A four-dimensional soil moisture drought event tracking and warming signal identification method

Abstract

The present invention discloses a method for tracking four-dimensional soil moisture drought events and identifying warming signals, the method comprising the following steps: data acquisition; four-dimensional soil moisture drought event identification; four-dimensional soil moisture drought event feature quantification; human-induced warming signal detection of the temporal evolution of four-dimensional soil moisture drought features in historical periods; four-dimensional soil moisture drought future evolution prediction; and four-dimensional soil moisture drought event driving factors. The present invention has the beneficial effects of exploring the future spatiotemporal evolution characteristics of four-dimensional soil moisture drought, analyzing the driving factors and physical mechanisms of four-dimensional soil moisture drought, and providing a strong scientific basis for responding to global climate change, ensuring water resource security, and formulating disaster prevention and mitigation strategies.

Description

Technical Field

[0001] The present invention relates to a four-dimensional soil moisture drought event tracking and warming signal recognition method. Background Art

[0002] Soil moisture drought, characterized by soil moisture deficits, is one of the most severe natural disasters, adversely affecting water resources, ecosystems, plant production, and crop yields. Influenced by climate, vegetation, topography, soil properties, and land use, soil moisture variations vary across soil depth at the regional scale. Water stored in deep soil layers can be directly used for vegetation growth. Therefore, water deficits in deep soil layers can cause greater damage to plants and agricultural production than those in surface soils. Given the high spatial heterogeneity and strong variability in root depth among vegetation, understanding the global vertical structure of soil moisture drought is crucial for more accurately monitoring and assessing its adverse impacts on ecosystems. Against the backdrop of global warming, greenhouse gas emissions are exacerbating soil drought. Currently, there are no detailed studies examining the continuous evolution of soil moisture drought in four dimensions (longitude, latitude, depth, and time), and whether it is influenced by climate change caused by anthropogenic warming remains unclear. Summary of the Invention

[0003] The purpose of the present invention is to propose a four-dimensional soil moisture drought event tracking and warming signal identification method to solve the technical problem that the spatiotemporal evolution of the vertical structure of global soil moisture drought has not been carefully studied.

[0004] The present invention provides a four-dimensional soil moisture drought event tracking and warming signal identification method, comprising the following steps:

[0005] Step S1: Data collection: Collect global climate data;

[0006] Step S2: Identification of four-dimensional soil moisture drought events: Based on the data obtained in step S1, the 10th percentile of the climatological growing season soil moisture value of each grid point is calculated as the drought threshold, and grid points below the threshold are identified as drought grid points. Continuous four-dimensional soil moisture drought events are identified using the Lagrangian four-dimensional tracking method.

[0007] Step S3: Quantification of four-dimensional soil moisture drought event characteristics: Combined with the four-dimensional drought events obtained in step S2, soil moisture drought event characteristics are calculated, and soil moisture drought events are divided into surface drought and deep drought;

[0008] Step S4: Detecting anthropogenic warming signals from the temporal evolution of the four-dimensional soil moisture drought characteristics over the historical period; combining the temporal and intensity characteristics of the two types of drought events obtained in step S3, using the correlation coefficient method and the optimal fingerprint method, detect anthropogenic warming signals from drought characteristic indicators calculated based on reanalysis data and historical climate experiments of climate models;

[0009] Step S5: Predicting the future evolution of four-dimensional soil moisture drought; combining the time and intensity characteristics of the two types of drought events obtained in step S3, combined with the historical climate experiments of the climate model and the future socio-economic development path experiments, to obtain the spatiotemporal evolution characteristics of drought events in the future period; combining the global land cover data obtained in step S1, to obtain the spatiotemporal evolution characteristics of drought events under different vegetation types in the future period;

[0010] Step S6: Four-dimensional soil moisture drought event driving factors; combining the time and intensity characteristics of the two types of drought events obtained in step S3 with the near-surface air temperature, precipitation, and leaf area index data obtained in step S1, the two types of drought events are attributed to the three types of antecedent driving factors, and the contribution of different driving factors to drought events is quantified.

[0011] A storage medium stores instructions and data for implementing a four-dimensional soil moisture drought event tracking and warming signal recognition method.

[0012] A four-dimensional soil moisture drought event tracking and warming signal identification device includes: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a four-dimensional soil moisture drought event tracking and warming signal identification method.

[0013] The beneficial effects provided by the present invention are:

[0014] (1) The present invention provides a method for identifying four-dimensional soil moisture drought, which helps the academic community deepen its understanding of this new type of soil drought. The present invention divides four-dimensional soil drought into two different types, quantifies the spatiotemporal characteristic indicators of four-dimensional soil drought, and clarifies the changes and impacts of four-dimensional soil moisture drought in the four-dimensional spatiotemporal domain of longitude, latitude, depth, and time. The present invention quantifies the differences in the spatiotemporal evolution of two types of four-dimensional soil drought in the future period under different land cover types, providing a scientific basis for a deeper understanding of the dynamic evolution process of drought events.

[0015] (2) The present invention quantitatively detects anthropogenic warming signals in the spatiotemporal evolution of four-dimensional soil moisture drought characteristics, identifies the impact of anthropogenic warming on the changes in two four-dimensional drought events, and estimates the changes in four-dimensional drought events in the future, providing theoretical support for decision makers to make decisions on adapting to and mitigating the effects of climate change on soil moisture drought.

