Satellite-ground rainfall data fusion method and device, electronic equipment and medium

Abstract

The invention belongs to the technical field of hydrology and meteorology, and particularly discloses a satellite-ground precipitation data fusion method and device, electronic equipment and a medium. The method comprises the following steps: acquiring first rainfall data of a first spatial resolution inversed by satellite remote sensing in a target area, and performing spatial interpolation on the first rainfall data of the first spatial resolution to obtain first rainfall data of a second spatial resolution; performing spatial resolution resampling on the rainfall observation data of each ground station in the target area and the plurality of topographic factor data to obtain second rainfall data and a plurality of topographic factor data under the first spatial resolution and the second spatial resolution; and performing geographically weighted regression analysis based on the first rainfall data and the second rainfall data under the first spatial resolution and the second spatial resolution and the multiple pieces of topographic factor data, and determining fused rainfall data under the second spatial resolution in the target area. According to the invention, the spatial resolution and precision of rainfall data can be effectively improved.

Description

Technical Field

[0001] The present application belongs to the field of hydrology and meteorology technology, and more specifically, to a satellite-ground precipitation data fusion method, device, electronic equipment and medium. Background Art

[0002] Precipitation, as an important part of the water cycle, is one of the important input parameters of the hydrological model. Its quality directly affects the accuracy of hydrological simulation. Therefore, how to obtain high-quality and high-precision precipitation data has become one of the important topics in today's hydrological research.

[0003] At present, precipitation data are mainly obtained through ground station observations, weather radar measurements, and satellite remote sensing inversion. Among them, ground station observations are the traditional way to obtain precipitation data, but due to the sparse number of stations and uneven spatial distribution, the uncertainty of precipitation data estimated by the station network has been increased to a certain extent. Precipitation products based on weather radar are easily interfered by external factors such as electronic signals, and it is difficult to ensure the accuracy of precipitation data. As an indirect observation product, satellite-inverted precipitation data has good spatiotemporal continuity and a wide observation range, which can make up for the shortcomings of ground station observations to a certain extent. However, due to the influence of climate, topography and other factors, satellite precipitation also has problems of low spatial resolution and poor accuracy, which makes it difficult to meet the needs of refined hydrological modeling. Summary of the invention

[0004] In view of the defects of the prior art, the present application aims to solve the problem that the existing precipitation data has low spatial resolution and low accuracy.

[0005] To achieve the above objectives, in a first aspect, the present application provides a method for fusing satellite-ground precipitation data, comprising: Acquire first precipitation data of a first spatial resolution inverted by satellite remote sensing in a target area, and perform spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; the second spatial resolution is higher than the first spatial resolution; Resampling the precipitation observation data and multiple terrain factor data at each ground station in the target area at spatial resolution to obtain second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution; A geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine fused precipitation data at the second spatial resolution in the target area.

[0006] Optionally, the performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and the plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area includes: Using a geographically weighted regression model, a regression analysis is performed on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine regression residual data and precipitation estimation value data of the geographically weighted regression model at the second spatial resolution; The precipitation estimation value data and the regression residual data at the second spatial resolution are summed to determine fused precipitation data at the second spatial resolution in the target area.

[0007] Optionally, the using a geographically weighted regression model to perform regression analysis on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine regression residual data and precipitation estimation value data of the geographically weighted regression model at the second spatial resolution includes: Using a geographically weighted regression model to perform regression analysis on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution, and determine the regression coefficient and regression residual data of the geographically weighted regression model at the first spatial resolution; Performing spatial interpolation calculation on the regression coefficient and regression residual data of the first spatial resolution to obtain the regression coefficient and regression residual data of the geographically weighted regression model at the second spatial resolution; Substitute the first precipitation data and multiple terrain factor data at the second spatial resolution, and the regression coefficient of the second spatial resolution into the geographically weighted regression model to determine the precipitation estimation value data of the geographically weighted regression model at the second spatial resolution.

[0008] Optionally, after performing spatial interpolation on the first precipitation data of the first spatial resolution to obtain the first precipitation data of the second spatial resolution, the method further includes: Performing principal component analysis on the multiple terrain factors corresponding to the multiple terrain factor data to determine multiple terrain principal component data; Resampling the precipitation observation data of each ground station in the target area and the plurality of terrain principal component data at spatial resolutions to obtain second precipitation data and a plurality of terrain principal component data at the first spatial resolution and the second spatial resolution; A geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain principal component data at the first spatial resolution and the second spatial resolution to determine fused precipitation data at the second spatial resolution in the target area.

[0009] Optionally, the multiple terrain factors include elevation, slope and aspect, and also include at least one of terrain relief and mountain shadow.

[0010] Optionally, the fused precipitation data is fused monthly precipitation data; correspondingly, after performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: Determine the ratio data of the daily precipitation data of each of the ground stations to its monthly precipitation data; Using the inverse distance weighted method to perform spatial interpolation on each of the ratio data to obtain ratio data corresponding to the target area; Determining fused daily precipitation data at a second spatial resolution within the target area based on the ratio data corresponding to the target area and the fused monthly precipitation data at a second spatial resolution within the target area; The fused monthly precipitation data at the second spatial resolution in the target area are accumulated to determine the fused annual precipitation data at the second spatial resolution in the target area.

[0011] Optionally, after performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and the plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: The fused precipitation data at the second spatial resolution is used to perform hydrological runoff simulation analysis to determine the runoff information of the target area at different time scales.

[0012] In a second aspect, the present application provides a satellite-ground precipitation data fusion device, comprising: An interpolation module, used for acquiring first precipitation data of a first spatial resolution inverted by satellite remote sensing in a target area, and performing spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; the second spatial resolution is higher than the first spatial resolution; A resampling module, used for performing spatial resolution resampling on the precipitation observation data and multiple terrain factor data of each ground station in the target area to obtain second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution; A fusion module is used to perform geographically weighted regression analysis based on the first precipitation data, the second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area.

