A runoff change attribution analysis method based on underlying surface spatial characteristics

Abstract

The application discloses a runoff change attribution analysis method based on underlying surface spatial characteristics, and belongs to the technical field of runoff attribution analysis. First, in the constructed underlying surface spatial characteristic factor set, the key underlying surface spatial characteristic factors capable of effectively explaining the change of Budyko parameters are screened through collinearity test and stepwise regression. Second, an optimal parameter relationship model between the Budyko parameters and the key underlying surface spatial characteristic factors is established. Finally, combined with a water-heat coupling balance equation, the runoff change in different analysis periods is attributed and decomposed, so that the contribution amount of climate change, the contribution amount of underlying surface spatial characteristic change and the absolute contribution proportion are obtained. Through the relationship between the parameters and the underlying surface spatial characteristics, the underlying surface spatial heterogeneity is allowed to participate in the runoff change attribution analysis in a parameterized manner, so that the expression capability of the attribution result to the comprehensive influence of the underlying surface is improved, and guidance is provided for the basin water resource regulation and ecological protection management.

Description

Technical Field

[0001] This invention belongs to the field of runoff attribution analysis technology, and relates to a runoff change attribution analysis method based on the spatial characteristics of the underlying surface. Background Technology

[0002] In recent years, under the combined effects of climate change and human activities, the underlying surface conditions of watersheds have continued to evolve, and runoff processes have exhibited significant spatiotemporal variability. Accurately identifying the contributions of climate change and underlying surface changes to runoff variation has become a critical technical issue that urgently needs to be addressed in watershed water resource management, ecological protection, and integrated watershed governance.

[0003] Existing studies on runoff variation attribution based on the Budyko hydrothermal equilibrium equation mainly focus on the quantitative analysis of the contributions of climate and underlying surface to runoff variation. For example, Chinese invention patent CN110852572B proposes a method, system, and device for quantitative attribution of runoff variation at arbitrary temporal and spatial scales. It quantitatively assesses runoff variation by constructing attribution equations at different time and spatial scales. However, CN110852572B primarily uses Budyko parameters as predetermined parameters in runoff variability analysis, failing to adequately characterize the underlying surface features and their changing processes when attributing runoff variation. Chinese invention patent CN115344815B proposes a method and system for attributing natural runoff variation considering vegetation spatial changes, using NDVI changes to attribute natural runoff variation. However, CN115344815B mainly uses a single vegetation index to characterize underlying surface changes, while the actual Budyko parameters in a watershed are influenced not only by vegetation but also by topography, soil, and human activities. Existing methods are insufficient in characterizing the spatial features of the underlying surface in the Budyko parameters, making it difficult to fully reflect the impact of the underlying surface spatial structure and its changes on runoff.

[0004] In summary, there is an urgent need for a method to comprehensively characterize the spatial features and changes of the underlying surface and establish its relationship with the Budyko parameter to attribute runoff changes. This would enable a quantitative and spatial analysis of the contribution of climate change and underlying surface changes to runoff, providing a scientific basis for watershed water resources management and ecological protection. Summary of the Invention

[0005] To address the problems of existing technologies, this invention proposes a runoff variation attribution analysis method based on underlying surface spatial characteristics, aiming to solve the problem of insufficient characterization of underlying surface spatial characteristics in existing runoff attribution methods. Under the Budyko equation, this invention introduces multiple underlying surface spatial characteristic factors, including landscape pattern, topography, soil, and human activities, to establish a quantitative relationship between Budyko parameters and underlying surface spatial characteristic factors; that is, it establishes Budyko parameters based on the Budyko equation. This invention establishes functional relationships between underlying surface spatial characteristics such as landscape, topography, soil, and human activities, enabling the spatial structure and changes of the underlying surface to participate in the runoff attribution process through parameters. Based on this, it quantitatively decomposes the contributions of climate change and changes in underlying surface spatial characteristics to runoff variation, thereby improving the ability of runoff change attribution analysis to express the spatial heterogeneity of the underlying surface. This invention can fully consider the spatial characteristics of the underlying surface and accurately identify the contributions of climate change and underlying surface changes to runoff variation.

