A statistical correction method for terrestrial water resources availability simulation data

Abstract

The present invention discloses a statistical correction method for terrestrial water resource availability simulation data, belonging to the field of climate data analysis technology. A spatiotemporal distribution simulation capability assessment model is constructed to screen global climate system models with good simulation effects, determine historical simulation data #imgabs0# and future simulation data #imgabs1#; establish a cumulative distribution function to quantify the amount and rate of change from historical simulation data #imgabs2# at the same quantile to future simulation data #imgabs3#; select a model simulation trend preservation method, and use the quantile mapping method to obtain bias-corrected future simulation data #imgabs4#; sort the bias-corrected future simulation data #imgabs5# to obtain time-series-based future simulation data #imgabs6#, and perform bias correction on multi-period simulation data to form a complete historical simulation data set and future simulation data set. The present invention uses WA as the correction target variable to fundamentally improve the WA simulation accuracy, analyzes the statistical correction effect of the quantile mapping method on the WA variable, and meets the needs of most climate-scale research.

Description

Technical Field

[0001] The present invention belongs to the technical field of climate data analysis, and in particular relates to a statistical correction method for terrestrial water resource availability simulation data. Background Art

[0002] In the field of climate research, terrestrial water availability (WA) is defined as the difference between precipitation and evapotranspiration. On monthly and longer timescales, according to the atmospheric water vapor balance equation, WA is balanced with the water vapor flux divergence of the entire atmosphere; according to the terrestrial water balance equation, WA can generally be regarded as the main contributor to local runoff and terrestrial water storage. Changes in WA are closely linked to the water cycle and are an important indicator of water resource changes, with significant impacts on human social activities and the natural ecological environment. Global climate system models are objective tools and means for conducting climate predictions. However, the simulated precipitation and evapotranspiration values ​​output by current models generally deviate significantly from the observed values, resulting in poor simulation of the inferred terrestrial WA. Therefore, when analyzing WA simulation data, the model output data should be post-processed first.

[0003] Existing studies of the spatiotemporal distribution characteristics of global WA often directly analyze model simulation results from an equal-weighted multi-model ensemble average, or perform different simulation data corrections based on the research objectives. For example, in comparative studies of future changes in WA over land and sea, only evaporation in the ocean region is corrected to ensure a global balance of total water volume over land and sea. Another example is the study of the impact of combined changes in precipitation and evapotranspiration in different climate zones on WA. A Bayesian model averaging method is used to perform multi-model weighted average corrections on model outputs of precipitation and evapotranspiration, analyzing the contribution and dominance of precipitation and evapotranspiration to WA changes. In studies of the spatiotemporal evolution of water resources dominated by meteorological conditions and their ecological and environmental effects, there is an urgent need to correct biases in global gridded future WA simulation data. However, there is currently little discussion on grid-by-grid correction methods for WA simulation data. Summary of the Invention

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a statistical correction method for terrestrial water resource availability simulation data.

[0005] The present invention is implemented by the following technical solution: a statistical correction method for terrestrial water resource availability simulation data, comprising the following steps:

[0006] The WA historical observation value is calculated based on the precipitation and evapotranspiration observation data of the historical period, and the WA historical observation value is used as the historical reference data. Based on the simulated historical precipitation and evapotranspiration data, the WA historical simulation value is calculated, and a spatiotemporal distribution simulation capability evaluation model is constructed to screen out the global climate system models with better WA simulation effects, and use their simulated WA historical data as historical simulation data. , WA future data as future simulation data ;

[0007] Create historical simulation data and future simulation data The cumulative distribution function of quantifies the historical simulation data of the same quantile Simulating data into the future The amount of change and rate of change , is the cumulative probability;

[0008] Establish historical reference data The cumulative distribution function of and historical simulation data The corresponding relationship between the quantile function values ​​is used to select the appropriate mode of model simulation trend preservation; the quantile mapping method is used to obtain bias-corrected future simulation data containing historical reference data and model simulation future trend information. ;

[0009] Based on future simulation data The quantile series of bias-corrected future simulated data Sort and get future simulation data based on time series .