[0013] In a third aspect, the present application provides an electronic device comprising: at least one memory for storing programs; and at least one processor for executing the programs stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method described in the first aspect or any possible implementation of the first aspect.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0015] In a fifth aspect, the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0016] It can be understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0017] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the prior art: The present application provides a satellite-to-ground precipitation data fusion method, device, electronic device and medium. By introducing multi-source fusion data consisting of satellite-to-ground precipitation data and multiple types of terrain factor data, the spatial autocorrelation of precipitation data is fully considered, and the multi-source fusion data with low spatial resolution is converted into multi-source fusion data with high spatial resolution by using spatial interpolation. The multi-source fusion data with high and low spatial resolutions are further used to perform geographically weighted regression analysis and data fusion calculation, establish a regression relationship between precipitation and auxiliary variable information, reduce the spatial non-stationarity between precipitation and each auxiliary variable, and effectively obtain high spatial resolution and high precision precipitation data in the target area, thereby greatly improving the spatial resolution and precision of precipitation data. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is one of the flow charts of the satellite-ground precipitation data fusion method provided in the embodiment of the present application; Figure 2(a), (b) and (c) are respectively scatter diagrams of satellite precipitation data and ground station precipitation data on the time scales of day, month and year provided in the embodiments of the present application; (d), (e) and (f) are respectively scatter diagrams of fused precipitation data and ground station precipitation data on the time scales of day, month and year provided in the embodiments of the present application; Figure 3 (a), (b) and (c) are schematic diagrams of spatial distribution of satellite precipitation data, fused precipitation data and ground station precipitation data provided in the embodiments of the present application respectively; Figure 4 (a), (c) and (e) are schematic diagrams for comparing the daily runoff simulated by using the ground site precipitation data, satellite precipitation data and fused precipitation data obtained at the hydrological site 1 provided in the embodiment of the present application and the actual daily runoff; (b), (d) and (f) are schematic diagrams for comparing the daily runoff simulated by using the ground site precipitation data, satellite precipitation data and fused precipitation data obtained at the hydrological site 2 provided in the embodiment of the present application and the actual daily runoff; Figure 5 (a), (c) and (e) are schematic diagrams for comparing the monthly runoff simulated by using the ground site precipitation data, satellite precipitation data and fused precipitation data obtained at the hydrological site 1 provided in the embodiment of the present application and the measured monthly runoff; (b), (d) and (f) are schematic diagrams for comparing the monthly runoff simulated by using the ground site precipitation data, satellite precipitation data and fused precipitation data obtained at the hydrological site 2 provided in the embodiment of the present application and the measured monthly runoff; Figure 6 This is the second flow chart of the satellite-ground precipitation data fusion method provided in the embodiment of the present application; Figure 7 It is a structural schematic diagram of a satellite-ground precipitation data fusion device provided in an embodiment of the present application; Figure 8 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0020] The terms "first" and "second" in the specification and claims of this application are used to distinguish different objects rather than to describe a specific order of objects. For example, the first spatial resolution and the second spatial resolution are used to distinguish spatial resolutions of different precisions rather than to describe the spatial resolution of a response message.

[0021] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0022] In the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more than two. For example, a plurality of terrain factor data refers to two or more terrain factor data.

[0023] Some studies have emphasized that combining the advantages of ground station observation precipitation data and satellite inversion precipitation data is a feasible strategy to improve the quality of precipitation products. The mainstream methods currently proposed for fusing precipitation products include mean deviation correction, linear regression model correction, dual-kernel smoothing correction, Bayesian fusion, and geographic difference analysis. However, the above fusion methods do not fully consider the spatial autocorrelation and spatiotemporal heterogeneity of precipitation, resulting in the spatial resolution and accuracy of the final precipitation data still need to be improved.

[0024] To this end, the present application provides a satellite-ground precipitation data fusion method, device, electronic device and medium to solve the defects in the above-mentioned prior art.

[0025] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0026] Figure 1 This is one of the flow charts of the satellite-ground precipitation data fusion method provided in the embodiment of the present application, such as Figure 1 As shown, including: Step S1, obtaining first precipitation data of a first spatial resolution inverted by satellite remote sensing in a target area, and performing spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; the second spatial resolution is higher than the first spatial resolution; Step S2, resampling the precipitation observation data and multiple terrain factor data of each ground station in the target area at a spatial resolution to obtain second precipitation data and multiple terrain factor data at a first spatial resolution and a second spatial resolution; Step S3, performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution, to determine fused precipitation data at the second spatial resolution in the target area.

[0027] Specifically, the target area described in the embodiment of the present application refers to the location of the target monitoring area in precipitation monitoring, which can be specifically determined based on the actual monitored research area.

[0028] The first spatial resolution described in the embodiment of the present application refers to the spatial resolution of precipitation data monitored by existing remote sensing satellites, which can be used to characterize a relatively low spatial resolution, generally 0.1°×0.1°.

[0029] The second spatial resolution described in the embodiment of the present application can be used to represent a spatial resolution higher than the first spatial resolution, which can be determined according to actual application requirements. For example, the second spatial resolution can be 0.01°×0.01°.

[0030] The first precipitation data described in the embodiment of the present application refers to precipitation data obtained by remote sensing inversion calculation of precipitation in the target area using satellite precipitation products.

[0031] The second precipitation data described in the embodiment of the present application refers to precipitation observation data monitored by various ground stations within the target area.

[0032] The multiple terrain factor data described in the embodiments of the present application refer to data obtained by detecting multiple types of terrain factors in the target area, wherein the multiple types of terrain factors may include DEM elevation, slope, slope direction, terrain undulation and mountain shadow, etc.

[0033] In an embodiment of the present application, in step S1, by adopting satellite precipitation products, such as IMERG-F products, precipitation data with a spatial resolution of 0.1°×0.1° inverted by satellite remote sensing in the target area, that is, first precipitation data with a first spatial resolution, can be obtained; and the first precipitation data with the first spatial resolution can be spatially interpolated by using the Kriging interpolation method to downscale the precipitation data with a spatial resolution of 0.1°×0.1° to 0.01°×0.01° to obtain precipitation data with a spatial resolution of 0.01°×0.01°, that is, the first precipitation data with a second spatial resolution is obtained.

