Climate-driven analysis methods, equipment, and media based on spatiotemporal evolution of soil moisture

Abstract

The present invention discloses a climate-driven analysis method, equipment and medium based on the spatiotemporal evolution of soil moisture. The method comprises: using multiple sets of surface soil moisture data sets, including satellite remote sensing, land surface assimilation, diagnostic models, reanalysis models and climate model simulations, to quantify the spatiotemporal variations of global surface soil moisture under the background of anthropogenic climate change, clarify the consistency of the long-term spatiotemporal evolution characteristics of surface soil moisture between different data sets and its driving factors; clarifying the atmospheric physical mechanism behind the global surface soil moisture changes, and using maximum covariance analysis to reveal the response of global surface soil moisture changes to anthropogenic climate change and internal climate variability; by clarifying the spatiotemporal evolution characteristics of surface soil moisture and its climate driving factors, the present invention provides a scientific basis for improving the quality of soil moisture data sets and enhancing the prediction ability of cascading extreme weather events, thereby providing scientific support for achieving the goal of mitigating global warming.

Description

Technical Field

[0001] The present invention relates to a climate-driven analysis method, equipment and medium based on the spatiotemporal evolution of soil moisture. Background Art

[0002] Increased greenhouse gas emissions are an important factor leading to global warming. Against the backdrop of global warming, soil moisture has shown a global decrease due to the influence of anthropogenic climate change. However, soil moisture is an important state variable in the hydrological process, which has an important impact on the land-air water cycle and energy exchange. In the water cycle, soil moisture is the main influencing factor of terrestrial evapotranspiration. On a longer time scale, soil moisture is also a prerequisite for extreme hydrological phenomena such as floods and droughts. Therefore, clarifying the changes in soil moisture under the background of global warming is crucial to improving the accuracy of climate and hydrological process simulation and forecasting. The present invention provides a reference for further research on climate warming and its impact on the natural environment and human production and life by exploring the spatiotemporal distribution characteristics of climate factors such as precipitation and temperature directly acting on soil moisture under the background of global warming.

[0003] Soil moisture datasets can be obtained through different technical means, including field measurements, satellite remote sensing, and model simulations. However, traditional soil moisture change analysis is usually based on a single dataset, without considering the differences in spatiotemporal variations and their driving factors between different datasets. Based on this, analyzing and comparing the spatiotemporal variations of surface soil moisture in different datasets, exploring its relationship with climate drivers (such as precipitation and temperature), and studying the climate dynamics behind these changes are of great significance for assessing the relationship between soil moisture changes and extreme hydrological phenomena caused by land-air water cycle and energy exchange. At the same time, it can reveal which long-term changes are consistent between different datasets, thereby improving confidence in the assessment of soil moisture changes and providing dataset developers with insights to improve the quality of soil moisture datasets. Summary of the Invention

[0004] The purpose of the present invention is to propose a climate-driven analysis method, equipment and medium based on the spatiotemporal evolution of soil moisture, so as to solve the technical problem that the existing technology lacks a method to study the consistency and driving factors of long-term trends in surface soil moisture between different data types (satellite remote sensing, land surface models, diagnostic models, reanalysis and climate models) on a global scale.

[0005] The present invention provides a climate-driven analysis method based on the spatiotemporal evolution of soil moisture, comprising the following steps:

[0006] S1: Data collection: Collect multi-source soil moisture data, large-scale meteorological and environmental variable data;

[0007] S2: Comparison of temporal correlations of surface soil moisture datasets: Combine the surface soil moisture datasets obtained in step S1, perform linear detrending on the annual mean surface soil moisture values, compare the different datasets in pairs, and calculate the median of the time series correlation coefficient of each dataset for all selected grid cells;

[0008] S3: Spatial correlation distribution of surface soil moisture datasets; combining the correlation coefficients between different datasets obtained in step S2, select some datasets, calculate the Spearman correlation coefficients between different datasets, and obtain the spatial correlation distribution characteristics between the above datasets;

[0009] S4: Temporal evolution analysis of soil moisture spatial correlation. Combined with the spatial correlation analysis results of the surface soil moisture dataset obtained in step S3, several sets of the same datasets are selected and the Sen slope is used to estimate the spatial distribution characteristics of the multi-year soil moisture trend. This allows comparison of the multi-year average soil moisture trends between the multiple source datasets. For each dataset pair, the proportion of all grid cells with the same trend direction is calculated.