[0006] To achieve the above objectives, the technical solution provided by the present invention is as follows: A method for attributing runoff variation based on underlying surface spatial characteristics, the method comprising the following steps: Step 1 involves preprocessing the hydrological and meteorological data, underlying surface data, and watershed attribute data of the study basin, and constructing a set of spatial characteristic factors for the underlying surface. This set is then filtered to obtain key spatial characteristic factors of the underlying surface. Specifically: Step 1.1: Collect and process the hydrological and meteorological data, underlying surface data, and watershed attribute data of the study watershed required for runoff change attribution. Unify the hydrological and meteorological data, underlying surface data, and watershed attribute data to the sub-watershed analysis scale, and perform statistics according to the preset analysis period to obtain the hydrological and meteorological data and underlying surface spatial characteristic factor set of each sub-watershed in each analysis period.

[0007] Step 1.2: Based on the set of underlying surface spatial characteristic factors obtained in Step 1.1, the variance inflation factor (VIF) is used to perform a collinearity test on each underlying surface spatial characteristic factor, and underlying surface spatial characteristic factors with strong collinearity (VIF≥5) are removed. Subsequently, a stepwise regression method is used, with the minimum of the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) as the screening criteria, to determine the appropriate Budyko parameters. The optimal combination of explanatory variables is used to obtain the key underlying surface spatial characteristic factors.

[0008] Step 2: Based on the key underlying surface spatial characteristic factors and hydro-meteorological data obtained in Step 1, retrieve the Budyko parameters for each sub-basin and each analysis period. And construct the Budyko parameters A parametric relationship model between the model and key underlying surface spatial characteristic factors was developed, yielding time-series data of Budyko parameters and optimal parametric relationship models for each sub-basin and analysis period. Specifically: Step 2.1: Based on the hydrological and meteorological data of each sub-basin obtained in Step 1.1 for each analysis period, the Budyko parameters of each sub-basin for each analysis period are retrieved using the Budyko equation. The Budyko parameters for each sub-basin and each analysis period were obtained. Time series data: On a multi-year average scale, neglecting changes in watershed storage, the watershed water balance relationship can be expressed as: (1) in, For precipitation, This is the actual evaporation rate. This refers to runoff.

[0009] The Budyko parameters for each sub-basin during each analysis period were inverted using the Budyko hydrothermal coupling equilibrium equation in the Choudhury-Yang form. n : (2) in, Potential evapotranspiration; Budyko parameters characterize the overall features of the underlying surface of the watershed; The drying index is as follows: (3) From equation (1), we can obtain: (4) Based on the analysis period of each sub-basin , and The data is used to solve for the corresponding Budyko parameters. This generates time-series data of Budyko parameters for each sub-basin at different time periods.

[0010] Step 2.2: Invert the Budyko parameters obtained from each analysis period of each sub-basin. Using the key underlying surface spatial characteristic factors obtained in step 1.2 as the dependent variable, the Budyko parameters are established. The candidate parameter relationship model between the key underlying surface spatial feature factors and the Budyko parameters is considered. The potential linear or nonlinear relationships between them are investigated, and the key underlying surface spatial characteristic factors for each candidate are transformed using linear, logarithmic, and exponential forms, respectively. Budyko parameters are then constructed. General relationship model between key underlying surface spatial characteristic factors: (5) in, For the first The Budyko parameters corresponding to each analysis period; These are the key underlying surface spatial characteristic factors after functional transformation; , For the 1st, 2nd, The values ​​of key underlying surface spatial characteristic factors after functional transformation; For constant terms; For the 1st, 2nd, Regression coefficients of key underlying surface spatial characteristic factors.

[0011] Based on the Akaike Information Criterion (AIC) and the coefficient of determination By comparing and filtering candidate parameter relationship models established under different combinations of function forms, the Budyko parameters were finally obtained. The optimal parameter relationship model between the key underlying surface spatial characteristic factors and the underlying surface.

[0012] Step 3: Based on the time-series data of the Budyko parameters and the optimal parameter relationship model obtained in Step 2, and combined with the average annual runoff of each sub-basin within each analysis time window, the annual average runoff variation of the study basin is decomposed attributively to obtain the contribution of climate change, the contribution of spatial characteristic changes of the underlying surface, and their absolute contribution percentages. Specifically: Step 3.1, calculate the average annual runoff variation. From equations (3) and (4), the runoff expression is obtained as follows: (6) For any two analysis time windows and reference time window , No. The average annual runoff variation of each sub-basin is: (7) in, Indicates the sub-basin number, Indicates the analysis time window number. For the first The sub-basin in the The average annual runoff change of each analysis time window relative to the baseline time window; The average annual runoff for the baseline time window, To compare the average annual runoff over time windows.