[0034] In an embodiment of the present application, in step S2, precipitation observation data in the target area can be obtained by monitoring precipitation data at various ground stations in the target area, and multiple terrain factor data of the target area can be extracted by obtaining high-quality DEM data and using GIS software, including but not limited to terrain factor data such as elevation, slope, and aspect. Then, the precipitation observation data and multiple terrain factor data of various ground stations in the target area are spatially resampled. Specifically, the spatial resolution of the precipitation observation data and terrain factor data of the ground stations in the target area can be resampled to 0.1°×0.1° and 0.01°×0.01° by using the inverse distance weighting method, that is, the second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution are obtained.

[0035] It should be noted that, in the embodiments of the present application, the first precipitation data of the first spatial resolution and the precipitation observation data of each ground station are precipitation data obtained through data preprocessing in advance. These data preprocessing operations include conventional data preprocessing operations such as data outliers and missing values ​​to ensure that the obtained precipitation data is accurate and reliable. The present application does not make any specific limitations on this.

[0036] Furthermore, in an embodiment of the present application, in step S3, based on the first precipitation data, the second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution, the precipitation data and terrain factor data at the first spatial resolution (such as 0.1°×0.1°) can first be used to perform geographically weighted regression analysis to calculate the regression coefficients and regression residual data in the geographically weighted regression analysis, and then these regression coefficients and regression residual data can be converted to high spatial resolution through spatial interpolation, that is, the regression coefficients and regression residual data at the second spatial resolution (such as 0.01°×0.01°) are obtained, so that the first precipitation data, the second precipitation data and multiple terrain factor data at the second spatial resolution can be combined to perform geographically weighted regression analysis again, and finally the fused precipitation data at the second spatial resolution in the target area is determined.

[0037] The satellite-ground precipitation data fusion method of the embodiment of the present application introduces multi-source fusion data composed of satellite-ground precipitation data and multiple types of terrain factor data, fully considers the spatial autocorrelation of precipitation data, and uses spatial interpolation to convert multi-source fusion data with low spatial resolution into multi-source fusion data with high spatial resolution; further uses the multi-source fusion data with high and low spatial resolutions to perform geographically weighted regression analysis and data fusion calculation, establishes a regression relationship between precipitation and auxiliary variable information, reduces the spatial non-stationarity between precipitation and each auxiliary variable, and can effectively obtain high spatial resolution and high precision precipitation data in the target area, thereby greatly improving the spatial resolution and precision of precipitation data.

[0038] Based on the content of the above embodiment, as an optional embodiment, a geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, including: Using a Geographically Weighted Regression (GWR) model, a regression analysis is performed on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the regression residual data and precipitation estimation value data of the GWR model at the second spatial resolution; The precipitation estimation data and the regression residual data at the second spatial resolution are summed to determine the fused precipitation data at the second spatial resolution in the target area.

[0039] Specifically, the regression residual data described in the embodiment of the present application refers to the difference between the precipitation estimate data predicted by the model and the precipitation data observed at the ground station in the GWR analysis, which can be used to evaluate the degree of fit of the GWR model.

[0040] More specifically, in an embodiment of the present application, based on the GWR model, Geographically Weighted Regression Kriging (GWRK) is used to carry out precipitation fusion research.

[0041] In this embodiment, the GWRK method is an extended hybrid method based on geographically weighted regression (GWR), which interpolates the geographically weighted regression residuals in combination with the Kriging method. GWRK introduces the sample spatial location on the basis of the traditional linear regression method to estimate the spatial distribution of precipitation errors in the region and explore the heterogeneity of the spatial relationship between variables.

[0042] Based on the content of the above embodiment, as an optional embodiment, a regression analysis is performed on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution using a GWR model to determine the regression residual data and precipitation estimation value data of the GWR model at the second spatial resolution, including: Using the GWR model to perform regression analysis on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution, and determine the regression coefficient and regression residual data of the GWR model at the first spatial resolution; Perform spatial interpolation calculation on the regression coefficient and regression residual data of the first spatial resolution to obtain the regression coefficient and regression residual data of the geographically weighted regression model at the second spatial resolution; The first precipitation data and the plurality of terrain factor data at the second spatial resolution, as well as the regression coefficient of the second spatial resolution are substituted into the geographically weighted regression model to determine the precipitation estimation value data of the geographically weighted regression model at the second spatial resolution.

[0043] Specifically, in the embodiment of the present application, the first precipitation data, the second precipitation data and the plurality of terrain factor data at the first spatial resolution are firstly subjected to regression analysis using the GWR model. The expression of the GWR model can be expressed as: ; In the formula, Indicates ground station i The latitude and longitude coordinates of; Indicates ground station i GWR precipitation estimates in mm; p Indicates the number of influencing variables; Indicates ground station i The constant term (intercept term) of ; Indicates k Influencing variables at ground stations i The regression coefficient on ; Indicates ground station i No. k Data on the variables that affect it; Indicates ground station i The regression residual data, in mm, can be interpolated by the Kriging method to calculate the regression residual in the target area.

[0044] Here, the data affecting the variables include the first precipitation data and multiple terrain factor data, namely, satellite precipitation data and parameter data of multiple terrain factors. In the specific regression analysis, the ground station precipitation data at a spatial resolution of 0.1°×0.1°, namely, the second precipitation data, is substituted into the model. The satellite precipitation data with a spatial resolution of 0.1°×0.1° and multiple terrain factor data were used as the data of each influencing variable to substitute into the model. The regression coefficient of the GWR model at a spatial resolution of 0.1°×0.1° was determined by regression calculation. and regression residual data .

[0045] Furthermore, in the embodiment of the present application, the regression coefficients at a spatial resolution of 0.1°×0.1° are calculated using the inverse distance weighting method. By performing spatial interpolation calculation, the regression coefficient at a spatial resolution of 0.01°×0.01° can be obtained. At the same time, the Kriging interpolation method can be used to interpolate the regression residual data at a spatial resolution of 0.1°×0.1° Perform spatial interpolation calculations to obtain regression residual data at a spatial resolution of 0.01°×0.01° .