[0010] In a further embodiment, the construction process of the spatiotemporal distribution simulation capability evaluation model is as follows:

[0011] Create a scoring model and use it to calculate the Taylor score corresponding to the global climate system model and interannual variability scores ; Among them, Taylor score The higher the value, the better the simulation effect of the corresponding global climate system model on the spatial distribution of WA values. The lower it is, the better the corresponding global climate system model simulates the interannual variability of WA values;

[0012] Based on Taylor rating and interannual variability scores , using comprehensive scoring The global climate system model's ability to simulate the spatiotemporal distribution of WA values ​​is obtained, and the comprehensive score is The calculation formula is as follows:

[0013] ;

[0014] Where, The number of global climate system models that participated in the scoring, is the number of scoring methods, , The scoring method The order of descending order of preference is The larger the value, the stronger the ability of the corresponding global climate system model to simulate the spatiotemporal distribution characteristics of WA.

[0015] In a further embodiment, the variation The solution formula is as follows:

[0016] Where, Indicates the cumulative probability of future simulation data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of ;

[0017] The rate of change The calculation formula is as follows:

[0018] .

[0019] In a further embodiment, the mode simulation trend saving mode includes at least: an additive simulation trend saving mode, a multiplicative simulation trend saving mode, and a mixed simulation trend saving mode; correspondingly, the steps of selecting and adapting the mode simulation trend saving mode are as follows:

[0020] Step 101: Obtain the same cumulative probability The quantile function value of the historical reference data and the quantile function value of historical simulation data , if the quantile function value and quantile function values If the signs of are the same, then step 102 is executed; otherwise, the additive simulation trend saving mode is selected;

[0021] Step 102: If , then select the multiplication simulation trend saving mode, otherwise execute step 103;

[0022] Step 103: If , then select the mixed simulation trend saving method, otherwise select the additive simulation trend saving method.

[0023] In a further embodiment, the bias-corrected future simulation data The process of obtaining is as follows:

[0024] Based on the quantile mapping method, the changing trend of the model simulation is coupled to the historical reference data through the adaptive model simulation trend preservation method. , and its coupling form is as follows:

[0025] , Simulate trend saving method for selecting the appropriate mode;

[0026] Correspondingly, future simulation data based on time series The expression is:

[0027] ,in, Indicates the year Future simulation data, is the empirical cumulative distribution function of future simulated data, Quantile function for bias-corrected future simulated data, Indicates the year Future simulation values ​​of the bias correction.

[0028] In a further embodiment, the quantile mapping method of the additive simulation trend preservation method is as follows:

[0029] Where, Year Historical reference data of yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data.

[0030] In a further embodiment, the quantile mapping method of the multiplication simulation trend preservation method is as follows:

[0031] Where, Year Historical observation data, yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data, .

[0032] In a further embodiment, the quantile mapping method of the hybrid simulation trend preservation method is as follows:

[0033] ;in, Year Historical observation data, yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data, is a mixed correction operator.

[0034] In a further embodiment, the hybrid correction operator The calculation formula is as follows:

[0035] , where Indicates the cumulative probability of historical reference data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of .

[0036] In a further embodiment, the following steps are also included:

[0037] The historical period is kept as the modeling period, and the future period is used as the correction period. A sliding window is used to select different correction periods to correct the future simulation data, forming a long time series of bias-corrected historical simulation data and future simulation data;

[0038] The above correction processing is performed on the future simulation data of the global climate system on a grid-by-grid, monthly, and individual basis to form bias-corrected historical simulation data sets and future simulation data sets, thereby obtaining complete historical simulation data sets and future simulation data sets.

[0039] The present invention directly uses WA as the target variable for correction, skipping the variable-by-variable correction process for individual components. By globally optimizing the distribution characteristics of the difference, it fundamentally improves WA simulation accuracy. This design avoids the cumulative error problem associated with traditional step-by-step corrections and is particularly well-suited to the physical mechanism where WA is influenced by the combined effects of precipitation and evapotranspiration. This ensures that the corrected data more accurately reflects the actual state of water surplus / deficit.