[0013] Equation (6) shows that runoff variation is influenced by both climate change and the Budyko parameter. The combined effects of changes can be represented as: (8) in, Indicates the first The sub-basin in the The contribution of each analysis time window to climate change relative to the baseline time window; Indicates the first The sub-basin in the The contribution of each analysis time window to the change in the spatial characteristics of the underlying surface relative to the baseline time window.

[0014] Step 3.2: Based on the average annual runoff variation obtained in Step 3.1, perform total differential on Equation (6) to obtain the sensitivity expression of the average annual runoff variation to precipitation, potential evapotranspiration, and watershed characteristic parameters: (9) Accordingly, the runoff change is decomposed into a climate change term and an underlying surface change term, wherein: (10) (11) in, Indicates the first The sub-basin in the The change in precipitation within each analysis time window relative to the baseline time window; This represents the corresponding potential change in evapotranspiration; This indicates the corresponding change in the parameter. Specifically: (12) (13) (14) in: , , The first The sub-basin in the Precipitation, potential evapotranspiration, and Budyko parameters for each analysis time window; , , This is the value corresponding to the baseline time window.

[0015] Step 3.3, based on formulas (10) and (11) from step 3.2, let... This represents the number of analysis time windows excluding the baseline time window. Given the number of sub-basins, the overall average contribution to climate change is... Contribution of overall average spatial characteristics of underlying surface They are represented as follows: (15) (16) In obtaining and Subsequently, the contribution rates of climate change and changes in underlying surface spatial characteristics to total runoff variation were calculated separately. To facilitate comparison of the relative impact of the two driving factors on runoff variation, the absolute contribution percentages were used to characterize the contribution intensity of climate change and changes in underlying surface spatial characteristics, and their expressions are as follows: (17) (18) in, The percentage of absolute contribution to climate change The percentage of absolute contribution to the spatial characteristics of the underlying surface.

[0016] By calculating using formulas (6) to (18), the average annual runoff climate change contribution, underlying surface spatial characteristic change contribution, and corresponding absolute contribution ratio of each sub-basin under different analysis time windows are obtained, thereby realizing quantitative attribution analysis of the average annual runoff change process of the study basin.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention reduces the interference of redundant factors on the spatial feature representation of the underlying surface by screening key underlying surface spatial feature factors, thereby improving the reliability of the spatial feature representation results of the underlying surface; and improves the reliability by constructing Budyko parameters. The parametric relationship model between the underlying surface and key spatial characteristic factors enables the spatial characteristics of the underlying surface to participate in the attribution analysis of mean annual runoff changes in a parametric manner, improving the ability of the attribution results to express the spatial differences of the underlying surface. By decomposing the mean annual runoff changes of different sub-basins and different analysis time windows, the contribution of climate change, the contribution of underlying surface spatial characteristic changes, and their absolute contribution ratio can be quantitatively obtained. This enables quantitative attribution of the mean annual runoff change process in the study basin and allows for more accurate identification of the dominant driving factors and their spatial differences in mean annual runoff changes of different sub-basins. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention.

[0019] Figure 2 This is a graph showing the attribution results of runoff changes under different time windows in the embodiments of the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to specific embodiments.

[0021] This embodiment selects the Yellow River Basin as the research object. The Yellow River Basin covers a vast area with complex natural geographical conditions, and there are significant differences in climate, topography, and underlying surface characteristics between its upper, middle, and lower reaches. In recent decades, under the combined effects of climate change and large-scale human activities, the Yellow River Basin has seen continuous progress in vegetation restoration, land use adjustment, and reservoir construction, resulting in significant changes in the spatial structure of the underlying surface. Therefore, it is suitable as a typical region for conducting runoff attribution analysis under the combined effects of climate change and underlying surface changes.