[0016] (3) This paper analyzes the contribution of driving factors to four-dimensional soil moisture drought events, clarifies the contribution of different driving factors to four-dimensional soil moisture drought in history and future periods, quantifies the dominant driving factors of two types of four-dimensional soil moisture drought, and provides theoretical support for the attribution of four-dimensional soil moisture drought. BRIEF DESCRIPTION OF THE DRAWINGS

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

[0018] Figure 2 This is an example map of surface drought and deep drought defined using the Lagrangian four-dimensional tracking framework;

[0019] Figure 3 It is the spatial distribution map of the centroids of global surface drought and deep drought in historical period;

[0020] Figure 4 The duration and intensity of global surface drought and deep drought have changed over the historical period;

[0021] Figure 5 It is a probability density map of the duration and intensity trend of global surface drought and deep drought in historical period;

[0022] Figure 6 It is the spatial and temporal evolution of the duration of global surface drought and deep drought events in the future period;

[0023] Figure 7 is the spatiotemporal evolution of the intensity of global surface drought and deep drought events in the future period;

[0024] Figure 8 The differences in duration and intensity of global deep and surface droughts under different land cover conditions;

[0025] Figure 9 is the contribution of different driving factors to the duration of surface drought and deep drought events;

[0026] Figure 10 is the contribution of different driving factors to the intensity of surface drought and deep drought events;

[0027] Figure 11 are the time series of the three driving factors during the growing season under different scenarios;

[0028] Figure 12 It is a working diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0029] 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.

[0030] Before formally explaining the present invention, the scheme of the present invention is first generally explained for easy understanding.

[0031] Please refer to Figure 1 , Figure 1 It is a schematic flow diagram of the method of the present invention.

[0032] The present invention provides a four-dimensional soil moisture drought event tracking and warming signal identification method, comprising the following steps:

[0033] Step S1: Data collection: Collect global climate data;

[0034] Specifically, the present invention collects soil moisture, near-surface air temperature and precipitation data from the fifth generation land surface reanalysis dataset (ERA5-Land) of the European Centre for Medium-Range Weather Forecasts, the Global Atmosphere and Land Surface Reanalysis Product (CRA-Land) of the China Meteorological Administration, and the Global Land Data Assimilation System Model (GLDAS-Noah); collects leaf area index data from the global long-term leaf area index GLOBMAP data; collects precipitation data from the Global Soil Moisture Project (GSWP-3); collects soil moisture, near-surface air temperature, precipitation and leaf area index data from the sixth phase of the Coupled Comparison Program (CMIP6); and collects global land cover data from the Moderate Resolution Imaging Spectroradiometer Land Cover Climate Simulation Grid (MCD12C1);

[0035] Step S2: Identification of four-dimensional soil moisture drought events: Based on the data obtained in step S1, the 10th percentile of the climatological growing season soil moisture value of each grid point is calculated as the drought threshold, and grid points below the threshold are identified as drought grid points. Continuous four-dimensional soil moisture drought events are identified using the Lagrangian four-dimensional tracking method.

[0036] It should be noted that in step S2, the soil moisture data is preprocessed: according to the GSWP-3 precipitation data, the grid points in Antarctica, Greenland and desert areas with annual precipitation less than 100 mm are removed; the soil moisture values ​​in the growing season (May to September in the Northern Hemisphere and November to March in the Southern Hemisphere) are used; the soil moisture value unit is changed from mass unit (kg·m -2 ) is converted to volume units (m 3 ·m 3 ); linearly interpolate multi-layer soil moisture values ​​to a common vertical profile (ranges 0-7, 7-28, 28-100, and 100-289 cm); transform datasets at different time scales to daily scales; use second-order conservative remapping to interpolate all datasets to a 1° × 1° resolution; and use an 11-day smoothing average to eliminate abnormal series fluctuations;

[0037] The spatiotemporal evolution of soil moisture drought was simulated in four dimensions using a Lagrangian approach as follows: for each grid point in a single layer, a drought grid point was defined when the soil moisture value was below the 10th percentile of the growing season during the climatological period (1981-2010); adjacent drought grid points in a single layer were merged into a two-dimensional spatially continuous drought patch; for each time step, all adjacent two-dimensional drought patches within 26 adjacent regions (nine in the upper layer, excluding eight central grid points, and nine in the lower layer) were connected into a single, continuous three-dimensional drought patch; any spatially continuous three-dimensional drought patches with an overlap greater than 50% in consecutive time steps were merged into a single four-dimensional (longitude-latitude-depth-time) drought event; and four-dimensional drought events lasted at least three days and had a projected area greater than 100,000 square kilometers.