[0040] This paper describes the statistical correction steps for the WA variable using the quantile mapping method. The multiplication and hybrid simulation trend preservation methods in the quantile mapping method are extended to algorithms that are universally applicable to positive and negative values. The correction effects of the addition, multiplication, and hybrid simulation trend preservation methods on WA are analyzed and compared. The optimal simulation trend preservation method is selected to form a global gridded WA simulation correction dataset on a monthly time scale. This data distribution meets the needs of most climate-scale studies. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 4 is a flow chart of the statistical correction method for the land water resource availability simulation data in Example 1.

[0042] Figure 2 It is a schematic diagram of the cumulative distribution function of historical reference data, historical simulation data, future simulation data, and future simulation data corrected by the quantile mapping method using multiple simulation trend preservation methods.

[0043] Figure 3 It is an extended hybrid correction operator in the hybrid simulation trend preservation method. Schematic diagram of . DETAILED DESCRIPTION

[0044] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0045] Example 1

[0046] This embodiment discloses a statistical correction method for land water resource availability simulation data, such as Figure 1 Shown, including:

[0047] The WA historical observation value is calculated based on the precipitation and evapotranspiration observation data of the historical period, and the WA historical observation value is used as the historical reference data. Based on the historical precipitation and evapotranspiration data of the model simulation period, the WA historical simulation value is calculated, and a spatiotemporal distribution simulation capability evaluation model is constructed to screen out the global climate system models with better WA simulation effects, and use their simulated WA historical data as historical simulation data. , WA future data as future simulation data .

[0048] Create historical simulation data and future simulation data The cumulative distribution function of quantifies the historical simulation data of the same quantile Simulating data into the future The amount of change and rate of change , is the cumulative probability;

[0049] Establish historical reference data The cumulative distribution function of and historical simulation data The corresponding relationship between the quantile function values ​​is used to select the appropriate mode of model simulation trend preservation; the quantile mapping method is used to obtain bias-corrected future simulation data containing historical reference data and model simulation future trend information. ;

[0050] Based on future simulation data The quantile series of bias-corrected future simulated data Sort and get future simulation data based on time series .

[0051] It is worth noting that the historical precipitation and evapotranspiration observation data described in this example are all derived from the ERA5-Land reanalysis land surface meteorological dataset provided by the European Centre for Medium-Range Weather Forecasts. These datasets are characterized by high precision, gridding, and multivariate coherent changes, such as the monthly precipitation and evapotranspiration data with an accuracy of 0.1° from 1950 to 2023. ,in, is the precipitation, = evapotranspiration during the same period. Therefore, the WA historical reference data can be directly calculated from the precipitation and evapotranspiration observation data of the historical period.

[0052] Furthermore, the global climate system models described in this embodiment can be the 29 global climate system models from the sixth Coupled Model Intercomparison Project (CMIP6). These 29 global climate system models include precipitation and evapotranspiration simulation data from historical simulation experiments from 1850 to 2014 and the Scenario Model Intercomparison Project (ScenarioMIP) from 2015 to 2100. Based on the above example, a spatiotemporal distribution simulation capability assessment model is used to screen out global climate system models with good WA simulation performance from among the 29 global climate system models.

[0053] Furthermore, the construction process of the spatiotemporal distribution simulation capability evaluation model is as follows:

[0054] Create a scoring model and use it to calculate the Taylor score corresponding to the global climate system model and interannual variability scores ; Among them, Taylor score The higher the value, the better the corresponding global climate system model simulates the spatial distribution of WA values. The highest score is 1; interannual variability score The lower the value, the better the simulation effect of the corresponding global climate system model on the interannual variability of WA value. The minimum score is 0.