[0022] Traditional Budyko attribution methods often use fixed parameters or single vegetation or land use area indicators to characterize underlying surface features, which is insufficient for representing spatial characteristics of the underlying surface and hinders the ability of runoff change attribution analysis to express spatial heterogeneity. Therefore, this embodiment adopts a runoff change attribution analysis method based on underlying surface spatial characteristics, and the flowchart of the method is as follows. Figure 1 As shown, taking the Yellow River basin as an example, the specific steps are as follows: Step 1 involves uniformly preprocessing the hydrological and meteorological data, underlying surface data, and watershed attribute data of the study basin, and constructing a set of spatial characteristic factors of the underlying surface. Based on this set of spatial characteristic factors, collinearity tests and stepwise regression screening are performed to obtain key spatial characteristic factors of the underlying surface. Specifically: Step 1.1: Collect and process hydrological and meteorological data, underlying surface data, and watershed attribute data for the study basin. The study period is from 1985 to 2018. In this embodiment, the Yellow River Basin is divided into 29 sub-basins according to the spatial boundaries of the national three-level watershed zoning, and each sub-basin serves as the basic analysis unit. To ensure the consistency of multi-source data, the collected data are compiled and analyzed, typical erroneous data are identified and corrected, and all types of data are uniformly preprocessed and standardized to the sub-basin analysis scale. To avoid over-reliance on single abrupt change points in runoff attribution results and to characterize the continuous evolution of watershed hydrological and meteorological elements and underlying surface spatial structure, a 10-year sliding time window is used to construct multiple analysis periods, obtaining hydrological and meteorological data and underlying surface spatial characteristic factor sets for each sub-basin within each analysis time window.

[0023] Step 1.2: Based on the set of underlying surface spatial characteristic factors obtained in 1.1, the variance inflation factor (VIF) is used for collinearity testing, and underlying surface spatial characteristic factors with VIF ≥ 5 are eliminated. Subsequently, stepwise regression is used to further screen underlying surface spatial characteristic factors with better explanatory power for the Budyko parameter n, and the minimum Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) are used as screening criteria to determine key underlying surface spatial characteristic factors. After screening, nine key underlying surface spatial characteristic factors are finally determined, namely, Landscape Diversity Index (SHDI), Maximum Patch Index (LPI), Patch Number (NP), Precipitation (P), Perimeter-Area Fractal Dimension (PAFRAC), Topographic Moisture Index (TWI), Soil Organic Carbon Content (soc), and Soil Saturated Hydraulic Conductivity (k). s And the reservoir's storage capacity per unit area (res).

[0024] Step 2: Based on the key underlying surface spatial characteristic factors and hydro-meteorological data obtained in Step 1, invert the Budyko parameters for each sub-basin and each analysis period. And construct the Budyko parameters A parametric relationship model between the model and key underlying surface spatial characteristic factors was developed, yielding time-series data of Budyko parameters and optimal parametric relationship models for each sub-basin and analysis period. Specifically: Step 2.1: Based on the hydrological and meteorological data of each sub-basin and each analysis time window obtained in Step 1.1, the Budyko parameters of each sub-basin and each analysis time window are retrieved to obtain the time series data of the Budyko parameters of each sub-basin and each analysis time window. Since this embodiment uses a 10-year sliding time window, the change in watershed storage at this scale can be approximately ignored, therefore the actual evapotranspiration satisfies: (1) in, This is the actual evaporation rate. For precipitation, This refers to runoff.

[0025] The Budyko parameters characterizing the comprehensive underlying surface features of each sub-basin were retrieved using the Budyko equation in the Choudhury-Yang form during each analysis period. Its expression is: (2) in, Potential evapotranspiration; Budyko parameters characterize the overall features of the underlying surface of the watershed; The drying index is as follows: (3) Therefore, the corresponding expression for average annual runoff is: (4) Based on the analysis period of each sub-basin , and The data was used to solve for the Budyko parameters within the corresponding time window. This results in the formation of parameter time series sequences for each sub-basin within different time windows.

[0026] Step 2.2, based on the Budyko parameters obtained in Step 2.1 Time series data, including Budyko parameters Using the key underlying surface spatial characteristic factors obtained in step 1.2 as the dependent variable, the Budyko parameters are established. A candidate parameter relationship model between the key underlying surface spatial characteristic factors and the underlying parameters. This model considers the Budyko parameters. There may be nonlinear response relationships between the variables and different key underlying surface spatial characteristic factors. Functional transformations were performed on each variable using linear, logarithmic, and exponential forms, and candidate parameter relationship models were established under different combinations of these forms. Based on the Akaike Information Criterion (AIC) and the coefficient of determination... By comparing and filtering candidate parameter relationship models established under different combinations of function forms, the Budyko parameters were finally obtained. The optimal parameter relationship model between the key underlying surface spatial characteristic factors is as follows: (19) Step 2.3, based on the optimal parameter relationship model obtained in Step 2.2, to further illustrate the Budyko parameters in this embodiment... To assess the effectiveness of the representation method, a fixed-parameter model and a single NDVI index model were set as control models and compared with the model of this invention. (Regarding Budyko parameters...) Perform fitting and prediction, and further analyze the predicted Budyko parameters. Substitute the values ​​into equation (4) to calculate the average annual runoff and obtain the predicted average annual runoff. A comparison of the accuracy of each model is shown in Table 1.

