A method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin

Abstract

The present invention belongs to the field of evapotranspiration prediction and provides a method for constructing evapotranspiration simulation models for different ecosystems in the Pearl River Basin. The method comprises the following steps: collecting meteorological data, land use data and hydrological data of the Pearl River Basin; simulating past and present evapotranspiration of the Pearl River Basin using a SWH model; predicting future land use changes in the Pearl River Basin using a CA-Markov model and a PLUS model; simulating future evapotranspiration of different ecosystems in the Pearl River Basin using the SWH model based on the predicted future land use data in combination with a GFDL-ESM2M model and a CMIP6 climate scenario; statistically simulating evapotranspiration data of different ecosystems in different periods; and analyzing driving mechanisms of evapotranspiration of different ecosystems in the Pearl River Basin.

Description

Technical Field

[0001] The present 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 a crucial component of the hydrological cycle and climate regulation, reflecting the evaporation of water and transpiration from plants within ecosystems. The Pearl River Basin, a typical complex ecosystem in southern China, encompasses diverse ecosystems such as forests, grasslands, farmlands, and wetlands. Its evapotranspiration is influenced by multiple factors, including climate change, land use change, and ecosystem function. Due to climate change and intensified human activities, evapotranspiration within the Pearl River Basin is undergoing significant temporal and spatial variations, directly impacting regional water resource distribution, ecological balance, and land use decisions. Therefore, establishing an accurate and dynamic evapotranspiration simulation model capable of comprehensively analyzing evapotranspiration changes across diverse ecosystems is of great significance for water resource management, climate change response, and ecological protection.

[0003] Existing simulations of watershed evapotranspiration often use single models or focus solely on climate factors, neglecting the impacts of land-use change and ecosystem complexity on evapotranspiration. This simplistic approach fails to accurately reflect the dynamics of complex ecosystems. Therefore, developing a simulation method that integrates climate change, land-use change, and ecosystem function for multi-dimensional evapotranspiration prediction and analysis has significant research and application value. 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 for the Pearl River Basin. The meteorological data includes temperature, precipitation, humidity, and solar radiation; the land use data includes remote sensing images and land cover classification data; and the hydrological data includes flow and water level data at hydrological stations within the basin.

[0006] 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;

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

[0008] Step 4: Simulating 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 future evapotranspiration of different ecosystems in the Pearl River Basin;

[0009] Step 5: Spatiotemporal evolution analysis: By using statistically simulated evapotranspiration data from different periods and ecosystems, we analyze the spatiotemporal evolution of evapotranspiration across forests, grasslands, and farmland ecosystems in the Pearl River Basin.

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

[0011] This paper provides a method for constructing a simulation model for evapotranspiration in different ecosystems in the Pearl River Basin. This method combines meteorological, land use, and hydrological data, using the SWH model, CA-Markov model, PLUS model, GFDL-ESM2M model, and CMIP6 climate scenarios to comprehensively simulate and predict past, present, and future evapotranspiration changes in the Pearl River Basin. Through this 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. It 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 existing technology.

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

[0014] (3) By combining the temporal characteristics of the CA-Markov model with the spatial complexity processing capabilities of the PLUS model, this paper ensures the accuracy of temporal trend prediction and spatial distribution simulation of land use change. Furthermore, 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, this invention overcomes the limitations of existing technologies by combining multiple models, achieving a comprehensive, accurate, and dynamic simulation of evapotranspiration in different ecosystems in the Pearl River Basin, and can provide strong support for ecological protection, climate change response, and water resource 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 following briefly introduces the drawings required for use in the embodiments. 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 any creative work.

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

[0020] Below, the invention is preferably described with reference to the accompanying drawings and specific embodiments.

[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 existing 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. Evapotranspiration refers to the process by which water in an ecosystem is transpired through plant transpiration and soil evaporation. It reflects an important indicator of ecosystem water consumption and energy balance, and affects the regional hydrological cycle and climate change. Based on existing data, methods, and models, the present invention establishes a mathematical model or computer model through a systematic modeling process that can predict and simulate changes in evapotranspiration in different ecosystems in the Pearl River Basin 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 includes temperature, precipitation, humidity and solar radiation, the land use data includes remote sensing images and land cover classification data, and the hydrological data includes flow and water level data of hydrological stations in the basin.