[0038] Step S3: Quantification of four-dimensional soil moisture drought event characteristics: Combined with the four-dimensional drought events obtained in step S2, soil moisture drought event characteristics are calculated, and soil moisture drought events are divided into surface drought and deep drought;

[0039] It should be noted that in step S3, the temporal and spatial characteristics of the four-dimensional soil moisture drought event are duration and intensity, respectively. Duration is the duration of the four-dimensional soil moisture drought event, and intensity is the weighted average of the grid area and depth of the soil moisture deficit (threshold minus soil moisture value) of all grid points in the four-dimensional drought event, as follows:

[0040]

[0041] Where i is the grid point, t is the time, D is the duration, l is the soil layer, d is the soil depth, s is the grid area, thrse is the drought threshold, sm is the soil water value, and cluster is the drought event;

[0042] These four-dimensional drought events were categorized into three types: surface drought with a heavy top-type inverted iceberg structure, defined as when the surface soil water deficit area exceeds the deep soil water deficit area for more than 60% of the total duration; deep drought with a heavy bottom-type iceberg structure, defined as when the deep soil water deficit area exceeds the surface soil water deficit area for more than 60% of the total duration; and when neither of these conditions occurs, the drought is classified as a columnar structure. For this classification, the top two layers (0-7 cm and 7-28 cm) were selected as the surface soil layer, and the third layer (28-100 cm) was selected as the deep soil layer.

[0043] Step S4: Detecting anthropogenic warming signals from the temporal evolution of the four-dimensional soil moisture drought characteristics over the historical period; combining the temporal and intensity characteristics of the two types of drought events obtained in step S3, using the correlation coefficient method and the optimal fingerprint method, detect anthropogenic warming signals from drought characteristic indicators calculated based on reanalysis data and historical climate experiments of climate models;

[0044] It should be noted that in step S4, the anthropogenic warming signal of the four-dimensional soil moisture drought event is detected based on the correlation coefficient method, and the Spearman correlation coefficient between the changes in the global surface and deep drought characteristic sequences based on the reanalysis (average of three reanalysis data sets) and the CMIP6 historical experiment simulation (multi-model average) is calculated; the CMIP6 pre-industrial simulation experiment is divided into multiple blocks of years (40 years) corresponding to the study period, and random sampling is performed 2000 times to generate a large number of samples, and the Spearman correlation coefficient between the samples and the multi-model average of the historical experiment simulation is calculated; and the correlation coefficient between the drought event characteristic sequence of the reanalysis data and the historical experiment simulation is analyzed to see whether it exceeds the 95th or 99th percentile of the correlation coefficient set between the multiple blocks of the historical simulation experiment and the pre-industrial simulation experiment; if the correlation between the reanalysis and the historical simulation experiment is greater than the 95th or 99th percentile of the correlation between the historical simulation experiment and the 2000 repeated 40-year pre-industrial simulation experiment blocks, it indicates that the changes in the surface and deep drought characteristics have external forcing signals and can be attributed to anthropogenic warming;

[0045] The detection of anthropogenic warming signals of four-dimensional soil moisture drought events based on the optimal fingerprint method is calculated as follows:

[0046] Y=(X-α)β+ε

[0047] Where Y is the global mean characteristic change of surface and deep drought based on reanalysis, x is the global mean characteristic change of surface and deep drought based on model simulation, α is the sampling uncertainty, β is the scaling factor, and ε is the internal climate variability estimated from pre-industrial simulation experiments. To reduce the noise caused by interannual variability, the characteristic series of surface and deep drought are averaged over consecutive three-year intervals. If the scaling factor and its 90% confidence interval are both greater than zero, the external forcing signal is considered detectable at the 5% significance level, that is, the anthropogenic warming signal can be detected.

[0048] Step S5: Predicting the future evolution of four-dimensional soil moisture drought; combining the time and intensity characteristics of the two types of drought events obtained in step S3, combined with the historical climate experiments of the climate model and the future socio-economic development path experiments, to obtain the spatiotemporal evolution characteristics of drought events in the future period; combining the global land cover data obtained in step S1, to obtain the spatiotemporal evolution characteristics of drought events under different vegetation types in the future period;

[0049] It should be noted that in step S5, the spatiotemporal evolution of global surface and deep drought characteristics in the future period is calculated using the CMIP6 historical simulation experiments and representative concentration pathways and shared socioeconomic pathways (SSPs) of future scenarios, including the SSP245 and SSP585 scenarios;

[0050] Combined with the land cover dataset, the changes in surface and deep drought characteristics of different vegetation types were estimated. Deep-rooted vegetation cover categories (i.e., forests, shrubs, and savannahs) were selected, and the area-weighted average of the differences in deep and surface drought characteristics (duration and intensity) in different vegetation areas was calculated.

[0051] Step S6: Four-dimensional soil moisture drought event driving factors; combining the time and intensity characteristics of the two types of drought events obtained in step S3 with the near-surface air temperature, precipitation, and leaf area index data obtained in step S1, the two types of drought events are attributed to the three types of antecedent driving factors, and the contribution of different driving factors to drought events is quantified.

[0052] It should be noted that in step S6, surface and deep drought events are attributed to different driving factors, such as extreme heat (TEM), precipitation deficit (PRE), and vegetation greening (high leaf area index; LAI). The average values ​​of the three variables in the 30 days before the drought event are calculated, and their severity index (SI) is calculated as follows:

[0053] SI v =(v i,j -μ j ) / σ j

[0054] Where i represents year, j represents month, v represents driving factor, μ and σ represent the climate mean and standard deviation of monthly variables in the climate state (historical period: 1981-2010, future period: 2061-2100), respectively.

[0055] S62: Define extreme heat as SI TEM >0.8, precipitation deficit is defined as SI PRE <-0.8, vegetation greening is defined as SI LAI >0.8. If a drought event and an extreme event that occurred in the previous 30 days occurred at the same grid point, the drought event at that grid point was attributed to this type of driving factor, and the normalized contributions of different driving factors to the history and intensity of the two types of drought events were calculated.