[0046] Furthermore, in an embodiment of the present application, the first precipitation data and multiple terrain factor data at a spatial resolution of 0.01°×0.01° are used as influencing variables, combined with the regression coefficients at a spatial resolution of 0.01°×0.01°, and the data are substituted into the GWR model again, and the precipitation estimation value data of the GWR model at a spatial resolution of 0.01°×0.01° can be calculated.

[0047] The method of the embodiment of the present application introduces multi-source fusion data consisting of satellite-ground precipitation data and multiple types of terrain factor data to perform geographically weighted regression analysis, and uses a spatial interpolation algorithm to convert the model regression coefficients and residuals at low spatial resolution to high spatial resolution. It can accurately output high spatial resolution precipitation estimation data and regression residual data, which is conducive to further improving the accuracy of subsequent fusion precipitation calculation results.

[0048] Secondly, the GWRK fused precipitation data estimate consists of two parts: regression residual and GWR precipitation estimate. Its specific expression can be expressed as: ; In the formula, Indicates ground station i GWRK precipitation estimates, i.e. fused precipitation data, in mm; Represents the ground station after Kriging interpolation i The regression residual data of , in mm.

[0049] Furthermore, the precipitation estimation data at a spatial resolution of 0.01°×0.01° and regression residual data at a spatial resolution of 0.01°×0.01° By summing, we can get the fused precipitation data at a spatial resolution of 0.01°×0.01° in the target area. .

[0050] It can be understood that when the data input into the above GWR model uses the first precipitation data, the second precipitation data and multiple terrain factor data under monthly statistics, after being processed by the above regression analysis method, the model can output fused monthly precipitation data at a spatial resolution of 0.01°×0.01°.

[0051] The method of the embodiment of the present application introduces the geographically weighted regression kriging method to construct a satellite-ground fusion precipitation model, combines the advantages of the geographically weighted regression method and the kriging interpolation method, fully considers the spatial autocorrelation of precipitation and the spatial non-stationarity between precipitation and auxiliary variables, and to a certain extent improves the shortcomings of the traditional fusion precipitation method that does not fully consider the spatial autocorrelation of precipitation and the spatiotemporal heterogeneity, and significantly improves the quality of precipitation data.

[0052] In practical applications, in the process of precipitation data fusion, auxiliary variables are important input data, and their high dimensionality and redundancy will also affect the accuracy and stability of the fusion results.

[0053] Based on the content of the above embodiment, as an optional embodiment, after performing spatial interpolation on the first precipitation data with the first spatial resolution to obtain the first precipitation data with the second spatial resolution, the method further includes: Based on the multiple terrain factor data, principal component analysis is performed on the multiple terrain factors corresponding to the multiple terrain factor data to determine multiple terrain principal component data; Resampling the precipitation observation data and multiple terrain principal component data at various ground stations in the target area at different spatial resolutions to obtain second precipitation data and multiple terrain principal component data at the first spatial resolution and the second spatial resolution; A geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain principal component data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area.

[0054] Specifically, in the embodiments of the present application, satellite monthly precipitation data, longitude and latitude, and a variety of terrain factors are used as influencing variables. Among them, combined with the correlation between various terrain factors and precipitation, elevation, slope, and slope aspect are mainly selected for research.

[0055] Based on the content of the above embodiment, as an optional embodiment, the multiple terrain factors include elevation, slope and aspect, and also include at least one of terrain relief and mountain shadow.

[0056] It should be noted that the effects of terrain relief and mountain shadow on precipitation data are intertwined. Terrain relief not only directly affects the precipitation process through dynamic uplift and rain shadow effect, but may also affect precipitation indirectly by changing local climate and solar radiation distribution. Mountain shadow, as an indicator of the interaction between terrain and solar radiation, further increases the complexity of precipitation distribution.

[0057] The method of the embodiment of the present application can further improve the accuracy of precipitation data by considering the influence of terrain undulation and mountain shadow on precipitation data and introducing terrain factors of terrain undulation and mountain shadow to fuse precipitation calculation.

[0058] Preferably, in an embodiment of the present application, five types of terrain factor data, namely, elevation, slope, aspect terrain relief and mountain shadow, are used for subsequent fused precipitation calculation.

[0059] First, the original data needs to be preprocessed to verify whether the selected terrain factors are independent of each other and solve the redundancy problem, thereby improving the regression accuracy. The principal component analysis (PCA) method can be used to reduce the dimensionality of the above five types of terrain factors to eliminate multiple linear relationships and redundancy problems.

[0060] In a specific embodiment of the present application, the eigenvectors and eigenvalues ​​of each terrain factor are shown in Table 1 and Table 2.

[0061] Table 1

[0062] Table 2

[0063] It can be seen from Tables 1 and 2 above that the cumulative contribution rate of principal components 1, 2 and 3 reaches 90.51%, which has the largest contribution to the terrain factor. It can be considered that the data of the first three terrain principal components can better replace the data of the above five types of terrain factors. Therefore, the data group of the first three terrain principal components can be finally selected to represent the terrain factor variables for subsequent calculations.

[0064] Further, in the embodiment of the present application, the precipitation observation data of each ground station in the target area and the three selected terrain principal component data are spatially resampled to obtain the second precipitation data and three terrain principal component data at a low spatial resolution of 0.1°×0.1° and a high spatial resolution of 0.01°×0.01°. Then, according to the above method, a geographically weighted regression analysis can be performed based on the first precipitation data, the second precipitation data and multiple terrain principal component data at a low spatial resolution of 0.1°×0.1° and a high spatial resolution of 0.01°×0.01°, and finally the fused precipitation data at a high spatial resolution of 0.01°×0.01° in the target area is determined.

[0065] The method of the embodiment of the present application adopts the PCA method to perform principal component analysis on multiple terrain factor data, thereby achieving dimensionality reduction processing of high-dimensional terrain data and eliminating multiple linear relationships and redundancy problems between data. This not only reduces the dimensionality and simplifies the data and reduces the consumption of computing resources, but also further improves the quality of the fused precipitation data.

