Method for constructing evapotranspiration simulation model of different ecological systems in Pearl River basin

Abstract

The invention belongs to the field of evapotranspiration prediction, and provides a Pearl River basin different ecosystem evapotranspiration simulation model construction method, which comprises the steps of collecting meteorological data, land utilization data and hydrological data of the Pearl River basin; simulating past and current evapotranspiration of the Pearl River basin by using an SWH model; using a CA-Markov model and a PLUS model to predict the future land utilization change of the Pearl River basin; based on the predicted future land utilization data, combining a GFDL-ESM2M model and a CMIP6 climate scene, and utilizing an SWH model to simulate evapotranspiration of different future ecosystems of the Pearl River basin; carrying out statistics on evapotranspiration data of different ecosystems in different periods; and analyzing a driving mechanism of evapotranspiration of different ecological systems in the Pearl River basin.

Description

Technical Field

[0001] The invention belongs to the field of evapotranspiration prediction, and in particular relates to a method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin. Background Art

[0002] Evapotranspiration is an important component of the hydrological cycle and climate regulation, reflecting the evaporation of water in the ecosystem and the transpiration process of plants. The Pearl River Basin is a typical complex ecosystem in southern China, covering a variety of ecosystems such as forests, grasslands, farmlands, and wetlands. Its evapotranspiration process is affected by multiple factors such as climate change, land use change, and ecosystem functions. Due to climate change and intensified human activities, evapotranspiration in the Pearl River Basin is undergoing significant temporal and spatial changes, which directly affects the regional water resources distribution, ecological balance, and land use decisions. Therefore, establishing an accurate and dynamic evapotranspiration simulation model that can comprehensively analyze the evapotranspiration changes of different ecosystems is of great significance for water resources management, climate change response, and ecological protection.

[0003] In the existing technology, the simulation of basin evapotranspiration mostly adopts a single model or only focuses on climate factors, while ignoring the impact of land use change and ecosystem complexity on the evapotranspiration process. This simplistic treatment method is difficult to accurately reflect the dynamic changes of complex ecosystems. Therefore, it is of great research and application value to develop a simulation method that can combine climate change, land use change and ecosystem function for multi-dimensional evapotranspiration prediction and analysis. Summary of the invention

[0004] In order to solve the problems in the prior art, the present invention provides a method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin, the method comprising the following steps:

[0005] Step 1, data collection: collect meteorological data, land use data and hydrological data of the Pearl River Basin, wherein the meteorological data include temperature, precipitation, humidity and solar radiation, the land use data include remote sensing images and land cover classification data, and the hydrological data include flow and water level data of hydrological stations in the basin;

[0006] Step 2, SWH model simulation of evapotranspiration: based on the meteorological data, land use data and hydrological data, using the SWH model to simulate the past and present evapotranspiration of the Pearl River Basin;

[0007] Step 3, future land use prediction: The CA-Markov model and PLUS model are used to predict the future land use change in the Pearl River Basin. The prediction process is based on the future climate scenario data provided by the GFDL-ESM2M model and CMIP6.

[0008] Step 4: Simulation of future evapotranspiration: Based on the predicted future land use data, combined with the GFDL-ESM2M model and CMIP6 climate scenarios, the SWH model is used to simulate the evapotranspiration of different ecosystems in the Pearl River Basin in the future;

[0009] Step 5, spatiotemporal evolution analysis: by statistically simulating the evapotranspiration data of different periods and different ecosystems, the spatiotemporal evolution dynamics of evapotranspiration of different ecosystems such as forests, grasslands and farmlands in the Pearl River Basin are analyzed;

[0010] Step 6: Analysis of driving mechanism: Based on the spatiotemporal variation data of evapotranspiration, combined with climate change, land use change and ecosystem function factors, the driving mechanism of evapotranspiration in different ecosystems in the Pearl River Basin is analyzed.

[0011] The present invention provides a method for constructing a simulation model for evapotranspiration of different ecosystems in the Pearl River Basin. The method combines meteorological data, land use data and hydrological data, and uses the SWH model, CA-Markov model, PLUS model, GFDL-ESM2M model and CMIP6 climate scenario to comprehensively simulate and predict the changes in evapotranspiration in the Pearl River Basin in the past, present and future. Through the effective combination of models, the present invention has the following beneficial effects:

[0012] (1) The model of the present invention combines multiple data sources and models, including meteorological data, land use data and hydrological data, and can comprehensively analyze the dynamic changes of evapotranspiration from multiple dimensions such as climate change, land use change and ecosystem function, thus making up for the shortcomings of the single model in the prior art.