[0056] As an implementation case, the present invention further describes the present invention by taking the global scope from 1981 to 2020 as an example. The implementation case is used to illustrate the present invention, but this case is not used to limit the scope of application of the present invention. It is also applicable to different regions and other time periods.

[0057] The implementation flow chart of the four-dimensional soil moisture drought event tracking and warming signal identification method of the present invention is as follows: Figure 1 The specific steps are as follows:

[0058] (1) Collection of test data;

[0059] In this implementation case, the hourly soil moisture (unit: m2) with a spatial resolution of 0.1°×0.1° from 1981 to 2020 was collected from the fifth generation land surface reanalysis dataset of the European Centre for Medium-Range Weather Forecasts (ERA5-Land). 3 ·m -3 ), depths of 0-7cm, 7-28cm, 28-100cm, and 100-289cm; as well as near-surface air temperature and precipitation data;

[0060] The 3-hour soil moisture data (unit: m2) with a spatial resolution of 34 km × 34 km were collected from the China Meteorological Administration's Global Atmosphere and Land Surface Reanalysis product (CRA-Land). 3 ·m -3 ), depths of 0-10cm, 10-40cm, 40-100cm and 100-200cm; as well as near-surface air temperature and precipitation data;

[0061] The 3-hourly soil moisture content (unit: kg·m ) at a spatial resolution of 34 km × 34 km was collected from the Global Land Data Assimilation System (GLDAS-Noah) model version 2.0 for 1981–2014 and version 2.1 for 2000–2020. -2 ), depths of 0-10cm, 10-40cm, 40-100cm and 100-200cm; as well as near-surface air temperature and precipitation data;

[0062] The 8km×8km leaf area index data of the global long-term leaf area index GLOBMAP data were collected with a temporal resolution of half a month during 1981-2000 and a temporal resolution of 8 days during 2001-2020;

[0063] Collect precipitation data from the Global Soil Moisture Project (GSWP-3); collect global land cover data with a spatial resolution of 0.05°×0.05° from 2001 to 2022 from the Moderate Resolution Imaging Spectroradiometer Land Cover Climate Simulation Grid (MCD12C1);

[0064] Experimental data from the sixth phase of the International Coupled Comparison Program (CMIP6) were also collected. The experimental scenarios included historical climate model experiments, future socioeconomic development path simulation experiments, and pre-industrial simulation experiments. The variables were soil moisture, near-surface air temperature, precipitation, and leaf area index. The time series were 1981-2020 and 2021-2100, as shown in Table 1 below.

[0065]

[0066] (2) Four-dimensional soil moisture drought event identification;

[0067] This implementation case is global in scope. Based on the preprocessing of soil water values ​​in three layers of three global reanalysis datasets from 1981 to 2020, the 10th percentile threshold of soil water values ​​for the growing season of each grid point (1981-2010) is calculated. Grid points below this threshold are identified as drought grid points. The growing season is defined as May to September in the Northern Hemisphere and November to March in the Southern Hemisphere. A four-dimensional continuous drought event is obtained using the Lagrangian method.

[0068] First, the soil moisture data were preprocessed: based on the GSWP-3 precipitation data, the grid points in Antarctica, Greenland, and desert areas with annual precipitation less than 100 mm were removed; the soil moisture values ​​during the growing season (May to September in the Northern Hemisphere and November to March in the Southern Hemisphere) were used; the units of soil moisture values ​​were changed from mass units (kg·m -2 ) is converted to volume units (m 3 ·m 3 ); linearly interpolate multi-layer soil moisture values ​​to a common vertical profile (ranges 0-7, 7-28, 28-100, and 100-289 cm); transform datasets at different time scales to daily scales; use second-order conservative remapping to interpolate all datasets to a 1° × 1° resolution; and use an 11-day smoothing average to eliminate abnormal series fluctuations;

[0069] The Lagrangian method is used to simulate the spatiotemporal evolution of soil moisture drought in four dimensions, as follows: for each grid point in a single layer, when the soil moisture value is lower than the 10th percentile of the growing season during the climatological period (1981-2010), the grid point is defined as a drought grid point; adjacent drought grid points in a single layer are merged into two-dimensional spatially continuous drought patches; for each time step, all adjacent two-dimensional drought patches in 26 adjacent regions (9 in the upper layer, 8 excluding the central grid point in this layer, and 9 in the lower layer) are connected into an independent continuous three-dimensional drought patch; any spatially continuous three-dimensional drought patches are merged into a single four-dimensional (longitude-latitude-depth-time) drought event if the overlap rate in consecutive time steps is greater than 50%; in this embodiment, the four-dimensional drought event lasts at least 3 days and has a projected area greater than 100,000 square kilometers.

[0070] In this implementation case, the Lagrangian four-dimensional framework is used to determine the deep drought and surface drought. Figure 2 As shown, during the 2012 local growing season, a deep drought event occurred in Central Asia. During 62% of the duration, the drought area of ​​the deep soil was larger than that of the surface soil. During the 2017 local growing season, a surface drought event occurred in Australia. During 86% of the duration, the drought area of ​​the surface soil was larger than that of the deep soil.