[0066] Based on the content of the above embodiment, as an optional embodiment, the fused precipitation data is fused monthly precipitation data; correspondingly, after performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: Determine the ratio of daily precipitation data of each ground station to the monthly precipitation data; The inverse distance weighted method is used to perform spatial interpolation on each ratio data to obtain the ratio data corresponding to the target area; Determine the fused daily precipitation data at the second spatial resolution in the target area based on the ratio data corresponding to the target area and the fused monthly precipitation data at the second spatial resolution in the target area; The fused monthly precipitation data at the second spatial resolution in the target area are accumulated to determine the fused annual precipitation data at the second spatial resolution in the target area.

[0067] Specifically, in an embodiment of the present application, after obtaining the fused precipitation data at the second spatial resolution in the target area, the proportional index method can be further selected according to the ratio of daily precipitation at the ground station to the precipitation of the month, and the aforementioned monthly fused precipitation can be proportionally decomposed into precipitation data at the daily scale.

[0068] The specific steps are as follows: calculate the ratio data of the daily precipitation data of the ground station and its monthly precipitation data, and then use the inverse distance weighted method to perform spatial interpolation on the ratio data to obtain the ratio data of the entire target area. Furthermore, multiply the fused monthly precipitation data at the second spatial resolution in the target area with the interpolated ratio data to effectively allocate the fused monthly precipitation data to the daily scale, thereby obtaining the fused daily precipitation data at the second spatial resolution in the target area.

[0069] Similarly, the fused monthly precipitation data at the second spatial resolution in the target area are accumulated month by month to obtain the fused annual precipitation data at the second spatial resolution in the target area.

[0070] The method of the embodiment of the present application combines the proportional index method and the spatial interpolation method, and uses the fused monthly precipitation data at high spatial resolution to perform proportional calculation, so as to obtain precipitation data of the target area on a daily and annual time scale with high data quality.

[0071] In order to compare the applicability of different precipitation products, the satellite precipitation data and fused precipitation data were interpolated to the ground station scale using the inverse distance weighted method. The effectiveness of the satellite precipitation products used was quantitatively evaluated using four accuracy evaluation indicators, namely the Pearson correlation coefficient (R), root mean square error (RMSE), standard deviation (SD) and relative bias (BIAS).

[0072] Figure 2(a), (b) and (c) are scatter diagrams of satellite precipitation data and ground station precipitation data at daily, monthly and annual time scales provided in the embodiments of the present application; (d), (e) and (f) are scatter diagrams of fused precipitation data and ground station precipitation data at daily, monthly and annual time scales provided in the embodiments of the present application. As shown in the figure, the correlation between satellite precipitation and fused precipitation and station precipitation generally shows an upward trend with the increase of time scale. Specifically, the correlation between satellite precipitation and station precipitation is highest on the monthly scale (R=0.943) and lowest on the daily scale (R=0.486). Since monthly and annual satellite precipitation is accumulated from daily precipitation, their RMSE and SD values ​​increase accordingly, while the BIAS value remains stable at all time scales, all at 0.05, indicating that satellite precipitation overestimates the measured value to a certain extent. In contrast, the fused precipitation has a similar trend to that of satellite precipitation, but its performance is significantly improved, with R values ​​all above 0.96, and it has a very high correlation with station precipitation. Although the SD value of the fused precipitation is slightly higher than that of the original satellite precipitation, which may be related to the large deviation of some data from the mean, its RMSE and BIAS values ​​are significantly reduced. The fused precipitation shows better evaluation results at all time scales.

[0073] To further compare the spatial differences in precipitation between satellite, fusion and ground stations, Figure 3 The spatial distribution of the three monthly average precipitations is presented. It can be seen that the satellite monthly average precipitation significantly overestimates the actual precipitation in some areas, and underestimates the actual precipitation in other areas, and there is an overall phenomenon of "low value overestimation"; the spatial distribution of the fused monthly average precipitation is very similar to the station monthly average precipitation, and the overall distribution shows the characteristics of high in the middle and low around; this shows that the fused precipitation in the embodiment of the present application can effectively describe the spatial distribution characteristics of precipitation in the target area.

[0074] Based on the content of the above embodiment, as an optional embodiment, after performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: The fused precipitation data at the second spatial resolution is used to perform hydrological runoff simulation analysis to determine the runoff information of the target area at different time scales.

[0075] Specifically, in an embodiment of the present application, based on ground station precipitation data, satellite precipitation data and the fused precipitation data obtained by the above calculation, the SWAT hydrological model is driven to carry out daily and monthly runoff simulation research to verify the hydrological application value of the obtained fused precipitation data.

[0076] In this example, the SWAT model is a distributed hydrological model that can simulate the coupling process of multiple factors such as water, sediment, and agriculture in a complex watershed. The hydrological process of the SWAT model includes multiple links such as surface runoff, evapotranspiration, and soil water. Its core calculation formula is as follows: ; In the formula, , Day 0 and day t Soil moisture content of the day, in mm; t is the time step in days; For the i The daily precipitation, in mm; For the i The daily surface runoff, in mm; For the i Daily evapotranspiration, in mm; For the i The soil flow of the day, in mm; For the i The daily base flow in mm.

[0077] Among them, for the calculation of surface runoff, the SCS runoff curve method is mainly used, which can represent the runoff of different types of underlying surfaces; in view of the fact that the target area has data such as air temperature, wind speed, relative humidity and solar radiation, the Penman-Montieth method with greater application potential is selected to calculate the potential evapotranspiration, and the actual evapotranspiration is obtained through the exponential relationship between soil thickness and water content; the trough storage algorithm is used to calculate the subsoil flow; the base flow is simulated by the exponential decay weight function.

[0078] More specifically, in a specific embodiment of the present application, the SWAT hydrological model is driven to simulate runoff based on ground site precipitation, satellite precipitation and fused precipitation. The SWAT model construction steps are as follows: Step 1: Combine ArcGIS software to establish the basic database required for the SWAT model, including geospatial database, land use database, soil database and meteorological database. First, download the original DEM data with a resolution of 90×90m in the relevant geospatial data cloud, use ArcGIS software to perform projection conversion, splicing and clipping, build a geospatial database, and then obtain the elevation distribution map of the target area. Select the raster data of land use types at a specific time (such as 2018), use ArcGIS software to clip, project, convert, reclassify, etc., divide the land use types of the target area into six categories, and then build a land use database.