[0013] (2) The SWH model, combined with historical and current meteorological data, land use data, and hydrological data, can accurately simulate the past and present changes in evapotranspiration in the Pearl River Basin. In addition, by combining the CA-Markov model with the PLUS model, future land use changes can be predicted, and then combined with future climate scenarios, future evapotranspiration can be dynamically simulated, making the prediction results more comprehensive and accurate.

[0014] (3) The present invention combines the time-lapse characteristics of the CA-Markov model with the spatial complexity processing capability of the PLUS model to ensure the accuracy of the temporal trend prediction and spatial distribution simulation of land use change. At the same time, the future climate data provided by the GFDL-ESM2M and CMIP6 climate scenarios enable the model to simulate under multiple climate scenarios, ensuring the multi-scenario adaptability of evapotranspiration prediction.

[0015] (4) By statistically analyzing the evapotranspiration data of different ecosystems at different times, the present invention can reveal the spatiotemporal evolution of evapotranspiration of various ecosystems in the Pearl River Basin, helping managers to better understand the water balance and dynamic changes of ecosystems.

[0016] (5) By combining climate change, land use change and ecosystem function, a multivariate regression analysis model was established to clarify the main driving mechanism of evapotranspiration in different ecosystems in the Pearl River Basin, which helps to provide a scientific basis for basin water resources management, land use decision-making and climate change response.

[0017] In summary, the present invention overcomes the limitations of the existing technology by combining multiple models, and realizes a comprehensive, accurate and dynamic simulation of the evapotranspiration of different ecosystems in the Pearl River Basin, which can provide strong support for ecological protection, climate change response and water resources management in the Pearl River Basin. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0020] Below, the invention is preferably described in conjunction with the accompanying drawings and specific implementation methods.

[0021] This embodiment solves the above problem through the following steps:

[0022] In one embodiment, reference Figure 1 The present invention provides a method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin. The different ecosystems refer to various natural or artificial ecosystems in the Pearl River Basin, including but not limited to forests, grasslands, farmlands, wetlands, urban green spaces, etc. These ecosystems have significant differences in structure, function and land use characteristics; the evapotranspiration refers to the process of water in an ecosystem through plant transpiration and soil evaporation, reflecting an important indicator of water consumption and energy balance in the ecosystem, affecting regional hydrological cycles and climate change. Based on existing data, methods and models, the present invention establishes a mathematical model or computer model that can predict and simulate the changes in evapotranspiration of different ecosystems in the Pearl River Basin through a systematic modeling process for scientific research and management decision-making.

[0023] Step 1, data collection: collect meteorological data, land use data and hydrological data of the Pearl River Basin. The meteorological data include temperature, precipitation, humidity and solar radiation, etc. The land use data include remote sensing images and land cover classification data. The hydrological data include flow and water level data of hydrological stations in the basin.

[0024] This step involves collecting basic data of the Pearl River Basin to support subsequent evapotranspiration simulation analysis. The data include but are not limited to the following categories:

[0025] Meteorological data: includes multiple meteorological elements within the Pearl River Basin, including temperature, precipitation, humidity, solar radiation and other data. The temperature data refers to the temporal and spatial distribution of surface temperature in the basin, usually obtained through meteorological station observations or reanalysis data; the precipitation data refers to annual precipitation, monthly precipitation or daily precipitation in the basin and other parameters closely related to the water cycle, which can reflect the temporal and spatial variation characteristics of precipitation in the Pearl River Basin; the humidity data refers to the water vapor content in the air, including relative humidity and absolute humidity data; the solar radiation data refers to the solar energy received by the Pearl River Basin, reflecting the degree of influence of the climate system on evaporation, usually estimated through radiation observations or satellite data.

[0026] Land use data: This type of data includes information on land cover conditions in the Pearl River Basin, specifically including remote sensing image data and land cover classification data. The remote sensing image data refers to image data obtained through remote sensing satellites such as Landsat and MODIS, which can reflect the surface characteristics of the Pearl River Basin at different times; the land cover classification data refers to classification data based on remote sensing images or field surveys, which classifies land use types into forests, grasslands, farmland, urban construction land and other categories according to a specific classification system, which is used for subsequent land use change and evapotranspiration analysis.

[0027] Hydrological data: including hydrological observation data of the main river systems in the Pearl River Basin, including flow, water level and other data measured by hydrological stations established in the basin. The flow data refers to the water flow observation values ​​of the main rivers in the basin at different time periods, which are used to analyze the distribution of water resources and their changes; the water level data refers to the water level height of rivers, lakes, reservoirs and other water bodies, which can reflect the changing trend of hydrological conditions in the Pearl River Basin.

[0028] The collection of the above-mentioned meteorological data, land use data and hydrological data provides basic data support for the subsequent simulation of evapotranspiration of different ecosystems in the Pearl River Basin.