[0055] Based on Taylor rating and interannual variability scores , using comprehensive scoring The global climate system model's ability to simulate the spatial and temporal distribution of WA is obtained, and the comprehensive score is The calculation formula is as follows:

[0056] ;

[0057] Where, The number of global climate system models that participated in the scoring, is the number of scoring methods, , The scoring method The order of descending order of preference is The larger the value, the stronger the corresponding global climate system model's ability to simulate the spatiotemporal distribution characteristics of WA. , .

[0058] In this embodiment, Taylor scoring and interannual variability scores The calculation formulas are as follows:

[0059] ;

[0060] ;

[0061] Where, is the Taylor score, R represents the central spatial correlation coefficient between the future simulation data of the global climate system of WA and the historical reference data, is the maximum value of the spatial correlation coefficient, the superscript ref is the reference data, and sim is the simulated data. represents the standard deviation of the spatial dimension corresponding to the simulated data, is the standard deviation of the spatial dimension corresponding to the reference data; Score the interannual variability, is the standard deviation of WA simulation data in the time dimension, is the standard deviation of WA reference data in the time dimension.

[0062] Table 1 Five mode information and score rankings

[0063]

[0064] Based on this, this example screened the top 13 global climate system models (GCSMs) from 29 for their ability to simulate the spatiotemporal distribution of WA from 1951 to 2014 for subsequent calibration. Table 1 provides the performance of five of the 13 GCMs (CESM2-WACCM, CMCC-ESM2, CMCC-CM2-SR5, NorESM2-MM, TaiESM1, and MIROC6) and their corresponding comprehensive scores.

[0065] All data in this embodiment are processed to 1° accuracy using bilinear interpolation.

[0066] In the prior art, parameterized theoretical distribution functions are often used to determine the distribution of variables. However, a growing body of research indicates that these theoretical distribution functions are not applicable to all spatial grids and global climate system models. In contrast, cubic spline smoothing can easily yield parameter-free empirical cumulative distribution functions (CDFs), which better match reference data. Therefore, this embodiment employs the cubic spline smoothing method to obtain the CDF of WA data.

[0067] Historical simulation data of the same quantile Simulating data into the future The changing trend of can be quantified by the amount of change or the rate of change. The solution formula is as follows:

[0068] ;

[0069] Where, Indicates the cumulative probability of future simulation data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of ;

[0070] Rate of change The calculation formula is as follows:

[0071] ;

[0072] Where, Indicates the cumulative probability of future simulation data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of .

[0073] Furthermore, the quantile mapping method of multiple model simulation trend preservation methods is used to couple the quantitative model simulation future change trend to the historical observation data to form bias-corrected future simulation data. For example, the model simulation future change quantity can be used only The additive simulation trend preservation method is used to perform quantile mapping correction. However, even if the WA observation and simulation values ​​at the same grid point have the same sign, the additive simulation trend preservation method can still cause the sign of the corrected WA simulation data to change, confusing the local water surplus or deficit. Although the quantile mapping method of the multiplication simulation trend preservation method ensures the consistency of the correction result sign in this case, when the historical observation value is much larger than the historical simulation value , even if the future rate of change Even if the amplitude of the WA is limited to a reasonable range, the multiplication simulation trend preservation method will still lead to spurious abnormally large values. In addition, when the historical simulation and observation values ​​of WA have different signs, that is, the historical simulation results of the model fail to accurately simulate the WA sign, the multiplication simulation trend preservation method will pass the incorrect simulation sign to the subsequent correction value.

[0074] Therefore, this embodiment introduces the method of including the mode simulation variation and rate of change A hybrid simulation trend preservation method was developed, with additional steps for selecting and adapting the simulation trend preservation method, resulting in a quantile-adapted simulation trend preservation method. By comparing the correspondence between model simulation and observed data for the same quantile period, the simulation trend preservation method was refined for different historical simulation biases. A combination of additive, multiplicative, and hybrid simulation trend preservation methods was used to correct the bias of the quantile mapping method.

[0075] Therefore, the mode simulation trend saving method of this embodiment includes at least: an addition simulation trend saving method, a multiplication simulation trend saving method and a mixed simulation trend saving method.