[0079] The soil database consists of three parts: soil type distribution, physical properties, and a type index table. Among them, the soil type data is selected from the Harmonized World Soil Database (HWSD). Through processing such as clipping, projection transformation, and reclassification using ArcGIS software, the soils in the target area are mainly divided into 12 types, and the spatial distribution of soil types is drawn. The soil database is constructed with the reclassified soil data, mainly including parameters such as the number of soil layers, available water holding capacity, wet bulk density, hydrological components, and erodibility factor. The meteorological database is the main driving factor of the SWAT model. Among them, precipitation data respectively adopts ground station precipitation, satellite precipitation, and merged precipitation. Other meteorological data mainly comes from the daily value dataset of surface climate data (V3.0). A weather generator is used to simulate and fill in the missing data, and then the corresponding meteorological database is established.

[0080] Step 2: According to the basic conditions of the target area, set the catchment area threshold to 30,000 ha, select Hydrological Station 2 as the basin outlet, divide the sub-basins and divide them into several hydrological response units, and input the station, satellite, and merged daily and monthly precipitation respectively to construct the SWAT daily and monthly runoff simulation models.

[0081] Step 3: Select the optimal parameters of the SWAT model, and use the SUFI-2 algorithm to select the T-Stat and P-Value values for parameter sensitivity analysis to improve the simulation performance of the model. The SUFI-2 algorithm is based on comprehensive optimization and uses gradient search as a means, which can realize the simultaneous calibration of multiple parameters, with shorter required time and higher accuracy. The larger the absolute value of T-Stat and the closer the P-Value is to 0, the more sensitive the parameter is.

[0082] Step 4: Set the warm-up period of the model as 2001 - 2002, the calibration period as 2003 - 2011, and the verification period as 2012 - 2019. After parameter sensitivity analysis, calibrate and verify the model parameters, and conduct daily and monthly runoff simulation studies under different precipitation scenarios. Select the coefficient of determination R 2 , Nash-Sutcliffe efficiency coefficient NSE, and root mean square error RMSE to verify the hydrological application effect of the merged precipitation. R 2 is used to evaluate the goodness of fit between the measured value and the simulated value. The closer this value is to 1, the better the fit and the better the runoff simulation effect. NSE reflects the quality of the runoff simulation result. The larger this value is, the better the agreement between the measured value and the simulated value. RMSE reflects the average error degree. The smaller this value is, the higher the accuracy of the simulated data. For the evaluation of the SWAT model performance, mainly R 2 and NSE are selected. Combining the characteristics of the target area, the SWAT model performance is divided into four categories: excellent (0.7 < R 2 ≤1, 0.75 < NSE ≤ 1), good (0.6 < R 2 ≤0.7, 0.65 < NSE ≤ 0.75), satisfactory (0.5 < R2 ≤0.6, 0.5 < NSE ≤ 0.65), dissatisfied (R 2 ≤0.5, NSE ≤ 0.5).

[0083] Table 3

[0084] Table 3 shows the evaluation results of the daily and monthly runoff simulation accuracy of the SWAT model under different precipitation scenarios. The results show that in terms of daily runoff simulation, the runoff simulation based on the integrated precipitation performs best, with its R 2 and NSE both exceeding 0.63, and the performance reaching "satisfactory" or above; the runoff simulation results under the station precipitation scenario are the second; in contrast, the runoff simulation based on the original satellite precipitation performs generally, with the R 2 and NSE of Station 1 both less than 0.60. In terms of monthly runoff simulation, the runoff simulation based on the integrated precipitation still generally performs best, with the R 2 and NSE in each period both reaching above 0.88, and the performance reaching "excellent", significantly better than the daily runoff simulation results; generally speaking, the integrated precipitation data has good hydrological application effects in the target area and can effectively replace the original satellite precipitation data to carry out regional hydrological simulation.

[0085] To further evaluate the application value of the integrated precipitation, Figure 4 shows the comparison results of the simulated daily runoff and the measured daily runoff driven by different precipitation data obtained from Hydrological Station 1 and Hydrological Station 2 during the calibration period from 2003 to 2011 and the verification period from 2012 to 2019. It can be seen from Figure 4 that all three precipitation data inputs underestimated the peak flow and overestimated the base flow, which may be due to the narrow river valley, complex terrain, and frequent human activities (such as reservoir regulation, artificial water intake, etc.) in the region, resulting in an increase in the uncertainty of hydrological simulation; during the calibration period and the verification period, the simulated daily runoff driven by the integrated precipitation data is closest to the measured results; the results of the simulated daily runoff by the ground station precipitation data are the second; while the simulated daily runoff by the satellite precipitation data generally shows lower simulation performance, and the simulation results are lower than the previous two, with a large deviation; this result confirms the potential of the integrated precipitation provided in the embodiments of the present application in the daily runoff simulation of the target area.

[0086] Such as Figure 5As shown, it compares the simulated monthly runoff and the measured monthly runoff processes driven by different precipitation data obtained from hydrological stations 1 and 2 during the rate period of 2003-2011 and the validation period of 2012-2019. The results show that compared with other precipitation products, the monthly runoff simulated by the fusion precipitation in the embodiment of the present application is closest to the measured monthly runoff, and the simulation performance of satellite precipitation is relatively poor. However, it should be noted that compared with the simulation of daily runoff, satellite precipitation can better simulate the peak value of monthly runoff. This phenomenon may be due to the better accuracy of satellite precipitation in detecting high values ​​of monthly precipitation; in addition, the monthly runoff simulation results driven by the three precipitation data are closer to reality than the daily runoff simulation results.

[0087] Figure 6 This is the second flow chart of the satellite-ground precipitation data fusion method provided in the embodiment of the present application, such as Figure 6 As shown, in an embodiment of the present application, firstly, IMERG-F satellite precipitation data, ground station precipitation data, terrain factor variables and other basic data in the target area are obtained, and data variable preprocessing is carried out.