[0029] Step 2, SWH model simulation of evapotranspiration: Based on the meteorological data, land use data and hydrological data, the SWH model is used to simulate the past and present evapotranspiration of the Pearl River Basin.

[0030] This step is based on the meteorological data, land use data and hydrological data, and uses the SWH model (Soil-Water-Heat coupling model) to simulate the past and present evapotranspiration of the Pearl River Basin. The specific implementation method is as follows:

[0031] The SWH model, or coupled soil-water-heat model, is a dynamic simulation model that integrates soil water balance, vegetation physiological characteristics, and atmospheric energy exchange processes. The model can accurately simulate evapotranspiration by describing the interactions between soil, plants, and the atmosphere at multiple levels. The inputs to the SWH model include meteorological parameters (temperature, precipitation, humidity, solar radiation, etc.), land use parameters (vegetation coverage, land cover type), and hydrological parameters (soil moisture, water storage, and watershed hydrological characteristics), which work together to simulate the temporal and spatial distribution of evapotranspiration.

[0032] Evapotranspiration refers to the total amount of water lost to the atmosphere through soil evaporation and plant transpiration, and is a key component of the hydrological cycle. The magnitude of evapotranspiration is affected by meteorological conditions, vegetation conditions, land use, and hydrological characteristics. Therefore, simulating evapotranspiration at different time points can reveal the dynamic balance of water in the basin.

[0033] The specific method used by the SWH model to simulate past and present evapotranspiration in the Pearl River Basin is as follows:

[0034] First, the input parameters required for the SWH model are determined based on the collected meteorological data, land use data, and hydrological data. Specifically, they include:

[0035] Meteorological data: historical and current data such as temperature, precipitation, humidity, and solar radiation, mainly extracted from meteorological stations or reanalysis datasets, used to drive the evaporation and transpiration processes in the model;

[0036] Land use data: including the distribution and changes of different ecosystems (such as forests, grasslands, farmlands, etc.) in the Pearl River Basin. These data can be obtained through remote sensing image interpretation;

[0037] Hydrological data: data such as flow, water level, and soil moisture recorded by hydrological stations within the basin, reflecting the water status of the basin.

[0038] Model parameterization and calibration: During the simulation process, the key parameters in the SWH model were adapted and calibrated according to the actual situation of the Pearl River Basin, mainly involving the following parameters:

[0039] Soil parameters: including physical properties of soil, such as permeability, field water holding capacity, thermal conductivity of soil, etc.;

[0040] Vegetation parameters: such as vegetation type, leaf area index (LAI), root depth, etc., directly affect transpiration;

[0041] Hydrological characteristic parameters: such as basin slope, topography, river channel structure, etc.

[0042] Model operation and evapotranspiration simulation:

[0043] Historical meteorological data, land use data and hydrological data were input into the SWH model to simulate the evapotranspiration of the Pearl River Basin in different historical periods. Through step-by-step iterative calculations over multiple periods, the changes in evapotranspiration in the basin in the past were obtained.

[0044] By using current meteorological data and land use data and inputting them into the model, real-time or current period evapotranspiration simulation is carried out to obtain the current evapotranspiration distribution in the Pearl River Basin.

[0045] Through the above implementation steps, the SWH model can accurately simulate the past and present evapotranspiration in the Pearl River Basin, providing a scientific basis for evaluating the basin's water resources dynamics, climate change impacts, and ecological effects of land use.

[0046] Step 3, future land use prediction: The CA-Markov model and PLUS model are used to predict future land use changes in the Pearl River Basin. The prediction process is based on the future climate scenario data provided by the GFDL-ESM2M model and CMIP6.

[0047] This step combines the CA-Markov model with the PLUS model to predict future land use changes in the Pearl River Basin. The prediction process is based on the future climate scenario data provided by the GFDL-ESM2M model and CMIP6.

[0048] The CA-Markov model, or the Cellular Automaton-Markov Chain model, is a land use prediction model that combines the time-lapse characteristics of the Markov chain with the spatial dynamic change characteristics of the cellular automaton. The model first calculates the temporal conversion probability of land use categories through the Markov chain, and then uses the cellular automaton to dynamically simulate the spatial distribution of these probabilities to generate the spatial change pattern of land use. The CA-Markov model is good at describing the overall conversion trend and spatial pattern changes of land use, but it has certain limitations when dealing with complex ecosystem structures.

[0049] The PLUS model (Patch-generating Land Use Simulation) is a land use simulation model based on patch generation, which is specially used to deal with complex land use changes. The model can better capture the spatial heterogeneity and complex structural changes of land use by generating spatial patches and simulating the dynamic process of the ecosystem. The PLUS model combines multiple driving factors, such as socio-economic, environmental constraints, and policy changes, and is suitable for large-scale land use change prediction.