[0076] Furthermore, the subscript hist is used as the historical period and fut is used as the future period. The steps for selecting and adapting the mode of saving the model simulation trend are as follows:

[0077] Step 101: Obtain the same cumulative probability The quantile function value of the historical reference data and the quantile function value of historical simulation data , if the quantile function value and quantile function values If the signs of are the same, then step 102 is executed; otherwise, the additive simulation trend saving mode is selected;

[0078] Step 102: If , then select the multiplication simulation trend saving mode, otherwise execute step 103;

[0079] Step 103: If , then select the mixed simulation trend saving method, otherwise select the additive simulation trend saving method.

[0080] Furthermore, the bias-corrected future simulation data The process of obtaining is as follows:

[0081] Based on the quantile mapping method, the changing trend of the model simulation is coupled to the historical reference data through the adaptive model simulation trend preservation method. , and its coupling form is as follows:

[0082] , Simulate trend saving method for selecting the appropriate mode;

[0083] In order to better understand the mathematical principles of each simulation trend preservation method and its selection and adaptation steps in the quantile mapping method, taking the CESM2-WACCM future simulation data of July WA and the ERA5-Land observation data at the (30°S, 70°W) grid as examples, the additive simulation trend preservation method, the multiplicative simulation trend preservation method, and the hybrid simulation trend preservation method of the quantile mapping correction method are further explained.

[0084] Combine Figure 2 The black solid line represents the historical reference data, the blue solid line represents the historical simulation data, the blue dotted line represents the future simulation data, the orange dotted line represents the quantile mapping correction result using only the additive simulation trend preservation method, the red dotted line represents the quantile mapping correction result using only the multiplicative simulation trend preservation method, and the red solid line represents the quantile mapping correction result using the quantile adaptation simulation trend preservation method, respectively indicating the corresponding cumulative distribution functions. Figure 2 The dark blue and light blue curves show that there are significant differences in the distribution characteristics of the historical and future simulated values ​​of WA at the same grid point, so a correction method is needed to retain the model simulation trend.

[0085] like Figure 2 As shown by the green arrow in (a), the quantile mapping method of the additive simulation trend preservation method is as follows:

[0086] Where, Year Historical reference data, yes The cumulative probability of Represents the cumulative probability Corrected future simulation data.

[0087] Figure 2 (a) Indicates that the cumulative probability The simulated future changes calculated are linearly superimposed on the historical observation data. Figure 2 The orange dashed line in the figure is the correction result using only the additive simulation trend preservation method.

[0088] The quantile mapping method of multiplication simulation trend preservation method is as follows Figure 2 As shown by the green arrow in (b), it should be noted that the negative signs in historical observations and simulations are not Figure 2 (b) shows:

[0089] Where, Year Historical observation data, yes The cumulative probability of Indicates the cumulative probability The corrected future simulation data, . Figure 2 The red dashed line in the figure is the correction result of the quantile mapping method using only the multiplication simulation trend preservation method.

[0090] The quantile mapping method of the hybrid simulation trend preservation method is as follows:

[0091] ;in, Year Historical reference data, yes The cumulative probability of Represents the cumulative probability Corrected future simulation data, is a mixed correction operator.

[0092] Furthermore, the hybrid correction operator The calculation formula is as follows:

[0093] , where Indicates the cumulative probability of historical reference data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of .

[0094] Furthermore, the hybrid correction operator is extended using piecewise functions The value of can be used to integrate the selection and adaptation steps of the simulation trend preservation method into the formula of the hybrid simulation trend preservation method to form a quantile-adapted simulation trend preservation method. The formula is as follows:

[0095] ;

[0096] Where, Indicates the cumulative probability of historical reference data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of .