[0088] In this embodiment, terrain factors include elevation, slope, aspect, terrain undulation and mountain shadow, and other basic data may include land use, soil data, temperature, wind speed, relative humidity and solar radiation; IMERG-F satellite precipitation data with a temporal and spatial resolution of 1 day and 0.1°×0.1° for a specific period (such as 2001-2019) is obtained from the relevant satellite data and information service center; the terrain factors are extracted by processing and analyzing the DEM data provided by the relevant geospatial data cloud platform through ArcGIS software; after collecting the data, the outliers and missing values ​​are pre-processed; the temporal and spatial resolutions of satellite precipitation data, ground station precipitation data, terrain factors and other basic data are unified.

[0089] Then, the PCA method was used to screen and reduce the dimension of terrain factor variables, eliminate multiple linear relationships and redundancy problems, and extract representative terrain variables for subsequent research.

[0090] In this example, combined with the correlation between local terrain factors and precipitation, five types of terrain factors, namely elevation, slope, aspect, terrain undulation and mountain shadow, are selected as the main inputs of the precipitation fusion model; since there may be a strong correlation between the factors, the PCA method is used to reduce the dimensions of the local terrain factors and eliminate the repeated information between the factors; by standardizing the original indicators, the standardized correlation coefficient matrix, eigenvalue, eigenvector and contribution rate of the local terrain factors are calculated; when the cumulative contribution rate of the principal component reaches more than 90%, it can better replace the five types of terrain factors, thereby determining multiple terrain principal component data, and identifying them as representative terrain factor variables for subsequent research.

[0091] Next, the GWRK method was introduced to construct a satellite-ground-month precipitation fusion model. The GWR results were combined with the Kriging interpolation results to obtain the fused monthly precipitation data at the target spatial resolution of 0.01°×0.01°. According to the ratio of daily precipitation at the ground station to the monthly precipitation, the proportional index method was used to obtain the fused daily precipitation data, and the annual fused precipitation was accumulated from the monthly fused precipitation.

[0092] Specifically, the fusion steps of IMERG-F satellite and ground station precipitation data based on GWRK method are as follows: Step 1: Select Gauss kernel function as the spatial weight function, determine the bandwidth using AIC criterion, and build the GWRK satellite-earth-moon precipitation fusion model.

[0093] Step 2: The spatial resolution of the monthly precipitation data and topographic principal component data of the ground stations in the study area were resampled to 0.1°×0.1° and 0.01°×0.01° using the inverse distance weighted method, and the satellite monthly precipitation data were downscaled to 0.01°×0.01° using the Kriging interpolation method.

[0094] Step 3: Using satellite monthly precipitation and multiple terrain principal components at a resolution of 0.1°×0.1° as auxiliary variables and ground station monthly precipitation as the dependent variable, a GWRK satellite-ground-moon precipitation fusion model with a resolution of 0.1°×0.1° was constructed to obtain GWR precipitation estimation data and regression coefficients at the corresponding spatial scale, and the spatial resolution of the regression coefficients was interpolated to 0.01°×0.01° using the inverse distance weighted method.

[0095] Step 4: Substitute the satellite monthly precipitation with a resolution of 0.01°×0.01°, multiple terrain principal component data and regression coefficients back into the GWRK model to obtain the GWR precipitation estimation data with a resolution of 0.01°×0.01°.

[0096] Step 5: Calculate the regression residual data with a resolution of 0.1°×0.1°, that is, the difference between the monthly precipitation at the ground station and the fused monthly precipitation, and use the Kriging interpolation method to interpolate the resolution of the regression residual to 0.01°×0.01°, and add it to the GWR precipitation estimate data at the corresponding resolution to obtain the GWRK-corrected fused monthly precipitation data.

[0097] Step six, according to the ratio of daily precipitation at the ground station to the monthly precipitation, the proportional index method is used to decompose the corrected monthly fused precipitation into daily scale in proportion. The specific steps are: calculate the ratio of daily precipitation at the ground station to the monthly precipitation, use the inverse distance weighted method to perform spatial interpolation on it, and obtain the ratio data of the entire study area. The fused monthly precipitation data is multiplied by the interpolated ratio to effectively allocate the fused monthly precipitation data to the daily scale, and the fused annual precipitation data is accumulated from the fused monthly precipitation data.

[0098] Finally, according to the aforementioned implementation method, based on ground station precipitation data, satellite precipitation data and fused precipitation data, the SWAT hydrological model is driven to carry out daily and monthly runoff simulation research, demonstrating the hydrological application value of the fused precipitation data provided in the embodiment of this application.

[0099] The satellite-to-ground precipitation data fusion device provided in the present application is described below. The satellite-to-ground precipitation data fusion device described below and the satellite-to-ground precipitation data fusion method described above can refer to each other.

[0100] Figure 7 is a schematic diagram of the structure of the satellite-ground precipitation data fusion device provided in the embodiment of the present application, such as Figure 7 As shown, including: The interpolation module 10 is used to obtain first precipitation data of a first spatial resolution in a target area, and perform spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; the first precipitation data of the first spatial resolution is obtained by satellite remote sensing inversion; the second spatial resolution is higher than the first spatial resolution; The resampling module 20 is used to perform spatial resolution resampling on the precipitation observation data and multiple terrain factor data of each ground station in the target area to obtain second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution; The fusion module 30 is used to perform geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area.

[0101] It can be understood that the detailed functional implementation of each of the above-mentioned units / modules can be found in the introduction of the aforementioned method embodiment, and will not be repeated here.

[0102] It should be understood that the above-mentioned device is used to execute the method in the above-mentioned embodiment. The implementation principle and technical effect of the corresponding program module in the device are similar to those described in the above-mentioned method. The working process of the device can refer to the corresponding process in the above-mentioned method, which will not be repeated here.

[0103] The satellite-to-ground precipitation data fusion device of the embodiment of the present application introduces multi-source fusion data composed of satellite-to-ground precipitation data and multiple types of terrain factor data, fully considers the spatial autocorrelation of precipitation data, and uses spatial interpolation to convert multi-source fusion data with low spatial resolution into multi-source fusion data with high spatial resolution; further uses the multi-source fusion data with high and low spatial resolutions to perform geographically weighted regression analysis and data fusion calculation, establishes a regression relationship between precipitation and auxiliary variable information, reduces the spatial non-stationarity between precipitation and each auxiliary variable, and can effectively obtain high spatial resolution and high-precision precipitation data in the target area, thereby greatly improving the spatial resolution and accuracy of precipitation data.