[0050] The GFDL-ESM2M model, Geophysical Fluid Dynamics Laboratory Earth System Model, is a global climate model used to simulate future climate change. It can provide future climate change scenario data for the Pearl River Basin, including key meteorological elements such as temperature and precipitation. This model is part of the CMIP6 project.

[0051] CMIP6, or Coupled Model Intercomparison Project Phase 6, is an international climate change research program that aims to predict future climate through a variety of climate models and provide different climate scenario data, such as RCP (Representative Concentration Pathway) and SSP (Shared Socioeconomic Pathway), for assessing the impact of future climate on the environment and society.

[0052] Specific implementation steps:

[0053] Data input and preprocessing:

[0054] First, the future climate scenario data provided by the GFDL-ESM2M model and the CMIP6 project were preprocessed, which included meteorological elements such as future temperature and precipitation in the Pearl River Basin. These meteorological data were used to simulate the impact of future land use changes under different scenarios.

[0055] At the same time, current and historical land use data (such as forests, grasslands, farmlands, urban expansion, etc.) in the Pearl River Basin, as well as socio-economic and policy change factors were collected and input as initial inputs for the CA-Markov model and the PLUS model.

[0056] Time prediction of CA-Markov model:

[0057] The conversion probabilities between various land use types in the Pearl River Basin are calculated through the Markov chain, and the land areas of various land use types in the Pearl River Basin at different times are queried. Based on the land areas, the conversion probabilities between various land use types in the Pearl River Basin are calculated through the Markov chain:

[0058]

[0059] Among them, P ij A represents the conversion probability of the i-th type to the j-th type in each land use type. ij (t, t+1) represents the area converted from the i-th land use type to the j-th land use type within the time interval (t, t+1), A i(t) represents the area of ​​the i-th land use type at time t;

[0060] Using a preset matrix function to perform matrix reduction generation processing on the conversion probability to obtain a probability matrix;

[0061] Based on the probability matrix, a conversion matrix of different land use categories at a specific future time point is generated.

[0062] Subsequently, cellular automata combined with the transformation matrix were used to simulate the dynamic evolution of land use on a spatial scale and generate a preliminary land use change map for a certain period in the future, mainly to predict the overall spatial distribution pattern of land use.

[0063] In detail, the matrix function is a function used to construct a matrix, such as a zero matrix function, and the probability matrix is ​​a square matrix constructed by conversion probabilities.

[0064] Spatial simulation of PLUS model:

[0065] After obtaining the land use change pattern initially predicted by the CA-Markov model, the PLUS model was used to simulate and refine the spatial patches. The PLUS model can capture more complex land use change characteristics, including the patchy effect of ecosystems and the spatial heterogeneity of land use types.

[0066] The PLUS model combines multidimensional driving factors, such as policy influences, socio-economic drivers, and environmental constraints, and can simulate the spatial structure and functional zoning of future land use in more detail.

[0067] Advantages of model combination:

[0068] The advantage of the CA-Markov model is that it can predict the temporal changes in land use and can more accurately describe the conversion trends and time series changes of land use types.

[0069] The PLUS model performs well in dealing with spatial patches and ecological complexity, and can further refine the spatial pattern and accurately simulate the dynamic changes of land use in local areas.

[0070] Combining the CA-Markov model with the PLUS model can not only predict the overall change trend of land use types in time, but also accurately describe the detailed changes in land use in space, making the prediction results more comprehensive and accurate.

[0071] Prediction result output and application:

[0072] Finally, by combining the GFDL-ESM2M model and the climate scenarios provided by CMIP6, a joint simulation of the CA-Markov model and the PLUS model was used to generate a map of land use change in the Pearl River Basin at different times in the future. This result can provide data support for subsequent ecosystem evapotranspiration simulation and environmental management.

[0073] At the same time, the prediction results can be used to analyze the changing trends of land use under different climate scenarios, helping to formulate targeted land management policies and ecological protection measures.

[0074] Through the above steps, the combination of CA-Markov model and PLUS model not only improves the temporal accuracy of prediction, but also enhances the accuracy of spatial simulation, providing comprehensive prediction results for future land use changes in the Pearl River Basin.

[0075] Step 4: Simulation of future evapotranspiration: Based on the predicted future land use data, combined with the GFDL-ESM2M model and CMIP6 climate scenarios, the SWH model is used to simulate the evapotranspiration of different ecosystems in the Pearl River Basin in the future.

