertainty of climate change and thus provide more accurate prediction results.

[0079] The visualization output module further includes:

[0080] Generate spatio-temporal distribution maps of agro-climatic resources under multiple climate scenarios of SSP126, SSP245, SSP585;

[0081] Output the migration path of the suitable planting area and the early warning report of extreme climate risks.

[0082] Exemplarily, the visualization output module not only has the function of generating spatio-temporal change trend maps and risk assessment reports, but also can generate spatio-temporal distribution maps of agro-climatic resources under multiple scenarios (such as SSP126, SSP245, SSP585). The distribution maps will clearly show the spatial distribution and change trends of agro-climatic resources under different climate change scenarios. At the same time, this module can also output the migration path of the suitable planting area and the early warning report of extreme climate risks, providing important reference information for agricultural production.

[0083] This system also includes an adaptive strategy generation module, specifically:

[0084] Recommend crop variety adjustment plans according to the changing trend of climate resources;

[0085] Generate optimization suggestions for planting layout based on accumulated temperature balance.

[0086] Exemplarily, according to the changing trend of climate resources, the system will recommend adjustment plans for crop varieties to users and generate optimization suggestions for planting layout based on accumulated temperature balance to help users better cope with the challenges brought by climate change, thereby ensuring the sustainability and stability of agricultural production.

[0087] The data preprocessing module adopts the bilinear interpolation algorithm and unifies the climate model data to a 0.25°×0.25° grid through the CDO tool.

[0088] Exemplarily, in the data preprocessing module, the bilinear interpolation algorithm is adopted, and the climate model data is unified to a 0.25°×0.25° grid through the CDO tool to ensure the spatial consistency and comparability of the data, providing a solid foundation for subsequent analysis and evaluation.

[0089] This system also includes a Python script library, and the Python script library is used for:

[0090] Automatically download CMIP6 data;

[0091] Batch process the spatio-temporal matching of climate models and observed data.

[0092] Exemplarily, for the convenience of users' use and operation, this system also integrates a Python script library. Through the Python script library, users can easily achieve the automatic download of CMIP6 data and the batch spatio-temporal matching of climate models and observed data, so as to reduce the user's usage threshold and operation difficulty.

[0093] This system also includes a multi-level data output module, and the multi-level data output module is used for

[0094] Output the provincial agricultural climate resource assessment report;

[0095] Output the refined planting zoning map at the city and county scales;

[0096] A web-based interactive visualization platform for users to query and obtain decision support in real time.

[0097] Exemplarily, to meet the needs of different users and application scenarios, the system provides multi-level data output functions, including provincial-level agricultural climate resource assessment reports, refined planting zoning maps at the city and county scales, and a Web-based interactive visualization platform, etc., thus providing users with a rich variety of choices, facilitating them to query and use relevant information according to their actual needs. At the same time, the Web-based interactive visualization platform also provides users with real-time query and decision support functions, enabling users to more conveniently obtain and use agricultural climate resource assessment information.

[0098] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. The agricultural climate resource assessment system driven by climate model is characterized by: The system comprises: Data acquisition module, used to download global multi-climate model data from the Earth System Grid Federation ESGF platform and obtain the CN05.1 high-resolution observation data set; Data preprocessing module, which performs spatial interpolation, temporal alignment, regional cropping and format standardization on global multi-climate model data; The statistical downscaling module uses the Delta bias correction method to downscale global multi-climate model data to regional resolution and combines it with observational data to generate future climate projection data; Model evaluation module, which quantitatively scores the simulation performance of multiple climate models based on Taylor diagrams and selects the optimal model set; The multi-model ensemble module integrates multi-climate model data through a skill score weighted method to generate ensemble prediction results; The agricultural climate resource calculation module calculates agricultural climate resource indicators based on preset accumulated temperature and precipitation thresholds; Visual output module generates spatiotemporal trend graphs and risk assessment reports.

2. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: The statistical downscaling module specifically includes: Based on the observational data and climate model data of the historical base period, the Delta value of the future climate change signal is calculated using linear change and proportional change; The Delta value is superimposed on the observation data to generate a high-resolution future climate dataset with a target grid of 0.25°×0.25°.

3. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: The mode evaluation module further comprises: Construct Taylor diagrams using standard deviation, correlation coefficient, and root mean square error; The simulation ability of each climate model is quantified by the skill score S, and the model with correlation coefficient R>0.8 and root mean square error E'<1.5 is selected. The specific calculation formula is: S=aR-bE' Among them, S is the skill score, R is the correlation coefficient, E' is the root mean square error, and a and b are weight coefficients.

4. The agricultural climate resource assessment system based on climate model drive according to claim 1, characterized in that: The agricultural climate resource calculation module supports custom thresholds, including: The corresponding accumulated temperature is calculated with the thresholds of 0℃, 5℃, 10℃, and 15℃ as the lower limits and 35℃ and 40℃ as the upper limits; Combine daily precipitation to generate monthly, quarterly and annual precipitation indicators.

5. The agricultural climate resource assessment system based on climate model drive according to claim 1, characterized in that: Also includes a dynamic rebalancing module for: Bias correction is performed on climate model ensemble results every six months; Adjust the weight distribution according to the actual observation data to optimize the prediction accuracy.

6. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: The visual output module further comprises: Generate spatiotemporal distribution maps of agricultural climate resources under multiple climate scenarios of SSP126, SSP245, and SSP585; Output migration paths of suitable planting areas and extreme climate risk warning reports.

7. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: It also includes an adaptive strategy generation module, specifically: Recommend crop variety adjustment plans based on climate resource change trends; Generate planting layout optimization suggestions based on accumulated temperature balance.

8. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: The data preprocessing module uses a bilinear interpolation algorithm to unify the climate model data into a 0.25°×0.25° grid through the CDO tool.

9. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: Also included is a Python script library for: Automatic download of CMIP6 data; Batch processing of spatiotemporal matching of climate models and observational data.

10. The agricultural climate resource assessment system driven by climate model according to claim 1, characterized in that: It also includes a multi-level data output module, which is used to output a provincial agricultural climate resource assessment report; Output city and county-scale refined planting zoning maps; A web-based interactive visualization platform for users to query and make decisions in real time.

Images