[0104] Based on the method in the above embodiment, the embodiment of the present application provides an electronic device, as shown in the figure, the electronic device may include: a processor (Processor) 810, a communication interface (Communications Interface) 820, a memory (Memory) 830 and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call the logic instructions in the memory 830 to execute the method in the above embodiment.

[0105] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0106] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method in the above embodiment.

[0107] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method in the above embodiment.

[0108] It is understandable that the processor in the embodiment of the present application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0109] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, and the software modules can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC.

[0110] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions may be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

[0111] It should be understood that the various numerical numbers involved in the embodiments of the present application are only used for the convenience of description and are not used to limit the scope of the embodiments of the present application.

[0112] It should be understood that expressions such as "including" and "may include" that may be used in the present application indicate the existence of the disclosed functions, operations, or constituent elements, and do not limit one or more additional functions, operations, and constituent elements. In the present application, terms such as "including" and / or "having" may be interpreted as indicating specific characteristics, numbers, operations, constituent elements, components, or combinations thereof, but may not be interpreted as excluding the existence or possibility of adding one or more other characteristics, numbers, operations, constituent elements, components, or combinations thereof.

[0113] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for fusion of satellite and ground precipitation data, characterized in that: include: Acquire first precipitation data of a first spatial resolution inverted by satellite remote sensing in a target area, and perform spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; The second spatial resolution is higher than the first spatial resolution; Resampling the precipitation observation data and multiple terrain factor data at each ground station in the target area at spatial resolution to obtain second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution; A geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine fused precipitation data at the second spatial resolution in the target area.

2. The satellite-ground precipitation data fusion method according to claim 1, characterized in that: The performing of geographically weighted regression analysis based on the first precipitation data, the second precipitation data and the plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area includes: Using a geographically weighted regression model, a regression analysis is performed on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the regression residual data and precipitation estimation value data of the geographically weighted regression model at the second spatial resolution; The precipitation estimation value data and the regression residual data at the second spatial resolution are summed to determine fused precipitation data at the second spatial resolution in the target area.

3. The satellite-ground precipitation data fusion method according to claim 2, characterized in that: The method of performing regression analysis on the first precipitation data, the second precipitation data, and the plurality of terrain factor data at the first spatial resolution and the second spatial resolution using the geographically weighted regression model to determine the regression residual data and precipitation estimation value data of the geographically weighted regression model at the second spatial resolution includes: Using a geographically weighted regression model to perform regression analysis on the first precipitation data, the second precipitation data, and a plurality of terrain factor data at the first spatial resolution, and determine the regression coefficient and regression residual data of the geographically weighted regression model at the first spatial resolution; Performing spatial interpolation calculation on the regression coefficient and regression residual data of the first spatial resolution to obtain the regression coefficient and regression residual data of the geographically weighted regression model at the second spatial resolution; Substitute the first precipitation data and multiple terrain factor data at the second spatial resolution, and the regression coefficient of the second spatial resolution into the geographically weighted regression model to determine the precipitation estimation value data of the geographically weighted regression model at the second spatial resolution.

4. The satellite-ground precipitation data fusion method according to claim 1, characterized in that: After performing spatial interpolation on the first precipitation data of the first spatial resolution to obtain the first precipitation data of the second spatial resolution, the method further includes: Performing principal component analysis on the multiple terrain factors corresponding to the multiple terrain factor data to determine multiple terrain principal component data; Resampling the precipitation observation data of each ground station in the target area and the plurality of terrain principal component data at spatial resolutions to obtain second precipitation data and a plurality of terrain principal component data at the first spatial resolution and the second spatial resolution; A geographically weighted regression analysis is performed based on the first precipitation data, the second precipitation data and a plurality of terrain principal component data at the first spatial resolution and the second spatial resolution to determine fused precipitation data at the second spatial resolution in the target area.

5. The satellite-ground precipitation data fusion method according to claim 4, characterized in that: The plurality of terrain factors include elevation, slope and aspect, and also include at least one of terrain relief and hill shadow.

6. The satellite-ground precipitation data fusion method according to any one of claims 1 to 5, characterized in that: The fused precipitation data is fused monthly precipitation data; correspondingly, after performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data and a plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: Determine the ratio of the daily precipitation data of each ground station to the monthly precipitation data; Using the inverse distance weighted method to perform spatial interpolation on each of the ratio data to obtain ratio data corresponding to the target area; Determining fused daily precipitation data at a second spatial resolution within the target area based on the ratio data corresponding to the target area and the fused monthly precipitation data at a second spatial resolution within the target area; The fused monthly precipitation data at the second spatial resolution in the target area are accumulated to determine the fused annual precipitation data at the second spatial resolution in the target area.

7. The satellite-ground precipitation data fusion method according to any one of claims 1 to 5, characterized in that: After performing geographically weighted regression analysis based on the first precipitation data, the second precipitation data, and the plurality of terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area, the method further includes: The fused precipitation data at the second spatial resolution is used to perform hydrological runoff simulation analysis to determine the runoff information of the target area at different time scales.

8. A satellite-ground precipitation data fusion device, characterized in that: include: An interpolation module, used for acquiring first precipitation data of a first spatial resolution inverted by satellite remote sensing in a target area, and performing spatial interpolation on the first precipitation data of the first spatial resolution to obtain first precipitation data of a second spatial resolution; The second spatial resolution is higher than the first spatial resolution; A resampling module, used for performing spatial resolution resampling on the precipitation observation data and multiple terrain factor data of each ground station in the target area to obtain second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution; A fusion module is used to perform geographically weighted regression analysis based on the first precipitation data, the second precipitation data and multiple terrain factor data at the first spatial resolution and the second spatial resolution to determine the fused precipitation data at the second spatial resolution in the target area.

9. An electronic device, characterized in that: include: at least one memory for storing a computer program; At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program runs on a processor, the processor is caused to execute the method according to any one of claims 1 to 7.

Images