[0076] This step is based on the predicted future land use data, combined with the GFDL-ESM2M model and the CMIP6 climate scenario, and uses the SWH model to simulate the evapotranspiration of different ecosystems in the Pearl River Basin in the future.

[0077] Specific implementation steps for future evapotranspiration simulation:

[0078] Model input preparation:

[0079] First, the predicted future land use data (the future land use change results generated by the CA-Markov model and the PLUS model) were imported into the SWH model. These data provide the spatial distribution and pattern changes of different ecosystems in the Pearl River Basin in the future.

[0080] At the same time, the future meteorological data (such as future temperature, precipitation, humidity, etc.) provided by the GFDL-ESM2M model and the CMIP6 climate scenario are imported into the SWH model. These meteorological data are used to simulate the water cycle and energy transfer process of the ecosystem under different climate scenarios.

[0081] Parameter adjustment and calibration of SWH model:

[0082] According to the actual situation of the Pearl River Basin, the key parameters in the SWH model are adjusted and calibrated to ensure that the model can adapt to the evapotranspiration process of different ecosystems in the future. The parameters include soil type, vegetation coverage, soil moisture content, etc.

[0083] Soil parameters affect the evaporation process, vegetation parameters determine the transpiration intensity of plants, and hydrological parameters (such as surface runoff and soil permeability) affect the flow and storage of water in the ecosystem.

[0084] Evapotranspiration simulation process:

[0085] Using the input of future meteorological data and land use data, the SWH model dynamically simulates the evaporation and transpiration of the Pearl River Basin in different periods in the future by calculating the evaporation and transpiration process of water in each ecosystem. The SWH model can gradually simulate the changes in evaporation in specific years in the future (such as 2025 and 2030) and output the spatiotemporal distribution of evaporation in each ecosystem.

[0086] The combination of future land use data and future climate scenario data is crucial because land use and climate conditions jointly determine the evapotranspiration process. Land use changes affect the spatial pattern of different ecosystems in the basin. For example, an increase or decrease in forest cover will directly affect evapotranspiration, while climate change regulates the evapotranspiration process through meteorological factors such as temperature and precipitation. Therefore, relying on a single data source is not enough to fully reflect future changes in evapotranspiration.

[0087] The GFDL-ESM2M model and CMIP6 climate scenarios provide a variety of scenario assumptions for future climate conditions, enabling evapotranspiration predictions to be simulated multiple times under different greenhouse gas emission pathways and climate conditions, thereby improving the adaptability and scientificity of the predictions.

[0088] The future land use changes predicted by the CA-Markov model and the PLUS model can accurately reflect the changes in the spatial pattern of different ecosystems. When combined with the SWH model, these data can be input to simulate the direct impact of land use changes on evapotranspiration, ensuring that the simulation results are more in line with the actual scenario.

[0089] Simulation results analysis and output:

[0090] The SWH model outputs the spatiotemporal distribution of evapotranspiration of different ecosystems (such as forests, grasslands, farmlands, etc.) in the Pearl River Basin at different periods in the future, showing the dynamic changes of evapotranspiration in various periods in the future.

[0091] Combining different climate scenarios (such as high emission scenarios and low emission scenarios), the changing trend of future evapotranspiration is analyzed, and the potential impact of climate change and land use change on the water resources balance in the Pearl River Basin is evaluated.

[0092] Through the above steps, combined with future land use data, GFDL-ESM2M model and CMIP6 climate scenario, the SWH model can dynamically simulate the evapotranspiration changes of different ecosystems in the Pearl River Basin in the future. This model combination can provide more accurate evapotranspiration prediction results and provide data support for future water resources management, ecological protection and land use planning.

[0093] Step 5, spatiotemporal evolution analysis: By statistically simulating the evapotranspiration data of different periods and different ecosystems, the spatiotemporal evolution dynamics of evapotranspiration of different ecosystems such as forests, grasslands, and farmlands in the Pearl River Basin are analyzed.

[0094] This step analyzes the spatiotemporal evolution of evapotranspiration of different ecosystems such as forests, grasslands, and farmlands in the Pearl River Basin by statistically simulating evapotranspiration data of different periods and different ecosystems.

[0095] Spatiotemporal evolution dynamics refers to the temporal changes and spatial distribution characteristics of evapotranspiration of different ecosystems within a certain time and space range. By comprehensively analyzing the time dimension (such as the past, present and future) and the spatial dimension (such as different regions and different types of ecosystems), the changing patterns of evapotranspiration are revealed. The analysis of spatiotemporal evolution dynamics helps to understand the trend changes and potential driving factors in the hydrological cycle of ecosystems.