[0010] S5: Analysis of the physical mechanism of soil moisture changes; Combined with the multi-year average soil moisture trend obtained in step S4, the same method is used to calculate the observed data and climate model data to obtain the spatial distribution characteristics of the annual aridity index AI, annual precipitation PRCP, annual potential evapotranspiration PET, and annual surface air temperature SAT. Different climate zones are divided globally to explore the changes in soil moisture in different climate regions;

[0011] S6: Analysis of climate drivers of soil moisture changes; Combined with the soil moisture change characteristics of different climate regions obtained in step S5, explore the impact of different climate drivers on soil moisture changes. Combined with the sea surface temperature dataset and surface soil moisture dataset obtained in step S1, through maximum covariance analysis, explore the first and second dominant modes and further analyze the impact of sea surface temperature on the spatiotemporal changes of soil moisture.

[0012] A storage medium stores instructions and data for implementing a climate-driven analysis method based on the spatiotemporal evolution of soil moisture.

[0013] A climate-driven analysis device based on the spatiotemporal evolution of soil moisture comprises: a processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a climate-driven analysis method based on the spatiotemporal evolution of soil moisture.

[0014] The beneficial effects provided by the present invention are:

[0015] (1) This paper proposes a method for comparing the spatiotemporal changes of global surface soil moisture based on multiple datasets, analyzes the consistency and differences between different datasets, identifies the main characteristics of the long-term trend of global soil moisture, and provides a scientific basis for monitoring and evaluating soil moisture in the context of global climate change.

[0016] (2) Through a detailed analysis of climate driving factors, this paper reveals the dominant role of these climate factors in global surface soil moisture changes and clarifies the complex relationship between climate driving factors and soil moisture changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic flow chart of the method of the present invention;

[0018] Figure 2 It is a cross-correlation analysis between ESA CCI, GLEAM, GLDAS, reanalysis and CMIP6 datasets;

[0019] Figure 3 is the temporal variation of spatial correlations among annual soil moisture datasets from ESA CCI, GLEAM, GLDAS-EM, Reanalysis-EM, and CMIP6-EM;

[0020] Figure 4 The relationship between the Sen slope (red line) and the proportion of dry areas (bar graph) of the annual average soil moisture of each data set and the climate average aridity index (AI);

[0021] Figure 5 is the relationship between the Sen slope trend of the annual precipitation (PRCP) and surface air temperature (SAT) of each dataset and the annual mean soil moisture;

[0022] Figure 6 is the temporal variation of the first dominant mode (MCA1) of sea surface temperature (SST), soil moisture, and dryness index;

[0023] Figure 7 is the temporal variation of the second dominant mode (MCA2) of sea surface temperature (SST), soil moisture, and dryness index;

[0024] Figure 8 It is a working diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0025] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0026] Before formally explaining the present invention, the scheme of the present invention is first generally explained for easy understanding.

[0027] Please refer to Figure 1 , Figure 1 It is a schematic flow diagram of the method of the present invention.

[0028] The present invention provides a climate-driven analysis method based on the spatiotemporal evolution of soil moisture, comprising the following steps:

[0029] S1: Data collection: Collect multi-source soil moisture data, large-scale meteorological and environmental variable data;

[0030] It should be noted that in step S1 of the present invention, the soil moisture data are satellite remote sensing datasets: ESA-CCI (spatial resolution is 0.25°×0.25°); reanalysis data: ERA5 (0.25°×0.25°), MERRA2 (0.5°×0.625°) and CRA-Land (1°×1°); land surface model data: GLDAS-CLM, GLDAS-VIC and GLDAS-NOAH (1°×1°); diagnostic model data: GLEAM (0.25°×0.25°); climate model data: 18 CMIP6 simulation data; monthly meteorological and environmental data include precipitation, air temperature, potential evapotranspiration, sea surface temperature, meridional wind, zonal wind and geopotential height.