[0024] This step involves collecting basic data on 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, specifically temperature, precipitation, humidity, solar radiation, and other data. The temperature data refers to the spatiotemporal distribution of surface temperature within the basin, typically obtained through meteorological station observations or reanalysis data. The precipitation data refers to annual, monthly, or daily precipitation within the basin, parameters closely related to the water cycle, and can reflect the spatiotemporal variation of precipitation in the Pearl River Basin. The humidity data refers to the water vapor content in the air, including relative and absolute humidity data. The solar radiation data refers to the amount of solar energy received by the Pearl River Basin, reflecting the impact of the climate system on evaporation, and is typically estimated through radiation observations or satellite data.

[0026] Land use data: This data includes information on land cover within the Pearl River Basin, specifically remote sensing imagery data and land cover classification data. Remote sensing imagery data refers to imagery acquired through remote sensing satellites such as Landsat and MODIS, which can reflect the surface characteristics of the Pearl River Basin at different times. Land cover classification data refers to classification data based on remote sensing imagery or field surveys, which categorizes land use types into forest, grassland, farmland, urban construction land, and other categories according to a specific classification system. This data is used for subsequent land use change and evapotranspiration analysis.

[0027] Hydrological data: This includes hydrological observation data for the major river systems within the Pearl River Basin, specifically including flow and water level data measured by hydrological stations established within the basin. Flow data refers to the observed flow of major rivers within the basin over different time periods and is used to analyze water resource distribution and changes. Water level data refers to the water level of rivers, lakes, reservoirs, and other water bodies, reflecting the changing trends in hydrological conditions within 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 in 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 uses the SWH model (Soil-Water-Heat coupling model) to simulate the past and present evapotranspiration of the Pearl River Basin based on the meteorological data, land use data, and hydrological data. The specific implementation method is as follows:

[0031] The SWH model, or coupled soil-water-heat model, is a dynamic simulation that integrates soil water balance, vegetation physiological characteristics, and atmospheric energy exchange processes. By describing the interactions between soil, plants, and the atmosphere at multiple levels, the model accurately simulates evapotranspiration. 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). These inputs work together to simulate the temporal and spatial distribution of evapotranspiration.

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

[0033] Specific methods used by the SWH model to simulate past and present evapotranspiration in the Pearl River Basin:

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

[0035] Meteorological data: historical and current data such as temperature, precipitation, humidity, and solar radiation, mainly extracted from weather stations or reanalysis datasets, are 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, and farmland) 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 of the SWH model were adapted and calibrated according to the actual conditions of the Pearl River Basin. The following parameters were mainly involved:

[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., which 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, land use, and hydrological data were input into the SWH model to simulate evapotranspiration over the Pearl River Basin over different historical periods. Through step-by-step iterative calculations over multiple time periods, the basin's past evapotranspiration changes were determined.

[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 of the Pearl River Basin, providing a scientific basis for assessing the basin's water resources dynamics, climate change impacts, and the ecological effects of land use.

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

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

[0048] The CA-Markov model, or cellular automaton-Markov chain model, is a land use prediction model that combines the temporal nature of Markov chains with the spatial dynamics of cellular automata. The model first calculates temporal transition probabilities of land use categories using a Markov chain. It then uses cellular automata to dynamically simulate the spatial distribution of these probabilities, generating spatial patterns of land use change. While the CA-Markov model excels at describing overall land use transition trends and spatial patterns, it has limitations when dealing with complex ecosystem structures.

[0049] The PLUS model (Patch-generating Land Use Simulation) is a patch-generating land use simulation model specifically designed to handle complex land use change. By generating spatial patches and simulating ecosystem dynamics, the model can better capture the spatial heterogeneity and complex structural changes in land use. The PLUS model incorporates multiple driving factors, such as socioeconomic factors, environmental constraints, and policy changes, making it suitable for predicting large-scale land use change.