[0071] (3) Quantification of four-dimensional spatiotemporal characteristics of soil moisture drought;

[0072] This case study uses three reanalysis data sets and model simulation data from the sixth phase of the Coupled Comparison Program (CMIP6) to quantify the spatiotemporal characteristics of four-dimensional soil moisture droughts. The temporal and spatial characteristics of four-dimensional soil moisture drought events are duration and intensity, respectively. Duration is the duration of the four-dimensional soil moisture drought event, and intensity is the weighted average of the grid area and depth of the soil moisture deficit (threshold minus soil water value) at all grid points in the four-dimensional drought event. The specific characteristics are as follows:

[0073]

[0074] Where i is the grid point, t is the time, D is the duration, l is the soil layer, d is the soil depth, s is the grid area, thrse is the drought threshold, sm is the soil water value, and cluster is the drought event;

[0075] These four-dimensional drought events were categorized into three types: surface drought with a heavy top-type inverted iceberg structure, defined as when the surface soil water deficit area exceeds the deep soil water deficit area for more than 60% of the total duration; deep drought with a heavy bottom-type iceberg structure, defined as when the deep soil water deficit area exceeds the surface soil water deficit area for more than 60% of the total duration; and when neither of these conditions occurs, the drought is classified as a columnar structure. For this classification, the top two layers (0-7 cm and 7-28 cm) were selected as the surface soil layer, and the third layer (28-100 cm) was selected as the deep soil layer.

[0076] During the period 1981-2020, Figure 3As shown, 6,065 four-dimensional continuous drought events were identified and categorized into surface droughts and deep droughts based on whether they exhibited a top-heavy inverted iceberg structure or a bottom-heavy iceberg structure. Surface droughts are characterized by a larger surface soil moisture deficit and a smaller deep drought, resembling an inverted iceberg. Deep droughts, on the other hand, have a larger deep soil moisture deficit than surface droughts, resembling an iceberg. Globally, 3,557 surface droughts occurred, accounting for 58.7% of all drought events, while 1,655 deep droughts accounted for 27.3%. Deep droughts are more persistent and intense than surface droughts, particularly in the Amazon Basin, Africa, North America, East Asia, and high latitudes of the Northern Hemisphere.

[0077] (4) Detection and attribution of four-dimensional soil moisture drought indicators;

[0078] This implementation case calculates the duration and intensity of global surface and deep droughts each year based on the reanalysis (average of three reanalysis data sets) and the historical simulations of the CMIP6 phase 6 of the International Coupled Comparison Program (multi-model averages), and obtains the indicator series of global surface and deep drought duration and intensity ( Figure 4 The duration and intensity of surface droughts showed a significant (p < 0.01) upward trend from 1981 to 2020, while the duration and intensity of deep droughts increased much faster than those of surface droughts. The trends in duration and intensity of surface and deep droughts based on the reanalysis data exceeded the range of 2.5% to 97.5% of the trends in 2000 overlapping 40-year periods based on pre-industrial control simulations ( Figure 5 ), indicating that the increase in surface and deep drought characteristics cannot be fully explained by internal climate variability. When anthropogenic influences are taken into account, the CMIP6 historical simulations well capture the upward trends in characteristics of both types of drought, as well as the faster growth of deep drought.

[0079] Two detection and attribution methods (correlation coefficient method and optimal fingerprint method) are used to quantitatively assess the impact of anthropogenic warming on the changes in two types of four-dimensional drought events around the world. Based on the correlation coefficient method, the Spearman correlation coefficients are calculated between the changes in the duration and intensity characteristics of annual global surface and deep drought based on reanalysis (average of three reanalysis data sets) and the multi-model average of the CMIP6 historical simulation experiment of the sixth phase of the International Coupled Comparison Project (CMIP6). At the same time, the Spearman correlation coefficients ( ) are calculated between the multi-model average of the CMIP6 historical simulation experiment and 2000 overlapping 40-year time blocks of the two types of drought indicator series. Figure 4In this case study, the time series of two characteristic indicators of surface drought under the historical simulation experiment conditions were significantly correlated with the time series based on the reanalysis dataset (p < 0.05). The correlation between deep drought between the historical simulation experiment and the reanalysis was greater than the 99th percentile of all correlations between the historical simulation experiment and the pre-industrial simulation experiment ( Figure 4 ), confirming that anthropogenic emissions lead to increased drought across the entire vertical soil profile.

[0080] The regularized optimal fingerprint method is used to validate the results based on the correlation coefficient method and quantify the differences between the reanalysis and historical simulation experiments. The specific calculation is as follows:

[0081] Y=(X-α)β+ε

[0082] Where Y is the global mean change in surface and deep drought characteristics based on reanalysis, X is the global mean change in surface and deep drought characteristics based on model simulations, α is the sampling uncertainty, β is the scaling factor, and ε is the internal climate variability estimated from pre-industrial simulation experiments. To reduce noise caused by interannual variability, the surface and deep drought characteristic series are averaged over three-year intervals. If the scaling factor and its 90% confidence interval are both greater than zero, the external forcing signal is considered detectable at the 5% significance level, meaning that an anthropogenic warming signal is detectable.