[0096] The evapotranspiration data of the Pearl River Basin at different periods (e.g., 2025 and 2030) output from the SWH model are collated. The data include the evapotranspiration of different ecosystems in the Pearl River Basin, such as forests, grasslands, and farmlands. The evapotranspiration of each ecosystem is classified and recorded by time node and spatial location to ensure that the evapotranspiration characteristics of each ecosystem in different periods and regions can be analyzed in detail.

[0097] In the time dimension, the average evapotranspiration, total evapotranspiration and evapotranspiration change rate of each ecosystem in different periods are calculated by statistically simulating evapotranspiration data. Specific calculations can include interannual changes, seasonal changes and long-term trend analysis. Through these statistics, the temporal evolution trend of evapotranspiration of different ecosystems in the Pearl River Basin can be intuitively revealed.

[0098] In the spatial dimension, the spatial distribution of evapotranspiration of each ecosystem is analyzed and the spatial distribution map of evapotranspiration is drawn using GIS technology to intuitively display the spatial differences and pattern changes of evapotranspiration of various ecosystems in the Pearl River Basin in different periods. For example, it is analyzed whether there is a trend of increasing evapotranspiration in forest-covered areas, or whether grasslands and farmlands show different evapotranspiration patterns.

[0099] Step 6: Analysis of driving mechanisms: Based on the spatiotemporal variation data of evapotranspiration, combined with factors such as climate change, land use change and ecosystem function, the main driving mechanisms of evapotranspiration in different ecosystems in the Pearl River Basin are analyzed.

[0100] This step is based on the spatiotemporal variation data of evapotranspiration, combined with factors such as climate change, land use change and ecosystem function, to analyze the main driving mechanisms of evapotranspiration in different ecosystems in the Pearl River Basin.

[0101] Specific implementation steps:

[0102] Data preparation and preprocessing:

[0103] The spatiotemporal variation data of evapotranspiration obtained in step 5 are used as the basis, including the statistical data of evapotranspiration in different periods, different regions and different ecosystems.

[0104] Collect and integrate climate data related to evapotranspiration changes (such as temperature, precipitation, humidity, etc.), land use change data (such as urban expansion, forest cover changes, agricultural land changes, etc.) and ecosystem function data (such as vegetation cover, species diversity, soil type, etc.).

[0105] Multiple regression analysis of driving factor data:

[0106] By establishing a multiple regression model, taking the spatiotemporal variation data of evapotranspiration as the dependent variable, taking factors such as climate change, land use change and ecosystem function as independent variables, the contribution rate of each driving factor to the change of evapotranspiration is analyzed, wherein the steps of analyzing the contribution rate of each driving factor to the change of evapotranspiration are: performing data cleaning on the spatiotemporal variation data to obtain target spatiotemporal variation data, standardizing the target spatiotemporal variation data to obtain standard variation data, and based on the standard variation data, respectively calculating the variances corresponding to the factors of climate change, land use change and ecosystem function to obtain the variance of climate factors, the variance of land factors and the variance of ecological factors, and obtaining the residual coefficient corresponding to the multiple regression model, combining the variance of climate factors, the variance of land factors, the variance of ecological factors and the residual coefficient, and analyzing the contribution rate of each driving factor to the change of evapotranspiration through the multiple regression model:

[0107]

[0108] Among them, E1, E2, and E3 represent the contribution rate of each driving factor to the change of evapotranspiration, and F a represents the variance of climate factors, F b represents the variance of land factors, F c represents the variance of ecological factors, and β represents the residual coefficient.

[0109] Among them, the target spatiotemporal change data is the data obtained after the erroneous data in the spatiotemporal change data is cleared, the standard change data is the data obtained after the differences between the target spatiotemporal change data are eliminated, and the residual coefficient is the error corresponding to the multivariate regression model; further, the data cleaning of the spatiotemporal change data can be achieved through the box plot method; the standardization processing of the target spatiotemporal change data can be achieved through the Z-score standardization method; the variance corresponding to climate change, land use change and ecosystem function factors can be calculated through the variance calculation formula.

[0110] For example, climate change can reflect the intensity of its impact on evapotranspiration through the regression coefficients of temperature and precipitation changes; land use change can reflect its impact on evapotranspiration through the proportional changes of different land types; and ecosystem function can reflect its regulatory effect on evapotranspiration through parameters such as vegetation coverage and soil moisture.

[0111] Spatial-temporal correlation analysis:

[0112] Correlation analysis (such as the Pearson correlation coefficient) was used to assess the strength of the association between climate change, land use change and evapotranspiration in different time periods and spatial ranges. Spatiotemporal correlation analysis can reveal the main driving mechanisms of evapotranspiration in different ecosystems under different climate scenarios. For example, the driving effect of climate in alpine forest areas may be more significant than that in urbanized areas.