[0050] The GFDL-ESM2M model, or Geophysical Fluid Dynamics Laboratory Earth System Model, is a global climate model used to simulate future climate change. It provides data on future climate change scenarios 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 the Coupled Model Intercomparison Project Phase 6, is an international climate change research program that aims to predict future climate using a variety of climate models and provide different climate scenario data, such as RCPs (Representative Concentration Pathways) and SSPs (Shared Socioeconomic Pathways), for assessing the impact of future climate on the environment and society.

[0052] Specific implementation steps:

[0053] Data input and preprocessing:

[0054] First, the GFDL-ESM2M model and future climate scenario data provided by the CMIP6 project were preprocessed. These climate scenario data include meteorological elements such as future temperature and precipitation in the Pearl River Basin. These meteorological data are 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 socioeconomic 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. 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] Performing matrix reduction generation processing on the conversion probability using a preset matrix function to obtain a probability matrix;

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

[0062] Subsequently, cellular automata were used in combination with the transformation matrix 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] PLUS model of space simulation:

[0065] After obtaining a preliminary prediction of land-use change patterns using the CA-Markov model, the PLUS model was used to simulate and refine spatial patches. The PLUS model can capture more complex land-use change characteristics, including the patchy effects of ecosystems and the spatial heterogeneity of land-use types.

[0066] The PLUS model combines multidimensional driving factors, such as policy influences, socioeconomic 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 handling 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 depict 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 with the climate scenarios provided by CMIP6, and using a joint simulation of the CA-Markov model and the PLUS model, a map of land use change in the Pearl River Basin at different time periods in the future was generated. This result can provide data support for subsequent ecosystem evapotranspiration simulations 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 the CA-Markov model and the PLUS model not only improves the temporal accuracy of the prediction, but also enhances the precision of the spatial simulation, providing comprehensive prediction results for future land use changes in the Pearl River Basin.

[0075] Step 4: Simulate 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 future evapotranspiration of different ecosystems in the Pearl River Basin.

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

[0077] Specific implementation steps for future evapotranspiration simulation:

[0078] Model input preparation:

[0079] First, the predicted future land use data (results of future land use change 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, future meteorological data (such as future temperature, precipitation, and humidity) provided by the GFDL-ESM2M model and CMIP6 climate scenarios were imported into the SWH model. These meteorological data were used to simulate the water cycle and energy transfer processes of the ecosystem under different climate scenarios.

[0081] Parameter adjustment and calibration of SWH model:

[0082] Based on the actual conditions of the Pearl River Basin, key parameters in the SWH model were adjusted and calibrated to ensure that the model can adapt to the evapotranspiration processes of different future ecosystems. These parameters include soil type, vegetation cover, 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 infiltration rate) affect the flow and storage of water in the ecosystem.

[0084] Evapotranspiration simulation process:

[0085] Using input from future meteorological and land use data, the SWH model dynamically simulates evapotranspiration in the Pearl River Basin over different future periods by calculating the evaporation and transpiration processes within each ecosystem. The SWH model can gradually simulate changes in evapotranspiration for specific years in the future (e.g., 2025 and 2030), outputting the temporal and spatial distribution of evapotranspiration for each ecosystem.

[0086] Combining future land use data with future climate scenario data is crucial because land use and climate conditions jointly determine evapotranspiration. Land use change influences the spatial pattern of different ecosystems within a watershed. For example, increases or decreases in forest cover directly affect evapotranspiration, while climate change modulates evapotranspiration through meteorological factors such as temperature and precipitation. Therefore, relying solely on a single data source is insufficient to fully reflect future changes in evapotranspiration.

[0087] The GFDL-ESM2M model and CMIP6 climate scenarios provide multiple 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 scientific nature 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 consistent with the actual scenario.

[0089] Simulation result 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 time 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 assessed.