[0097] Figure 3 Extended hybrid correction operator The distribution characteristics of the quantile mapping method are shown, which shows the selection steps of different simulation trend preservation methods and the use of mixed trend preservation methods. The value of Figure 3 The horizontal coordinate QQ in is The purple circle is the correspondence between ERA5-Land values ​​and CESM2-WACCM simulation values ​​at the example grid point (30°S, 70°W) from 1951 to July 2014. Operator value. There is a situation where the historical simulation and observation data of the same quantile have different signs (i.e. ),at this time is 0, that is, the additive simulation trend preservation method is used, and the absolute values ​​of the observed values ​​of the remaining quantiles are appropriately larger than the simulated values ​​( ),at this time That is, use the mixed simulation trend saving method.

[0098] It can be seen that this method couples the future change trends simulated and calculated by the physical equations of the global climate system model on the basis of historical observation data, and the correction idea is intuitive and reasonable.

[0099] Furthermore, based on the future simulation data of the time series The expression is: ,in, Indicates the year Future simulation data, is the empirical cumulative distribution function of future simulated data, Quantile function for bias-corrected future simulated data, Indicates the year Future simulation values ​​of the bias correction.

[0100] In another embodiment, the following steps are also included:

[0101] The historical period is kept as the modeling period, and the future period is used as the correction period. A sliding window is used to select different correction periods to correct the future simulation data, forming a long time series of bias-corrected historical simulation data and future simulation data;

[0102] The above correction processing is performed on the future simulation data of the global climate system on a grid-by-grid, monthly, and individual basis to form bias-corrected historical simulation data sets and future simulation data sets, thereby obtaining complete historical simulation data sets and future simulation data sets.

[0103] For example, for the widely used global climate system simulation data of the Sixth International Coupled Model Intercomparison Project (CMIP6), the 36 years from 1979 to 2014 can be set as the modeling period, and the correction period can be set as a 36-year window of the same length. The historical simulation data can be corrected in sections, and the periods are divided into the following: 1850-1885, 1886-1921, 1922-1957, 1958-1993 and 1994-2029, of which 1994-2029 is the historical simulation data from 1994 to 2014. The bias-corrected historical simulation data from 1850 to 2014 are formed by combining the historical simulation data with the future simulation data under the SSP-5.85 scenario from 2015 to 2029. The correction periods for the future simulation data are divided into the following: 2015 to 2050 (the correction results from 2015 to 2043 are retained), 2040 to 2075 (the correction results from 2044 to 2072 are retained), and 2065 to 2100 (the correction results from 2073 to 2100 are retained), thus forming the bias-corrected future simulation data from 2015 to 2100.

[0104] In another embodiment, the method further includes: quantitatively evaluating and comparing the correction results of the quantile mapping method using the additive simulation trend preservation method, the multiplicative simulation trend preservation method, and the quantile-adapted simulation trend preservation method.

[0105] Furthermore, the central spatial correlation coefficient (R), mean absolute error (MAE; mm day -1 ) and sign agreement rate (SAR; %) are used to test the bias correction effect of the quantile mapping method with different simulation trend preservation methods on WA.

[0106] The analysis period was set to 1951–2014, a total of 64 years. This period was divided into four sub-periods with 10-year sliding windows. Thirty-two consecutive years served as the modeling period for the statistical calibration algorithm: 1951–1982, 1961–1992, 1971–2012, and 1981–2012. The remaining 32 years served as the validation period. The cross-validation results from these four data sets can reduce the impact of the modeling and validation period selection on the test results.

[0107] R represents the spatial similarity between the simulation and reference data, with an optimal value of 1; MAE measures the average difference between the simulation and reference values, with an optimal value of 0 mm day -1The optimal value for the agreement between the simulated and reference values ​​for SAR is 100%. The three indicators were calculated for each subperiod and for each global climate system model (MAE and SAR were first calculated for each grid point), and then averaged to obtain the overall indicator. Note that each annual mean parameter has 52 samples (4 subperiods × 13 models). Therefore, the nonparametric Mann-Whitney U test can be used to compare whether there are significant differences in the evaluation indicators under different simulation trend preservation methods. The significance level for this test is p < 0.05.