[0113] In detail, in different time periods and spatial ranges, the correlation strength between climate change, land use change and evapotranspiration can be evaluated through correlation analysis through the following steps: vectorize the climate change, land use change and evapotranspiration respectively to obtain the climate change vector, land use change vector and evapotranspiration vector, calculate the means corresponding to the climate change vector, land use change vector and evapotranspiration vector respectively, obtain the first vector mean, the second vector mean and the third vector mean, combine the climate change vector, evapotranspiration vector, the first vector mean and the third vector mean, and calculate the climate-evaporation correlation coefficient between the climate change and evapotranspiration through the following formula:

[0114]

[0115] Where H represents the correlation coefficient between climate change and evapotranspiration, n represents the total number of climate change vectors and evapotranspiration vectors, and FT e represents the e-th vector in the evapotranspiration vector, C f

[0116] represents the fth vector in the climate change vector, e and f represent the serial numbers corresponding to the evapotranspiration vector and the climate change vector, respectively. and denote the first vector mean and the third vector mean respectively;

[0117] The land-evaporation correlation coefficient between land-use change and evapotranspiration was calculated by combining the land-use change vector, evapotranspiration vector, second vector mean, and third vector mean;

[0118] Combining the climate-evaporation correlation coefficient and the land-evaporation correlation coefficient, the correlation between climate change, land use change and evapotranspiration was analyzed to obtain the first correlation and the second correlation;

[0119] The first and second correlations were combined to assess the strength of the association between climate change, land use change and evapotranspiration.

[0120] Among them, the climate change vector, land use change vector and evapotranspiration vector are the expression vectors corresponding to the evaluation of climate change, land use change and evapotranspiration respectively, and the first correlation and the second correlation are the correlations between climate change, land use change and evapotranspiration respectively.

[0121] Furthermore, the vectorization processing of evaluating climate change, land use change and evapotranspiration can be achieved through vector models, such as the word2vec model; combining the values ​​of the climate-evaporation correlation coefficient and the land-evaporation correlation coefficient, analyzing the correlation between climate change, land use change and evapotranspiration, such as the higher the correlation coefficient, the higher the correlation; combining the first correlation and the second correlation, evaluating the correlation strength between climate change, land use change and evapotranspiration, such as the higher the correlation, the higher the correlation strength.

[0122] Climate change, land use change and ecosystem function have different action paths and impact patterns on evapotranspiration. By combining different models (such as SWH model, CA-Markov model and PLUS model) to integrate these factors, it can ensure that the driving mechanism of evapotranspiration is fully reflected in time, space and ecological processes.

[0123] Climate models (such as GFDL-ESM2M and CMIP6) can provide future climate scenario data to help explain the impact of climate change on evapotranspiration; land use change models (such as CA-Markov and PLUS) can accurately simulate the spatial pattern of future land use and reveal the intervention of human activities in the evapotranspiration process; SWH models can simulate the dynamic circulation process of water in the ecosystem and reflect the contribution of different ecosystem functions to evapotranspiration.

[0124] The combination of models can capture the interaction of multiple driving factors. For example, climate change may affect soil moisture by changing precipitation, thereby changing evapotranspiration, while land use change may directly affect regional transpiration through changes in vegetation cover. By integrating these factors, the analysis of driving mechanisms will be more scientific and comprehensive.

[0125] Output and application of driving mechanism results:

[0126] By combining regression analysis, correlation analysis and models, we obtained a report on the main driving mechanisms of evapotranspiration in different ecosystems in the Pearl River Basin. The results can reveal which factors drive the changes in evapotranspiration of each ecosystem in different periods, such as rising temperatures, loss of forest cover or urban expansion.

[0127] These results can be used in areas such as watershed management, climate change adaptation and ecological protection, helping relevant decision makers to take targeted regulatory measures. For example, by adjusting land use planning or strengthening climate change monitoring, the negative impact of climate and human activities on evapotranspiration can be reduced.

[0128] Through the above steps, the driving mechanism analysis will reveal the main influencing factors and interactions of evapotranspiration changes in the Pearl River Basin. By combining climate models, land use models and ecosystem function data, the complex driving mechanisms behind the spatiotemporal evolution of evapotranspiration can be fully reflected, providing a basis for the formulation of water resources management and ecosystem protection policies.

[0129] The prior art mentioned in the aforementioned background technology part and specific embodiment part of the present invention may be regarded as a part of the present invention, and is used to understand the meaning of some technical features or parameters.