[0092] By combining these steps with future land use data, the GFDL-ESM2M model, and CMIP6 climate scenarios, the SWH model can dynamically simulate future evapotranspiration changes in different ecosystems within the Pearl River Basin. This combined approach provides more accurate evapotranspiration predictions, providing data support for future water resources management, ecological protection, and land use planning.

[0093] Step 5: Spatiotemporal evolution analysis: By using statistically simulated evapotranspiration data from different periods and 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 in different ecosystems such as forests, grasslands, and farmlands in the Pearl River Basin by statistically simulating evapotranspiration data for different periods and ecosystems.

[0095] Spatiotemporal dynamics refers to the temporal changes and spatial distribution of evapotranspiration across different ecosystems within a specific time and space. By comprehensively analyzing both temporal dimensions (e.g., past, present, and future) and spatial dimensions (e.g., across different regions and ecosystem types), we can reveal patterns of evapotranspiration change. This analysis of spatiotemporal dynamics helps us understand trends and potential drivers of ecosystem hydrological cycles.

[0096] Evapotranspiration data for the Pearl River Basin at different time periods (e.g., 2025 and 2030) generated from the SWH model were collated. This data included evapotranspiration from various ecosystems within the Pearl River Basin, including forests, grasslands, and farmland. Evapotranspiration for each ecosystem was categorized and recorded by time and spatial location, enabling detailed analysis of evapotranspiration characteristics for each ecosystem over time and across regions.

[0097] In the temporal dimension, the simulated evapotranspiration data are statistically analyzed to calculate the average evapotranspiration, total evapotranspiration, and rate of change of each ecosystem over time. Specific calculations include interannual variation, seasonal variation, and long-term trend analysis. These statistics can intuitively reveal the temporal evolution of evapotranspiration across different ecosystems in the Pearl River Basin.

[0098] In the spatial dimension, by analyzing the spatial distribution of evapotranspiration across ecosystems and mapping it using GIS technology, the study visually demonstrates the spatial differences and pattern changes in evapotranspiration across ecosystems in the Pearl River Basin over time. For example, the study examines whether there is a trend of increasing evapotranspiration in forested areas, or whether grassland and farmland exhibit different evapotranspiration patterns.

[0099] Step 6: Driving mechanism analysis: 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 evapotranspiration statistics of 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 change, agricultural land change, 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, and 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. The steps of analyzing the contribution rate of each driving factor to the change of evapotranspiration are as follows: 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 climate factor variance, land factor variance and ecological factor variance, and obtaining the residual coefficient corresponding to the multiple regression model. Combining the climate factor variance, the land factor variance, the ecological factor variance and the residual coefficient, the contribution rate of each driving factor to the change of evapotranspiration is analyzed 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 clearing the erroneous data in the spatiotemporal change data, the standard change data is the data obtained after eliminating the differences between the target spatiotemporal change data, and the residual coefficient is the error corresponding to the multiple 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 changes in temperature and precipitation; land use change can reflect its impact on evapotranspiration through changes in the proportion of different land types; and ecosystem function can reflect its regulatory effect on evapotranspiration through parameters such as vegetation coverage and soil moisture.

[0111] Spatiotemporal 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 across different time periods and spatial scales. Spatiotemporal correlation analysis can reveal the primary driving mechanisms of evapotranspiration in different ecosystems under different climate scenarios. For example, climate may play a more significant role in driving evapotranspiration in alpine forests than in urbanized areas.

[0113] In detail, the correlation analysis can be used to evaluate the correlation strength between climate change, land use change and evapotranspiration in different time periods and spatial ranges by 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 to 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 to calculate the climate-evaporation correlation coefficient between the climate change and evapotranspiration by 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 associations among 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, vector processing for evaluating climate change, land use change, and evapotranspiration can be achieved through vector models, such as the word2vec model. The correlation between climate change, land use change, and evapotranspiration can be analyzed by combining the values ​​of the climate-evaporation correlation coefficient and the land-evaporation correlation coefficient. For example, the higher the correlation coefficient, the higher the correlation. The first correlation and the second correlation can be combined to evaluate the strength of the association between climate change, land use change, and evapotranspiration. For example, the higher the correlation, the higher the strength of the association.