[0031] S2: Comparison of temporal correlations of surface soil moisture datasets: Combine the surface soil moisture datasets obtained in step S1, perform linear detrending on the annual mean surface soil moisture values, compare the different datasets in pairs, and calculate the median of the time series correlation coefficient of each dataset for all selected grid cells;

[0032] As an exemplary embodiment, in step S2, the time correlation of the data set needs to be calculated by linear detrending the annual average soil moisture value: the linear detrending is performed by the least square method. t , fitting linear model y t =a+bt+∈ t , where t is the time a and b are regression coefficients, ∈ t is the residual. Its trend part is y t ′=a+bt. Detrended data means y=y t -y′ t , where the least squares fitting formula for the slope is as follows:

[0033]

[0034] Where b is the slope, n is the total number of years in the study period, and x i is the i-th year of the study period, is the mean value of the years in the study period, yi is the value of the research variable in year i during the research period, is the mean of the research variable during the research period;

[0035] Then, the Spearman correlation coefficients between all selected grid points and different data sets were calculated to analyze the temporal correlation. The value was selected as the cross-correlation value, and the calculation formula was as follows:

[0036]

[0037] In the formula, x and y are two time series of the same length, n is the length of the series x and y, x i and y i represents the value of sequences x and y at position i, and cor(x,y) is the cross-correlation between sequences x and y.

[0038] S3: Spatial correlation distribution of surface soil moisture datasets; combining the correlation coefficients between different datasets obtained in step S2, select some datasets, calculate the Spearman correlation coefficients between different datasets, and obtain the spatial correlation distribution characteristics between the above datasets;

[0039] As an exemplary embodiment, in step S3 of the present invention, the spatial distribution characteristics of the Spearman correlation coefficient of the annual average soil moisture of ESA CCI, GLEAM, three sets of GLDAS datasets and three sets of reanalysis datasets are analyzed.

[0040] S4: Temporal evolution analysis of soil moisture spatial correlation. Combined with the spatial correlation analysis results of the surface soil moisture dataset obtained in step S3, several sets of the same datasets are selected and the Sen slope is used to estimate the spatial distribution characteristics of the multi-year soil moisture trend. This allows comparison of the multi-year average soil moisture trends between the multiple source datasets. For each dataset pair, the proportion of all grid cells with the same trend direction is calculated.

[0041] As an exemplary embodiment, step S4 of the present invention is specifically as follows:

[0042] The spatial correlation of the annual soil moisture time series was analyzed by calculating the Spearman correlation coefficient, and then the temporal evolution of the spatial correlation between different soil moisture datasets was obtained by comparison. Finally, the spatial variation trend of the multi-year average soil moisture at each grid point was analyzed using the Sen slope estimation.

[0043] S5: Analysis of the physical mechanism of soil moisture changes; Combined with the multi-year average soil moisture trend obtained in step S4, the same method is used to calculate the observed data and climate model data to obtain the spatial distribution characteristics of the annual aridity index AI, annual precipitation PRCP, annual potential evapotranspiration PET, and annual surface air temperature SAT. Different climate zones are divided globally to explore the changes in soil moisture in different climate regions;

[0044] It should be noted that step S5 of the present invention is specifically as follows:

[0045] S51: Using the same method as in step S4, the annual aridity index (AI), annual precipitation (PRCP), annual potential evapotranspiration (PET), and annual surface air temperature (SAT) are calculated using the CRU data and the multi-model mean (CMIP6-EM) data of 18 climate models in CMIP6;

[0046] S52: Divide the world into different climate zones based on aridity indices. Compare the relationships between soil moisture variables from various soil moisture datasets (ESACCI, GLEAM, GLDAS, reanalysis data, and CMIP6) and the climatological mean of the aridity index over the reference period (1980–2020). Use the aridity index from each dataset as a scaling factor for changes in the Sen slope and the change in the dryness region in the annual mean soil moisture.

[0047] S6: Analysis of climate drivers of soil moisture changes; Combined with the soil moisture change characteristics of different climate regions obtained in step S5, explore the impact of different climate drivers on soil moisture changes. Combined with the sea surface temperature dataset and surface soil moisture dataset obtained in step S1, through maximum covariance analysis, explore the first and second dominant modes and further analyze the impact of sea surface temperature on the spatiotemporal changes of soil moisture.

[0048] It should be noted that in step S6, geopotential height and wind fields from 1980 to 2020 were analyzed using ERA5 reanalysis data and multi-model averages of 18 CMIP6 climate models. A Mann-Kendall trend test was used to identify regions where geopotential height showed an increasing trend at the 5% significance level. A maximum covariance analysis was then used to investigate the impact of sea surface temperature anomalies on soil moisture changes across all datasets. The maximum covariance analysis method was as follows:

[0049]

[0050] Where X and Y are the spatial matrices of sea surface temperature anomalies and soil moisture, respectively, and N is the time span. m and q refer to the number of grid points in the spatial matrices of sea surface temperature anomalies and soil moisture, respectively. x,y is the covariance matrix. U and V are the spatial patterns of sea surface temperature anomalies and soil moisture, respectively. PCx,m and PC y,q are the time series of unit m in sea surface temperature anomalies and unit q in soil moisture.