[0083] like Figure 4 As shown in the right column, the influence of historical external forcing is detectable in the increases in both the duration and intensity of drought events in both categories. This confirms the correlation coefficient-based results, which indicate that anthropogenic emissions drive temporal changes in both drought characteristics at least at a 90% (5% to 95%) confidence level. With the exception of surface drought duration, the scaling factors for both drought characteristics are close to 1, indicating that the model accurately reproduces the changes in most drought characteristics based on the reanalysis.

[0084] This case study shows that anthropogenic climate change is very likely (90%) responsible for increases in the duration and intensity of global surface and deep droughts between 1981 and 2020. The strong external signal (primarily anthropogenic emissions) in the temporal changes of deep droughts further suggests that anthropogenic influences on global soil moisture droughts have penetrated deeper into the soil layer below the surface.

[0085] (5) Projection of future evolution of four-dimensional soil moisture drought;

[0086] In this implementation case, the spatial and temporal evolution of global surface and deep drought characteristics in the future period is projected using CMIP6 historical simulation experiments and representative concentration pathways and shared socioeconomic pathways (SSPs), including the SSP245 and SSP585 scenarios. Figure 6As shown in the SSP245 scenario, compared with the historical period (1981-2020), the projected increase in surface drought duration in the future period (2061-2100) mainly occurs in the middle and high latitudes of the Northern Hemisphere (above 30°N), and the intensity of surface drought in these regions is expected to increase by 335.9% ( Figure 7 In contrast, deeper droughts are projected to be more persistent (+154.9%) and more intense (+477.3%) over land globally, particularly in the Amazon basin, southern Africa, Australia, southeastern North America, the Middle East, and mid- and high-latitudes of the Northern Hemisphere ( Figure 6 and Figure 7 In the vertical soil profile, shallower soil depths led to greater increases in both surface drought characteristics, while deeper soil depths led to greater increases in deep drought. This suggests that under a warming climate, the duration and intensity of deep (surface) droughts are projected to increase from the bottom (top). Under the SSP585 scenario, the increases in both surface and deep droughts are even greater.

[0087] Figure 6 and Figure 7 The temporal variation trends of the characteristics of both types of drought events also show a significant upward trend. Under the SSP245 scenario, the duration of surface drought is expected to increase by 1.3 days per decade, while the intensity will increase by 5.4×10 -4 m 3 ·m -3 The two characteristics of deep drought increased (duration: 2.4 days per decade; intensity: 6.5×10 -4 m 3 ·m -3 ) is greater than surface drought. Under the influence of increased anthropogenic emissions, the increase in both surface and deep drought under the SSP585 scenario is further amplified by 1.6 to 2.0 times. This suggests that global deep droughts will last longer and be more intense, and the risk of intensified drought in the deep soil layer will also increase.

[0088] Combined with the land cover dataset, the changes in surface and deep drought characteristics of different vegetation types were estimated. Deep-rooted vegetation cover categories (i.e., forest, shrub, and savanna) were selected and the area-weighted average of the differences in deep and surface drought characteristics (duration and intensity) across different vegetation areas was calculated. Figure 8 As shown, for ecosystems known to contain deep-rooted species, such as forests, shrublands, and savannas, the difference in duration and intensity of both deep and surface droughts showed a significant upward trend between 1981 and 2100, with this increase also found in the Amazon basin.

[0089] (6) Physical drivers of surface and deep drought;

[0090] This implementation case attributes surface and deep drought events to different driving factors, such as extreme heat (TEM), precipitation deficit (PRE), and vegetation greening (high leaf area index; LAI). The average values ​​of these three variables in the 30 days before the drought event are calculated, and their severity index (SI) is calculated as follows:

[0091] SI v =(v i,j -μ j ) / σ j

[0092] Where i represents year, j represents month, v represents driving factor, μ and σ represent the climate mean and standard deviation of monthly variables in the climate state (historical period: 1981-2010, future period: 2061-2100), respectively.

[0093] Defining extreme high temperatures as SI TEM >0.8, precipitation deficit is defined as SI PRE <-0.8, vegetation greening is defined as SI LAI >0.8. If a drought event and an extreme event that occurred in the previous 30 days occurred at the same grid point, the drought event at that grid point was attributed to this type of driving factor, and the normalized contributions of different driving factors to the duration and intensity of the two types of drought events were calculated.

[0094] like Figure 9 and Figure 10 As shown in the three reanalysis datasets from 1981 to 2020, antecedent precipitation deficits accounted for 45% and 44% of the total duration and intensity of global surface droughts, respectively, and dominated 69% and 61% of global surface drought evolution, respectively. This indicates that global surface droughts are highly sensitive to precipitation changes. The main drivers of deep droughts exhibit significant spatial heterogeneity. In arid regions, antecedent extreme high temperatures are the primary driver of deep droughts, particularly in eastern North America, Europe, and the Middle East, accounting for 52% of the total duration and intensity. These regions are hotspots of soil moisture-air temperature coupling.

[0095] In humid regions, such as the tropics and eastern Asia, the total duration and intensity of deep droughts are mainly attributed to the simultaneous occurrence of antecedent precipitation deficits (40% and 42%) and extreme high temperatures (43% and 44%). In global vegetation greening hotspots, such as the high latitudes of the Northern Hemisphere, the role of antecedent vegetation greening cannot be ignored, with the first three driving factors each contributing 33% ( Figure 9 i and Figure 10 i) in the text.