[0027] Table 1: Effects of models with different parameters

[0028] As shown in Table 1, compared with the fixed parameter model and the single NDVI index model, the optimal parameter relationship model between the Budyko parameters and key underlying surface spatial characteristic factors constructed in this invention is superior in both parameter fitting and average annual runoff simulation performance. This indicates that the Budyko parameters are characterized by key underlying surface spatial characteristic factors. It can more effectively reflect the comprehensive characteristics of the underlying surface of the watershed, improve the accuracy of parameter expression and the reliability of subsequent runoff change attribution analysis.

[0029] Step 3: Based on the time-series data of the Budyko parameters and the optimal parameter relationship model obtained in Step 2, and combined with the average annual runoff of each sub-basin within each analysis time window, the attribution decomposition of the average annual runoff change in the study basin is performed to obtain the contribution of climate change, the contribution of underlying surface spatial characteristic changes, and their absolute contribution percentages. Specifically: Step 3.1: Based on the Budyko parameter time-series data and optimal parameter relationship model obtained in Step 2, calculate the average annual runoff change for each sub-basin within each analysis time window relative to the baseline time window, thus obtaining the average annual runoff change. In this embodiment, the first analysis time window, 1985–1994, is used as the baseline time window, with 1990 as its central year. For subsequent analysis time windows in each sub-basin, calculate the average annual runoff change relative to the baseline time window. Then, the... The sub-basin in the The average annual runoff change for each analysis time window relative to the baseline time window is expressed as follows: (7) Among them, among them, , , For the first The sub-basin in the The average annual runoff change of each analysis time window relative to the baseline time window; The average annual runoff for the baseline time window, To compare the average annual runoff over time windows.

[0030] Equation (6) shows that runoff variation is influenced by both climate change and the Budyko parameter. The combined effects of changes can be represented as: (8) in, Indicates the first The sub-basin in the The contribution of each analysis time window to climate change relative to the baseline time window; Indicates the first The sub-basin in the The contribution of each analysis time window to the change in the spatial characteristics of the underlying surface relative to the baseline time window.

[0031] Step 3.2: Based on the average annual runoff variation obtained in Step 3.1, perform total differential on Equation (6) to obtain the sensitivity expression of the average annual runoff variation to precipitation, potential evapotranspiration, and watershed characteristic parameters: (9) Accordingly, the runoff change is decomposed into a climate change term and an underlying surface change term, wherein: (10) (11) in, Indicates the first The sub-basin in the The change in precipitation within each analysis time window relative to the baseline time window; This represents the corresponding potential change in evapotranspiration; This indicates the corresponding change in the parameter. Specifically: (12) (13) (14) in: , , The first The sub-basin in the Precipitation, potential evapotranspiration, and Budyko parameters for each analysis time window; , , This is the value corresponding to the baseline time window.

[0032] Step 3.3, based on equations (10) and (11) in step 3.2, let... This represents the number of analysis time windows excluding the baseline time window. Given the number of sub-basins, the overall average contribution to climate change is... Contribution of overall average spatial characteristics of underlying surface They are represented as follows: (20) (twenty one) In obtaining and Subsequently, the contribution rates of climate change and changes in underlying surface spatial characteristics to total runoff variation were calculated separately. To facilitate comparison of the relative impact of the two driving factors on runoff variation, the absolute contribution percentages were used to characterize the contribution intensity of climate change and changes in underlying surface spatial characteristics, and their expressions are as follows: (17) (18) in, The percentage of absolute contribution to climate change The percentage of absolute contribution to the spatial characteristics of the underlying surface.

[0033] Then, for each sub-basin and each analysis time window, the total runoff change, the contribution of climate change, and the contribution of underlying surface spatial characteristic change were calculated, and further statistical analysis was performed to form the overall attribution results. To characterize the average change characteristics relative to the baseline period during the study period, the attribution results of each subsequent analysis time window for each sub-basin relative to the baseline period were averaged, and the results are shown in Table 2.