[0051] In order to better explain the above steps, the present invention provides a specific embodiment as follows:

[0052] (1) Collection of basic data;

[0053] In this implementation case, when analyzing the spatiotemporal variations in soil moisture, the ESA CCI (soil moisture dataset) was collected, using annual average soil moisture data to ensure temporal consistency and spatial coverage of the data. The GLEAM (Global Land Evaporation Amsterdam Model) data product was collected to consider the impact of soil moisture on evaporation processes. The GLDAS-CLM, GLDAS-VIC, and GLDAS-NOAH soil moisture datasets from the GLDAS (Global Land Data Assimilation System) were collected for further verification and comparative analysis. Monthly precipitation, surface air temperature, and potential evapotranspiration data from the CRU TS4.06 (Climate Research Unit Time Series Dataset) were collected to assess the driving effect of climate factors on soil moisture evolution. Monthly meteorological reanalysis data from the ERA5 reanalysis dataset, including key meteorological variables such as geopotential height, meridional wind, zonal wind, and soil moisture, were collected to analyze the impact of large-scale climate dynamics on soil moisture changes. The COBE SST (sea surface temperature dataset) was collected for CMIP 6 related analysis. The selected data information is shown in Table 1.

[0054] Table 1 Main data information

[0055]

[0056] (2) Comparison of temporal correlations of surface soil moisture datasets;

[0057] In this implementation case, in order to reduce the differences between the soil moisture datasets in data preprocessing, the spatial resolution of each dataset was unified, and the soil moisture data of all grid points were re-gridded to a spatial resolution of 1°×1° using the bilinear interpolation method. In order to eliminate the impact of long-term trends on soil moisture data analysis, the linear detrending method was used to process the annual average soil moisture data from 1980 to 2020: by performing linear regression on the annual average soil moisture value of each grid point, the linear trend of the grid point during the study period was calculated, and the calculated linear trend was subtracted from the original soil moisture data to obtain the soil moisture anomaly after removing the long-term trend. The least squares fitting formula used for linear regression is as follows:

[0058]

[0059] Where b is the slope, n is the total number of years in the study period, and x i is the i-th year of the study period, is the mean value of the years in the study period, y i is the value of the research variable in the i-th year of the research period, is the mean of the research variable during the research period;

[0060] Subsequently, the Spearman correlation coefficient was calculated to quantify the temporal correlation between all selected grid cells in different data sets, and the median of the temporal correlation was selected as the cross-correlation value. The correlation calculation formula is as follows:

[0061]

[0062] Where x and y are two time series of the same length, n is the length of the series x and y, x i and y i represents the value of sequences x and y in year i, and cor(x,y) is the cross-correlation between sequences x and y.

[0063] (3) Spatial correlation distribution of surface soil moisture dataset;

[0064] In this implementation case, based on the cross-correlation comparison results of the surface soil moisture dataset from 1980 to 2020 obtained in implementation case (2), four major soil moisture datasets were used: ESA CCI, GLEAM, the ensemble mean of three GLDAS datasets (GLDAS-EM), and the ensemble mean of three reanalysis datasets (Reanalysis-EM). By calculating the Spearman correlation coefficient of multi-year soil moisture at each spatial grid point between different datasets, the spatial distribution characteristics of different datasets were obtained.

[0065] The results show significant positive correlations among the four soil moisture datasets: GLEAM and the ensemble mean of three reanalysis datasets, the ensemble mean of GLEAM and three global land surface model datasets, and the ensemble mean of the three reanalysis datasets and the ensemble mean of the three global land surface model datasets. Positive correlations also exist between the ESA CCI and GLEAM, the ESACCI and the average of the three global land surface model datasets, and the ESA CCI and the ensemble mean of the three reanalysis datasets. The highest correlations are found between the ensemble mean of the three reanalysis datasets and the other datasets, reaching a significance level of 0.05 for 82.78% of the grids, particularly over Northeast Asia, Australia, southern South America, and North America.