[0096] like Figure 11As shown, under continued anthropogenic emissions, the main drivers of drought across different climate regions will change little between 2061 and 2100. Extreme heat days are projected to increase in eastern North America, Europe, and the Middle East. The combination of precipitation deficits and extreme heat waves will become more frequent in tropical and East Asia, and vegetation greening will become more persistent in high-latitude Northern Hemisphere regions. Anthropogenic climate change will exacerbate these driving processes, further exacerbating deep droughts.

[0097] See Figure 12 , Figure 12 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 four-dimensional soil moisture drought event tracking and warming signal identification device 401, a processor 402 and a storage medium 403.

[0098] A four-dimensional soil moisture drought event tracking and warming signal identification device 401: The four-dimensional soil moisture drought event tracking and warming signal identification device 401 implements the four-dimensional soil moisture drought event tracking and warming signal identification method.

[0099] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the four-dimensional soil moisture drought event tracking and warming signal recognition method.

[0100] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the four-dimensional soil moisture drought event tracking and warming signal identification method.

[0101] The beneficial effects of the present invention are:

[0102] (1) The present invention provides a method for identifying four-dimensional soil moisture drought, which helps the academic community deepen its understanding of this new type of soil drought. The present invention divides four-dimensional soil drought into two different types, quantifies the spatiotemporal characteristic indicators of four-dimensional soil drought, and clarifies the changes and impacts of four-dimensional soil moisture drought in the four-dimensional spatiotemporal domain of longitude, latitude, depth, and time. The present invention quantifies the differences in the spatiotemporal evolution of two types of four-dimensional soil drought in the future period under different land cover types, providing a scientific basis for a deeper understanding of the dynamic evolution process of drought events.

[0103] (2) The present invention quantitatively detects anthropogenic warming signals in the spatiotemporal evolution of four-dimensional soil moisture drought characteristics, identifies the impact of anthropogenic warming on the changes in two four-dimensional drought events, and estimates the changes in four-dimensional drought events in the future, providing theoretical support for decision makers to make decisions on adapting to and mitigating the effects of climate change on soil moisture drought.

[0104] (3) This paper analyzes the contribution of driving factors to four-dimensional soil moisture drought events, clarifies the contribution of different driving factors to four-dimensional soil moisture drought in history and future periods, quantifies the dominant driving factors of two types of four-dimensional soil moisture drought, and provides theoretical support for the attribution of four-dimensional soil moisture drought.

[0105] 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 four-dimensional soil moisture drought event tracking and warming signal identification method, characterized by: The method comprises the following steps: Step S1: Data collection: Collect global climate data; Step S2: Identification of four-dimensional soil moisture drought events: Based on the data obtained in step S1, the 10th percentile of the climatological growing season soil moisture value of each grid point is calculated as the drought threshold. Grid points below this threshold are identified as drought grid points, and continuous four-dimensional soil moisture drought events are identified using the Lagrangian four-dimensional tracking method. Step S3: Quantification of four-dimensional soil moisture drought event characteristics: Combined with the four-dimensional drought events obtained in step S2, soil moisture drought event characteristics are calculated, and soil moisture drought events are divided into surface drought and deep drought; Step S4: Detecting anthropogenic warming signals from the temporal evolution of the four-dimensional soil moisture drought characteristics over the historical period; combining the temporal and intensity characteristics of the two types of drought events obtained in step S3, using the correlation coefficient method and the optimal fingerprint method, detect anthropogenic warming signals from drought characteristic indicators calculated based on reanalysis data and historical climate experiments of climate models; Step S5: Prediction of the future evolution of four-dimensional soil moisture drought; combining the time and intensity characteristics of the two types of drought events obtained in step S3, combined with the historical climate experiments of the climate model and the future socio-economic development path experiments, to obtain the spatiotemporal evolution characteristics of drought events in the future period; Combined with the global land cover data obtained in step S1, the spatiotemporal evolution characteristics of drought events under different vegetation types in the future period are obtained; Step S6: Four-dimensional soil moisture drought event driving factors; combining the time and intensity characteristics of the two types of drought events obtained in step S3 with the near-surface temperature, precipitation, and leaf area index data obtained in step S1, the two types of drought events are attributed to the three types of antecedent driving factors, and the contribution of different driving factors to the drought events is quantified; Step S2 is specifically as follows: S21: Soil moisture data preprocessing: Based on GSWP-3 precipitation data, remove grid points in the Arctic and Antarctic regions and desert areas with annual precipitation less than 100 mm; use soil moisture values ​​during the growing season; convert soil moisture values ​​from mass to volume; linearly interpolate soil moisture values ​​from multiple layers of different data to a common vertical profile; convert datasets of different time scales to daily scale; Second-order conservative remapping was used to interpolate all datasets to a resolution of 1° × 1° to ensure consistent spatiotemporal resolution. An 11-day smoothing average was used to eliminate abnormal series fluctuations. S22: Use the Lagrangian tracing method to simulate the spatiotemporal evolution of soil moisture drought in four dimensions, as follows: For each grid point in a single layer, when the soil moisture value is lower than the 10th percentile of the growing season in the climatic period, the grid point is defined as a drought grid point; adjacent drought grid points in a single layer are merged into two-dimensional spatially continuous drought patches; for each time step, all adjacent two-dimensional drought patches within 26 adjacent areas are connected into an independent three-dimensional continuous drought patch; any spatially continuous three-dimensional drought patches are merged into a single four-dimensional drought event if the overlap rate in consecutive time steps is greater than a preset percentage; the four-dimensional drought event lasts for at least a preset number of days and the projected area is greater than a preset value.