Claims

1. A method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin, characterized in that: The method comprises the following steps: Step 1, data collection: collect meteorological data, land use data and hydrological data of the Pearl River Basin, wherein the meteorological data include temperature, precipitation, humidity and solar radiation, the land use data include remote sensing images and land cover classification data, and the hydrological data include flow and water level data of hydrological stations in the basin; Step 2, SWH model simulation of evapotranspiration: based on the meteorological data, land use data and hydrological data, using the SWH model to simulate the past and present evapotranspiration of the Pearl River Basin; Step 3, future land use prediction: The CA-Markov model and PLUS model are used to predict the future land use change in the Pearl River Basin. The prediction process is based on the future climate scenario data provided by the GFDL-ESM2M model and CMIP6. Step 4: Simulation of future evapotranspiration: Based on the predicted future land use data, combined with the GFDL-ESM2M model and CMIP6 climate scenarios, the SWH model is used to simulate the evapotranspiration of different ecosystems in the Pearl River Basin in the future; Step 5, spatiotemporal evolution analysis: by statistically simulating the evapotranspiration data of different periods and different ecosystems, the spatiotemporal evolution dynamics of evapotranspiration of different ecosystems such as forests, grasslands and farmlands in the Pearl River Basin are analyzed; Step 6: Analysis of driving mechanism: Based on the spatiotemporal variation data of evapotranspiration, combined with climate change, land use change and ecosystem function factors, the driving mechanism of evapotranspiration in different ecosystems in the Pearl River Basin is analyzed.

2. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The remote sensing image data refers to image data obtained by Landsat or MODIS remote sensing satellites.

3. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The simulation of the past and present evapotranspiration in the Pearl River Basin using the SWH model includes: Determine the input parameters required for the SWH model based on the collected meteorological data, land use data, and hydrological data; The historical meteorological data, land use data and hydrological data were input into the SWH model to simulate the evapotranspiration of the Pearl River Basin in different historical periods. Through the step-by-step iterative calculation of multiple periods, the changes in the evapotranspiration of the basin in the past were obtained. By using current meteorological data and land use data and inputting them into the model, real-time or current period evapotranspiration simulation is carried out to obtain the current evapotranspiration distribution in the Pearl River Basin.

4. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 3 comprises: Preprocess the future climate scenario data provided by the GFDL-ESM2M model and the CMIP6 project; The conversion probabilities between various land use types in the Pearl River Basin are calculated by using the Markov chain, the land areas of various land use types in the Pearl River Basin at different times are queried, and the conversion probabilities between various land use types in the Pearl River Basin are calculated by using the Markov chain according to the land areas; Using a preset matrix function to perform matrix reduction generation processing on the conversion probability to obtain a probability matrix; According to the probability matrix, a conversion matrix of different land use categories at a preset future time point is generated; By combining cellular automata with the transformation matrix, the dynamic evolution of land use is simulated on a spatial scale, a preliminary land use change map for a certain period in the future is generated, and the overall spatial distribution pattern of land use is predicted.

5. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 4, characterized in that: The step 3 also includes: After obtaining the land use change pattern preliminarily predicted by the CA-Markov model, the PLUS model was used to simulate and refine the spatial patches.

6. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 5, characterized in that: The step 3 also includes: Combining the GFDL-ESM2M model and the climate scenarios provided by CMIP6, a land use change map of the Pearl River Basin in different future periods was generated through the joint simulation of the CA-Markov model and the PLUS model.

7. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 4 comprises: The future land use change results generated by the CA-Markov model and the PLUS model are imported into the SWH model; Import future meteorological data provided by the GFDL-ESM2M model and CMIP6 climate scenarios into the SWH model; Using the input of future meteorological data and land use data, the SWH model dynamically simulates the evaporation and transpiration of the Pearl River Basin at different times in the future by calculating the evaporation and transpiration process of water in each ecosystem.

8. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 5 comprises: The evapotranspiration data of the Pearl River Basin at different periods output from the SWH model were collated; In the time dimension, the average evapotranspiration, total evapotranspiration and evapotranspiration change rate of each ecosystem in different periods are calculated by statistically simulating evapotranspiration data. In the spatial dimension, the spatial distribution of evapotranspiration of each ecosystem was analyzed and the spatial distribution map of evapotranspiration was drawn using GIS technology to intuitively display the spatial differences and pattern changes of evapotranspiration of various ecosystems in the Pearl River Basin in different periods.

9. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 6 comprises: By establishing a multivariate regression model, the spatiotemporal variation data of evapotranspiration were used as dependent variables, and climate change, land use change and ecosystem function factors were used as independent variables to analyze the contribution rate of various driving factors to the change of evapotranspiration.

10. The method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin according to claim 9, characterized in that: The step 6 also includes: Correlation analysis was used to assess the strength of associations among climate change, land use change, and evapotranspiration at different time periods and spatial scales.

Images