[0066] (4) Temporal evolution analysis of soil moisture spatial correlation;

[0067] In this implementation case, the temporal evolution trend of soil moisture spatial correlation is further analyzed based on the spatial correlation distribution characteristics of the 1980-2020 dataset obtained in implementation case (3). The spatial correlation between ESA CCI and other datasets has increased significantly in the past 10 years, which is mainly attributed to the improvement of ESA CCI product quality ( Figure 3 a). The spatial correlation between GLDAS and other datasets remains stable in time ( Figure 3 b) The interannual variation in spatial correlation between the ensemble mean of GLEAM, three sets of reanalysis data, and the ensemble mean of 18 climate model data from CMIP6 is small, indicating that the internal climate variability has a weaker impact on these datasets ( Figure 3 c) Subsequently, the temporal and spatial evolution characteristics of surface soil moisture were compared based on the temporal evolution trend of soil moisture spatial correlation from 1980 to 2020. The trend of annual mean soil moisture at each grid point was detected using the Sen slope estimation method.

[0068] Most datasets show a widespread decrease in surface soil moisture globally, but the ESA CCI dataset shows an increase. The GLEAM and ensemble mean of three reanalysis datasets show a more widespread and severe global trend in soil moisture decrease, with the largest decrease in soil moisture. Over East Asia and Australia, the ensemble mean of three global land surface model datasets and the GLEAM dataset show significant differences in soil moisture trends. The ensemble mean of 18 CMIP6 climate model datasets (CMIP6-EM) shows the smallest change of all datasets.

[0069] Spatial cross-correlation analysis of soil moisture trends across different datasets revealed low spatial correlations between the ESA CCI and the ensemble mean of 18 CMIP6 climate model datasets (CMIP6-EM) and other datasets. Spatial correlations between GLEAM, the ensemble mean of three reanalysis datasets, and the ensemble mean of three global land surface model datasets were high, with significance levels below 0.01. However, most datasets consistently exhibited significant soil moisture drying trends over regions such as North America, Europe, Northeast Asia, North Africa, and the Arabian Peninsula. This case study also calculated the proportion of grid points with consistent trend directions within each dataset pair. The GLEAM dataset showed the highest trend consistency with the other datasets, with an average of 60% of grid points sharing the same trend direction with the other datasets. The ESA CCI had the lowest proportion of grid points with the same trend sign as the other datasets. Compared to the other datasets, the ESACCI soil moisture trend diverged in over 50% of the regions.

[0070] (5) Analysis of the physical mechanism of soil moisture changes;

[0071] In this implementation case, the comparison results of spatiotemporal changes of surface soil moisture were obtained based on implementation case (4). Using the same method, the annual aridity index, annual precipitation, annual potential evapotranspiration, and annual surface air temperature were obtained by calculating the ensemble mean (CMIP6-EM) of CRU data and 18 climate model data of CMIP6. The results show that the annual potential evapotranspiration shows an increasing trend in most regions. Since the change of annual precipitation shows significant spatial differences in different datasets, especially in Europe, northeastern Asia, North America, central and southern South America, and north of the equator in Africa, the change of the aridity index (calculated as annual precipitation / annual potential evapotranspiration) can better reflect the dry or wet changes of the climate. The spatial patterns of the aridity index trend and the comprehensive soil moisture trend are very similar, which indicates that the ensemble mean of multiple soil moisture datasets can more reasonably represent the actual situation than a single dataset. In western North America, Europe, Northeast Asia, and western and eastern Australia, the significant soil moisture drying trend is consistent with the result of the decreasing aridity index. The annual precipitation in these regions also shows a decreasing trend. The Sahara region experienced an increase in precipitation between 1980 and 2015, with this increase being a key driver of soil moisture increases. Significant increases in the Sahara dryness index and annual precipitation confirm the results of CMIP6.

[0072] Subsequently, we analyzed soil moisture changes in different climate regions, dividing the globe into different climate zones. The global climate zones were divided according to the Aridity Index (AI). Soil moisture changes from various soil moisture datasets (ESA CCI, GLEAM, GLDAS, reanalysis data, and CMIP6) were compared with multi-year averages of the Aridity Index reference period (1980-2020). The Sen slope and the proportion of dry areas in the annual mean soil moisture were calculated as a function of the average of the Aridity Index reference period for each dataset.