2. The four-dimensional soil moisture drought event tracking and warming signal identification method according to claim 1, characterized in that: The global climate data in step S1 include: the ERA5-Land dataset of the Weather Forecast Center, the CRA-Land product data of the Global Atmosphere and Land Surface Reanalysis of the Meteorological Administration, the soil moisture, near-surface temperature and precipitation data of the Global Land Data Assimilation System GLDAS-Noah, the leaf area index data of the Global Long-term Leaf Area Index GLOBMAP data, the precipitation data collected from the Global Soil Moisture Project GSWP-3, the soil moisture, near-surface temperature, precipitation and leaf area index data of the sixth phase of the International Coupled Comparison Program CMIP6, and the global land cover data collected from the Moderate Resolution Imaging Spectroradiometer Land Cover Climate Simulation Grid MCD12C1.

3. The four-dimensional soil moisture drought event tracking and warming signal identification method according to claim 1, characterized in that: Step S3 is as follows: S31: The temporal and spatial characteristics of a four-dimensional soil moisture drought event are duration and intensity, respectively. Duration is the duration of the four-dimensional soil moisture drought event, and intensity is the weighted average of the grid area and depth of the soil moisture deficit at all grid points in the four-dimensional drought event. They are as follows: in, i is the grid point, t For time, D For the duration, l For the soil layer, d is the soil depth, s is the grid area, thres is the drought threshold, sm is the soil water value, cluster for drought events; S32: Four-dimensional drought events are divided into three types: surface drought with a heavy top inverted iceberg structure, defined as the total duration of the surface soil moisture deficit area being greater than the deep soil moisture deficit area for a period of more than a preset percentage; deep drought with a heavy bottom iceberg structure, defined as the total duration of the deep soil moisture deficit area being greater than the surface soil moisture deficit area for a period of more than a preset percentage; when neither of these two conditions occurs, it is a columnar drought structure; S33: During the division process, the top two layers were selected as surface soil and the third layer as deep soil.

4. The four-dimensional soil moisture drought event tracking and warming signal identification method according to claim 1, characterized in that: Step S4 is specifically as follows: S41: Detect and attribute anthropogenic climate change to four-dimensional soil moisture drought events based on the correlation coefficient method. Calculate the Spearman correlation coefficient between the changes in the global surface drought and deep drought characteristic sequences based on reanalysis data and CMIP6 historical simulations. Divide the CMIP6 pre-industrial simulations into multiple blocks of years corresponding to the study period and calculate the multi-model average Spearman correlation coefficient with the historical simulations. Analyze whether the correlation coefficients between the drought event characteristic sequences simulated by reanalysis data and historical simulations exceed the 95th or 99th percentile of the set of correlation coefficients between multiple blocks of historical simulations and pre-industrial simulations. S42: Detection and attribution of anthropogenic climate change to four-dimensional soil moisture drought events based on the optimal fingerprint method. The specific calculation is as follows: in, is the global mean characteristic change of surface and deep drought based on reanalysis, are the global mean changes in the surface and deep drought characteristics simulated by the models, is the sampling uncertainty, is the scale factor, Internal climate variability estimated for pre-industrial simulations.

5. The four-dimensional soil moisture drought event tracking and warming signal identification method according to claim 1, characterized in that: Step S5 is specifically as follows: S51: Calculate the spatiotemporal evolution of global surface and deep drought characteristics in the future period using CMIP6 historical simulations and future scenarios, including the Representative Concentration Pathways (SSPs) and Shared Socioeconomic Pathways (SSPs), including the SSP245 and SSP585 scenarios. S52: Combined with the land cover dataset, estimate the changes in surface drought and deep drought characteristics of different vegetation types, select deep-rooted vegetation cover categories, and calculate the area-weighted average of the differences in deep drought and surface drought characteristics in different vegetation areas.

6. The four-dimensional soil moisture drought event tracking and warming signal identification method according to claim 1, characterized in that: Step S6 is as follows: S61: Surface and deep drought events are attributed to three different driving factors, including extreme high temperature (TEM), precipitation deficit (PRE), and high leaf area index (LAI). The average values ​​of the three variables for the 30 days before the drought event are calculated, and the severity index (SI) is calculated as follows: in, i Indicates the year, j Indicates the month, represents the driving factor, and represent the climatological mean and standard deviation of the monthly variables of the climatological state; S62: Extreme heat is defined as >Default value, precipitation deficit is defined as <Default value, vegetation greening is defined as >Default value: If a drought event and an extreme event that occurred in the previous 30 days occur at the same grid point, the drought event at that grid point is attributed to this type of driving factor, and the contribution of different driving factors to the duration and intensity of the two types of drought events is calculated.

7. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a four-dimensional soil moisture drought event tracking and warming signal identification method as described in any one of claims 1 to 6.

8. A four-dimensional soil moisture drought event tracking and warming signal identification device, characterized by: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a four-dimensional soil moisture drought event tracking and warming signal identification method as described in any one of claims 1 to 6.

Images