[0034] Table 2: Overall Attribution Results

[0035] As shown in Table 2, in this embodiment, relative to the baseline time window, the overall average change in total runoff during the study period was 4.29 mm, of which climate change contributed -2.86 mm and changes in underlying surface spatial characteristics contributed 7.16 mm. The absolute contributions of climate change and changes in underlying surface spatial characteristics were 28.57% and 71.43%, respectively, indicating that, on an overall average basis, changes in underlying surface spatial characteristics had a more significant impact on runoff evolution, while climate change had a suppressive effect on total runoff change.

[0036] Figure 2 This is a graph showing the attribution results of runoff changes under different time windows in the embodiments of the present invention. Figure 2 It is evident that the contributions of climate change and the contribution of changes in the spatial characteristics of the underlying surface fluctuate significantly under different analytical time windows, indicating that the dominant driving factors of the average annual runoff change differ at different stages during the study period, which can provide a scientific basis for watershed water resources management and ecological regulation.

[0037] The technical solution of the present invention has been described in detail above with reference to specific embodiments, but the scope of protection of the present invention is not limited to the above embodiments. For those skilled in the art, any equivalent substitutions, modifications, improvements or combinations made to the present invention without departing from the technical concept and essence of the present invention should be considered as falling within the scope of protection of the present invention.

Claims

1. A method for attributing runoff variation based on the spatial characteristics of the underlying surface, characterized in that, The runoff change attribution analysis method includes the following steps: Step 1: Preprocess the hydrological and meteorological data, underlying surface data, and watershed attribute data of the study watershed, and construct a set of underlying surface spatial characteristic factors; screen the underlying surface spatial characteristic factor set to obtain key underlying surface spatial characteristic factors. Step 2: Based on key underlying surface spatial characteristic factors and hydro-meteorological data, invert the Budyko parameters for each sub-basin and each analysis period. And construct the Budyko parameters The parameter relationship model between the key underlying surface spatial characteristic factors was used to obtain the time series data of Budyko parameters and the optimal parameter relationship model for each sub-basin and each analysis period. Step 3: Based on the time series data of the Budyko parameters and the optimal parameter relationship model, and combined with the average annual runoff of each sub-basin in each analysis time window, the annual average runoff change of the study basin is decomposed by attribution to obtain the contribution of climate change, the contribution of the underlying surface spatial characteristics change and their absolute contribution ratio, so as to realize the quantitative attribution analysis of the average annual runoff change process of the study basin.

2. The runoff variation attribution analysis method based on underlying surface spatial characteristics according to claim 1, characterized in that, Specifically, step 1 is as follows: Step 1.1: Collect and process the hydrological and meteorological data, underlying surface data, and watershed attribute data of the study watershed required for runoff change attribution. Unify the hydrological and meteorological data, underlying surface data, and watershed attribute data to the sub-watershed analysis scale, and perform statistics according to the preset analysis period to obtain the hydrological and meteorological data and underlying surface spatial characteristic factor set of each sub-watershed in each analysis period. Step 1.2: Based on the set of underlying surface spatial characteristic factors obtained in Step 1.1, the variance inflation factor (VIF) is used to perform collinearity test on each underlying surface spatial characteristic factor, and underlying surface spatial characteristic factors with strong collinearity are removed. Subsequently, a stepwise regression method was used, with the minimum values ​​of the Akaike information criterion and the Bayesian information criterion as the screening criteria, to determine the appropriate parameters for the Budyko parameters. The optimal combination of explanatory variables is used to obtain the key underlying surface spatial characteristic factors.

3. The runoff variation attribution analysis method based on underlying surface spatial characteristics according to claim 2, characterized in that, In step 1.2, the variance expansion factor (VIF) ≥ 5 is a spatial characteristic factor of the underlying surface with strong collinearity.