[0073] The relationship between soil moisture changes and dryness index values ​​presents a V-shaped curve ( Figure 4 ). In particular, in the GLEAM, reanalysis data and CMIP6 datasets, the lowest point of the V-shaped structure (i.e., the most significant dryness) occurs in the area with aridity index values ​​between 0.8 and 1.2, mainly in the humid-arid transition zone. In the GLEAM dataset, the dryness peak of the V-shaped structure appears earlier than in the reanalysis data and CMIP6 datasets. In the humid-arid transition zone (dryness index values ​​between 0.8 and 1.2), the changes in potential evapotranspiration and precipitation are more significant. The increase in annual potential evapotranspiration in these areas is greater than the increase in annual precipitation, but due to enhanced soil evaporation and vegetation transpiration, soil moisture still tends to be dry. Global warming may cause the evapotranspiration mechanism in the humid-arid transition zone to shift to a water-limited mode, further intensifying the drying trend in the region.

[0074] (6) Analysis of climate-driven changes in soil moisture;

[0075] In this implementation case, based on the conclusions on soil moisture changes in different climate regions obtained in implementation case (5), the impact of different climate drivers on soil moisture changes is explored. This case is based on the ERA5 reanalysis data and the ensemble mean of 18 climate model data from CMIP6 (CMIP6-EM), and analyzes the atmospheric circulation, geopotential height, and wind field from 1980 to 2020. The red (blue) areas with black dots are areas where the geopotential height shows an upward trend at the 5% significance level according to the Mann-Kendall test.

[0076] During the winter of 1980-2020, the weakening of the Aleutian Low and the presence of an anomalous anticyclone slowed westerly winds south of the Aleutian Low, reducing moisture transport to North America and leading to reduced precipitation and drier soil moisture. East Asia and North Africa are more heavily influenced by the monsoon circulation during the warm season. An anomalous high-pressure center developed in Northeast Asia from June to August, hindering the transport of warm, moist air from the Pacific to Northeast Asia. The weakening of the East Asian summer monsoon prevented moist air from reaching inland and northern Asia. During the warm season, the West African monsoon brings moist air from the tropical Atlantic to the North African Sahel, increasing precipitation in the region (represented only in the ensemble mean of the CMIP6 and three global land surface model datasets). A low-pressure center south of South Africa from December to February strengthens the low-level pressure gradient between land and sea, intensifying easterly winds from the Indian Ocean, transporting more moist air and resulting in more rainfall, thus increasing the wetting trend over southern South Africa. This process can explain the wetting trend over southern South Africa outside the CMIP6 dataset. During the austral summer, an abnormally high-pressure center over Patagonia and unusual easterly winds to the west suppress warm, moist air from the Pacific, leading to drying of soil moisture in the region. Summer precipitation in eastern Australia shows a downward trend, while central and western Australia show an upward trend: from December to February, the easterly winds from the tropical western Pacific weaken, reducing the amount of moisture transported by these winds. From June to August, anomalous cyclones over central and western Australia bring warm, moist air, triggering more precipitation and a humid climate.

[0077] Based on the comparison of the spatiotemporal variation of soil moisture obtained in implementation case (5), this case further explores the impact of sea surface temperature on the spatiotemporal variation of soil moisture. This implementation case uses the maximum covariance analysis (MCA) method to explore the impact of SST anomalies (40°S-60°N) on soil moisture variation in all data sets. The maximum covariance analysis method is as follows:

[0078]

[0079] Where X and Y are the spatial matrices of sea surface temperature anomalies and soil moisture, respectively, and N is the time span. m and q refer to the number of grid points in the spatial matrices of sea surface temperature anomalies and soil moisture, respectively. x,y is the covariance matrix. U and V are the spatial patterns of sea surface temperature anomalies and soil moisture, respectively. PC x,m and PC y,q are the time series of unit m in sea surface temperature anomalies and unit q in soil moisture.