[0122] Climate change, land use change, and ecosystem function have different pathways and impacts on evapotranspiration. Integrating these factors through a combination of models (such as the SWH model, the CA-Markov model, and the PLUS model) ensures a comprehensive representation of the driving mechanisms of evapotranspiration across 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 on the evapotranspiration process; SWH models can simulate the dynamic circulation of water within ecosystems and reflect the contribution of different ecosystem functions to evapotranspiration.

[0124] Combining models can capture the interactions of multiple driving factors. For example, climate change may affect soil moisture through changes in precipitation, which in turn alters 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 modeling, the authors report on the main driving mechanisms of evapotranspiration in different ecosystems in the Pearl River Basin. The results reveal which factors, such as rising temperatures, loss of forest cover, or urban expansion, primarily drive evapotranspiration changes in each ecosystem over time.

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

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

[0129] The prior art mentioned in the aforementioned background technology section and specific embodiment section of the present invention may be regarded as a part of the present invention and 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 for the Pearl River Basin. The meteorological data includes temperature, precipitation, humidity, and solar radiation; the land use data includes remote sensing images and land cover classification data; and the hydrological data includes flow and water level data at hydrological stations within the basin. 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; Step 3: Future land use prediction: The CA-Markov model and the PLUS model are used to predict future land use changes in the Pearl River Basin. The prediction process is based on the GFDL-ESM2M model and future climate scenario data provided by CMIP6. Step 4: Simulating 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 future evapotranspiration of different ecosystems in the Pearl River Basin; Step 5: Spatiotemporal evolution analysis: By using statistically simulated evapotranspiration data from different periods and ecosystems, we analyze the spatiotemporal evolution of evapotranspiration across forests, grasslands, and farmland ecosystems in the Pearl River Basin. Step 6: Driving mechanism analysis: Based on the spatiotemporal evolution of evapotranspiration data, combined with climate change, land use change, and ecosystem function factors, the driving mechanisms of evapotranspiration in different ecosystems in the Pearl River Basin were analyzed; The step 3 includes: Preprocess the future climate scenario data provided by the GFDL-ESM2M model and the CMIP6 project; Calculate the conversion probability between various land use types in the Pearl River Basin through Markov chain, query the land area of ​​various land use types in the Pearl River Basin at different times, and calculate the conversion probability between various land use types in the Pearl River Basin through Markov chain based on the land area; Performing matrix generation processing on the conversion probability using a preset matrix function to obtain a probability matrix; generating, based on the probability matrix, a transition matrix for different land use categories at a preset future time point; 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. 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 input 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 processes of water in various ecosystems.

2. The method for constructing an evapotranspiration simulation model 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 an evapotranspiration simulation model for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The simulation of 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; 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 over 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 an evapotranspiration simulation model for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 3 further comprises: 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.

5. The method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin according to claim 4, characterized in that: The step 3 further comprises: Combining the GFDL-ESM2M model and the climate scenarios provided by CMIP6, a land use change map of the Pearl River Basin at different time periods in the future was generated through joint simulation of the CA-Markov model and the PLUS model.

6. The method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 5 comprises: The evapotranspiration data for the Pearl River Basin at different times, output from the SWH model, were collated. In the temporal dimension, the average evapotranspiration, total evapotranspiration, and evapotranspiration change rate of each ecosystem at different times were calculated by statistically analyzing the simulated evapotranspiration data. In the spatial dimension, by analyzing the spatial distribution of evapotranspiration of each ecosystem and using GIS technology to draw a spatial distribution map of evapotranspiration, the spatial differences and pattern changes of evapotranspiration of various ecosystems in the Pearl River Basin in different periods are intuitively displayed.

7. The method for constructing an evapotranspiration simulation model for different ecosystems in the Pearl River Basin according to claim 1, characterized in that: The step 6 comprises: By establishing a multiple regression model, the spatiotemporal evolution 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.

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

Images