4. The runoff variation attribution analysis method based on underlying surface spatial characteristics according to claim 3, characterized in that, Step 2 specifically includes: Step 2.1: Based on the hydrological and meteorological data of each sub-basin obtained in Step 1.1 for each analysis period, the Budyko parameters of each sub-basin for each analysis period are retrieved using the Budyko equation. The Budyko parameters for each sub-basin and each analysis period were obtained. Time series data; On a multi-year average scale, neglecting changes in watershed storage, the watershed water balance relationship can be expressed as: (1) in, For precipitation, This is the actual evaporation rate. This refers to runoff; The Budyko parameters for each sub-basin during each analysis period were retrieved using the Budyko hydrothermal coupling equilibrium equation. n : (2) in, Potential evapotranspiration; Budyko parameters characterize the overall features of the underlying surface of the watershed; The drying index is as follows: (3) From equation (1), we can obtain: (4) Based on the analysis period of each sub-basin , and The data is used to solve for the corresponding Budyko parameters. This generates time-series data of Budyko parameters for each sub-basin at different time periods; Step 2.2: Invert the Budyko parameters obtained from each analysis period of each sub-basin. Using the key underlying surface spatial characteristic factors obtained in step 1.2 as the dependent variable, the Budyko parameters are established. The candidate parameter relationship model between the key underlying surface spatial feature factors and the Budyko parameters is considered. The possible linear or nonlinear relationships between them were investigated, and the key underlying surface spatial characteristic factors of each candidate were transformed using linear, logarithmic, and exponential forms, respectively; the Budyko parameters were constructed. General relationship model between key underlying surface spatial characteristic factors: (5) in, For the first The Budyko parameters corresponding to each analysis period; These are the key underlying surface spatial characteristic factors after functional transformation; , For the 1st, 2nd, The values ​​of key underlying surface spatial characteristic factors after functional transformation; For constant terms; For the 1st, 2nd, Regression coefficients of key underlying surface spatial characteristic factors; Based on the Akaike Information Criterion (AIC) and the coefficient of determination By comparing and filtering candidate parameter relationship models established under different combinations of function forms, the Budyko parameters were finally obtained. The optimal parameter relationship model between the key underlying surface spatial characteristic factors and the underlying surface.

5. The runoff variation attribution analysis method based on the spatial characteristics of the underlying surface according to claim 4, characterized in that, Step 3 specifically includes: Step 3.1, calculate the average annual runoff variation. From equations (3) and (4), the runoff expression is obtained as follows: (6) For any two analysis time windows and reference time window , No. The average annual runoff variation of each sub-basin is: (7) in, Indicates the sub-basin number, Indicates the analysis time window number. For the first The sub-basin in the The average annual runoff change of each analysis time window relative to the baseline time window; The average annual runoff for the baseline time window, To compare the average annual runoff over time windows; Equation (6) shows that runoff variation is influenced by both climate change and the Budyko parameter. The combined effects of changes can be represented as: (8) in, Indicates the first The sub-basin in the The contribution of each analysis time window to climate change relative to the baseline time window; Indicates the first The sub-basin in the The contribution of each analysis time window to the change in the spatial characteristics of the underlying surface relative to the baseline time window; Step 3.2: Based on the average annual runoff variation obtained in Step 3.1, perform total differential on Equation (6) to obtain the sensitivity expression of the average annual runoff variation to precipitation, potential evapotranspiration, and watershed characteristic parameters: (9) Accordingly, the runoff change is decomposed into a climate change term and an underlying surface change term, wherein: (10) (11) in, Indicates the first The sub-basin in the The amount of precipitation change in each analysis time window relative to the baseline time window; This represents the corresponding potential change in evapotranspiration; This indicates the corresponding parameter change; specifically: (12) (13) (14) in: , , The first The sub-basin in the Precipitation, potential evapotranspiration, and Budyko parameters for each analysis time window; , , The value corresponding to the reference time window; Step 3.3, based on formulas (10) and (11) from step 3.2, let... This represents the number of analysis time windows excluding the baseline time window. Given the number of sub-basins, the overall average contribution to climate change is... Contribution of overall average spatial characteristics of underlying surface They are represented as follows: (15) (16) In obtaining and Then, the contribution rates of climate change and changes in underlying surface spatial characteristics to total runoff change were calculated separately. To facilitate comparison of the relative impact of the two types of driving factors on runoff change, the absolute contribution percentages were used to characterize the contribution intensity of climate change and changes in underlying surface spatial characteristics, and their expressions are as follows: (17) (18) in, The percentage of absolute contribution to climate change The percentage of absolute contribution to the spatial characteristics of the underlying surface; By calculating using formulas (6) to (18), the average annual runoff climate change contribution, underlying surface spatial characteristic change contribution, and corresponding absolute contribution ratio of each sub-basin under different analysis time windows are obtained, thus achieving quantitative attribution analysis.

Images