[0080] The first dominant MCA mode (MCA1) can be considered as a global warming signal in soil moisture variations. In all datasets, there is a strong positive correlation between the temporal coefficients of MCA1 for SST and SAT (using the Spearman method), with a correlation value of approximately 0.9 ( Figure 6 ). The maximum squared fractional covariance (SFC) explained by MCA1 in the ensemble mean of 18 climate model data from CMIP6 (CMIP6-EM) is 0.96, and the minimum is 0.5 in the ESA CCI dataset. In addition, the temporal and spatial patterns between MCA1 and the corresponding long-term changes have similar temporal evolution characteristics. The long-term changes in soil moisture can be largely explained by global warming ( Figure 6 Despite regional differences, the spatial patterns of MCA1 are consistent across datasets (e.g., GLEAM, the ensemble mean of three reanalysis datasets, the ensemble mean of three global land surface model datasets, and the ensemble mean of 18 climate models from CMIP6 (CMIP6-EM)). However, there are significant differences in the spatial patterns of MCA1 between the ESA CCI and the simulated datasets. Corresponding to global warming, the spatial patterns of MCA1 across all datasets show drying of soil moisture in North America, Europe, Northeast Asia, and southern South America. Some differences are observed in the MCA1 models for southern South Africa, the Sahel, the Indian subcontinent, and Australia.

[0081] The second dominant MAC mode (MCA2) represents the influence of El Niño-Southern Oscillation on soil moisture changes. In all datasets, the time coefficient of MCA2 is There is a high correlation (about 0.9) between the SSTs in region 3.4 (120°W-170°W, 5°W-5°S), and there is a similar spatial pattern between the SSTs caused by the El Niño-Southern Oscillation and the spatial pattern of the MCA2 ( Figure 7 Unlike the consistent upward trend in MCA1, MCA2 in soil moisture exhibits an interannual variation pattern similar to the SST anomaly fluctuation caused by ENSO ( Figure 7The spatial patterns of MCA2 differ significantly across datasets, suggesting that short-term soil moisture variations induced by the El Niño-Southern Oscillation depend on the selected soil moisture dataset. These patterns are less consistent with the temporal variations of soil moisture in the corresponding datasets.

[0082] See Figure 8 , Figure 8 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically includes: a climate-driven analysis device 401 based on the spatiotemporal evolution of soil moisture, a processor 402 and a storage medium 403.

[0083] A climate-driven analysis device 401 based on the spatiotemporal evolution of soil moisture: The climate-driven analysis device 401 based on the spatiotemporal evolution of soil moisture implements the climate-driven analysis method based on the spatiotemporal evolution of soil moisture.

[0084] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the climate-driven analysis method based on the spatiotemporal evolution of soil moisture.

[0085] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the climate-driven analysis method based on the spatiotemporal evolution of soil moisture.

[0086] The beneficial effects of the present invention are:

[0087] (1) This paper proposes a method for comparing the spatiotemporal changes of global surface soil moisture based on multiple datasets, analyzes the consistency and differences between different datasets, identifies the main characteristics of the long-term trend of global soil moisture, and provides a scientific basis for monitoring and evaluating soil moisture in the context of global climate change.

[0088] (2) Through a detailed analysis of climate driving factors, this paper reveals the dominant role of these climate factors in global surface soil moisture changes and clarifies the complex relationship between climate driving factors and soil moisture changes.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A climate-driven analysis method based on the spatiotemporal evolution of soil moisture, characterized by: The method comprises the following steps: S1: Data collection: Collect multi-source soil moisture data, large-scale meteorological and environmental variable data; S2: Comparison of temporal correlations of surface soil moisture datasets: Combine the surface soil moisture datasets obtained in step S1, perform linear detrending on the annual mean surface soil moisture values, compare the different datasets in pairs, and calculate the median of the time series correlation coefficient of each dataset for all selected grid cells; S3: Spatial correlation distribution of surface soil moisture datasets; combining the correlation coefficients between different datasets obtained in step S2, select some datasets, calculate the Spearman correlation coefficients between different datasets, and obtain the spatial correlation distribution characteristics between the above datasets; S4: Temporal evolution analysis of soil moisture spatial correlation. Combined with the spatial correlation analysis results of the surface soil moisture dataset obtained in step S3, several sets of the same datasets are selected and the Sen slope is used to estimate the spatial distribution characteristics of the multi-year soil moisture trend. This allows comparison of the multi-year average soil moisture trends between the multiple source datasets. For each dataset pair, the proportion of all grid cells with the same trend direction is calculated. S5: Analysis of the physical mechanism of soil moisture changes; Combined with the multi-year average soil moisture trend obtained in step S4, the same method is used to calculate the observed data and climate model data to obtain the spatial distribution characteristics of the annual aridity index AI, annual precipitation PRCP, annual potential evapotranspiration PET, and annual surface air temperature SAT. Different climate zones are divided globally to explore the changes in soil moisture in different climate regions; S6: Analysis of climate drivers of soil moisture changes; Combined with the soil moisture change characteristics of different climate regions obtained in step S5, explore the impact of different climate drivers on soil moisture changes. Combined with the sea surface temperature dataset and surface soil moisture dataset obtained in step S1, through maximum covariance analysis, explore the first and second dominant modes and further analyze the impact of sea surface temperature on the spatiotemporal changes of soil moisture.

2. The climate-driven analysis method based on the spatiotemporal evolution of soil moisture according to claim 1, characterized in that: In step S2, when calculating the temporal correlation of the dataset, the annual mean soil moisture values ​​are first linearly detrended using the least squares method; For time series data y t , fitting linear model y t =a+bt+∈ t , where t is the time a and b are regression coefficients, ∈ t is the residual; Its trend part is y′ t =a+bt; detrended data, i.e., y=y t -y′ t , where the least squares fitting formula for the slope is as follows: Where b is the slope, n is the total number of years in the study period, and x i is the i-th year of the study period, is the mean value of the years in the study period, y i is the value of the research variable in year i during the research period, is the mean of the research variable during the research period; Subsequently, the Spearman correlation coefficients between all selected grid points and different data sets were calculated to analyze the temporal correlation; the mean value was selected as the cross-correlation value, and the calculation formula was as follows: In the formula, x and y are two time series of the same length, n is the length of the series x and y, x i and y i represents the value of sequences x and y at position i, and cor(x,y) is the cross-correlation between sequences x and y.

3. The climate-driven analysis method based on the spatiotemporal evolution of soil moisture according to claim 2, characterized in that: In step S3, the spatial distribution characteristics of the Spearman correlation coefficient of the annual mean soil moisture of ESA CCI, GLEAM, three GLDAS datasets, and three reanalysis datasets are analyzed.

4. The climate-driven analysis method based on the spatiotemporal evolution of soil moisture according to claim 3, characterized in that: Step S4 is specifically as follows: The spatial correlation of the annual soil moisture time series was analyzed by calculating the Spearman correlation coefficient, and then the temporal evolution of the spatial correlation between different soil moisture datasets was obtained by comparison. Finally, the spatial variation trend of the multi-year average soil moisture at each grid point was analyzed using the Sen slope estimation.

5. The climate-driven analysis method based on the spatiotemporal evolution of soil moisture according to claim 4, characterized in that: Step S5 is specifically as follows: S51: Using the same method as in step S4, the annual aridity index AI, annual precipitation PRCP, annual potential evapotranspiration PET, and annual surface air temperature SAT are calculated using the CRU data and the multi-model average CMIP6-EM data of 18 climate model data of CMIP6; S52: Divide the world into different climatic regions based on the aridity index; compare the relationship between the soil moisture variables of each soil moisture dataset and the climatological mean of the aridity index during the reference period, and use the aridity index of each dataset as a scaling factor for the change in Sen's slope and the change in the dryness area in the annual mean soil moisture.

6. The climate-driven analysis method based on the spatiotemporal evolution of soil moisture according to claim 5, characterized in that: In step S6, the geopotential height and wind fields during the reference period are analyzed using reanalysis data and multi-model averages of multiple climate models. The Mann-Kendall trend test is used to identify areas where the geopotential height shows an increasing trend at the 5% significance level. The maximum covariance analysis method is then used to explore the impact of sea surface temperature anomalies on soil moisture changes in all data sets. The maximum covariance analysis method is as follows: where X and Y are the spatial matrices of sea surface temperature anomalies and soil moisture, respectively; N is the time span; m and q are the number of grid points in the spatial matrices of sea surface temperature anomalies and soil moisture, respectively; C x,y is the covariance matrix; U and V are the spatial patterns of sea surface temperature anomalies and soil moisture, respectively; PC x,m and PC y,q are the time series of unit m in sea surface temperature anomalies and unit q in soil moisture.

7. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a climate-driven analysis method based on spatiotemporal evolution of soil moisture as described in any one of claims 1 to 6.

8. A climate-driven analysis device based on the spatiotemporal evolution of soil moisture, characterized by: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a climate-driven analysis method based on the spatiotemporal evolution of soil moisture as described in any one of claims 1 to 6.

Images