[0108] While evaluating the global average simulation performance of the WA variable, the effectiveness of the WA correction under different drought scenarios was also assessed. Terrestrial drought severity was quantified using the Aridity Index (AI) averaged from 1951 to 2014. The AI ​​is defined as the ratio of annual precipitation to annual potential evapotranspiration: AI < 0.05 indicates extremely arid areas, 0.05 ≤ AI < 0.20 indicates arid areas, 0.20 ≤ AI < 0.50 indicates semi-arid areas, 0.50 ≤ AI < 0.65 indicates arid sub-humid areas, and AI ≥ 0.65 indicates humid areas.

[0109] Example 2

[0110] This embodiment discloses a statistical correction system for land water resource availability simulation data, which is used to implement the statistical correction method for land water resource availability simulation data as described in the embodiment, including:

[0111] The first module is configured to calculate the WA historical observation value based on the precipitation and evapotranspiration observation data of the historical period, and use the WA historical observation value as the historical reference data. Based on the simulated historical precipitation and evapotranspiration data, the WA historical simulation value is calculated, and a spatiotemporal distribution simulation capability evaluation model is constructed to screen out the global climate system models with better WA simulation effects, and use their simulated WA historical data as historical simulation data. , WA future data as future simulation data ;

[0112] The second module is set up to build historical simulation data and future simulation data The cumulative distribution function of quantifies the historical simulation data of the same quantile Simulating data into the future The amount of change and rate of change , is the cumulative probability;

[0113] The third module is set up to build historical reference data The cumulative distribution function of and historical simulation data The corresponding relationship between the quantile function values ​​is used to select the appropriate mode of model simulation trend preservation; the quantile mapping method is used to obtain bias-corrected future simulation data containing historical reference data and model simulation future trend information. ;

[0114] The fourth module is set up based on future simulation data The quantile series of bias-corrected future simulated data Sort and get future simulation data based on time series ;

[0115] The fifth module is set to keep the historical period as the modeling period unchanged, and use the future period as the correction period. A sliding window is used to select different correction periods to correct the future simulation data, forming a long time series of bias-corrected historical simulation data and future simulation data;

[0116] The sixth module is configured to perform the above correction processing on the future simulation data of the global climate system on a grid-by-grid, monthly, and individual basis, forming bias-corrected historical simulation data sets and future simulation data sets, thereby obtaining complete historical simulation data sets and future simulation data sets.

Claims

1. A statistical correction method for terrestrial water resource availability simulation data, characterized in that: The following steps are involved: The WA historical observation value is calculated based on the precipitation and evapotranspiration observation data of the historical period, and the WA historical observation value is used as the historical reference data. Based on the simulated historical precipitation and evapotranspiration data, the WA historical simulation value is calculated, and a spatiotemporal distribution simulation capability evaluation model is constructed to screen out the global climate system models with better WA simulation effects, and use their simulated WA historical data as historical simulation data. , WA future data as future simulation data ; Create historical simulation data and future simulation data The cumulative distribution function of quantifies the historical simulation data of the same quantile Simulating data into the future The amount of change and rate of change , is the cumulative probability; the change The solution formula is as follows: Where, Indicates the cumulative probability of future simulation data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of ; The rate of change The calculation formula is as follows: ； Establish historical reference data The cumulative distribution function of and historical simulation data The corresponding relationship between the quantile function values ​​is used to select the appropriate mode of model simulation trend preservation; the quantile mapping method is used to obtain bias-corrected future simulation data containing historical reference data and model simulation future trend information. ; The mode simulation trend saving mode includes at least: an addition simulation trend saving mode, a multiplication simulation trend saving mode and a mixed simulation trend saving mode; The bias-corrected future simulation data The process of obtaining is as follows: Based on the quantile mapping method, the changing trend of the model simulation is coupled to the historical reference data through the adaptive model simulation trend preservation method. , and its coupling form is as follows: , Simulate trend saving method for selecting the appropriate mode; Based on future simulation data The quantile series of bias-corrected future simulated data Sort and get future simulation data based on time series .

2. The statistical correction method for terrestrial water resource availability simulation data according to claim 1, characterized in that: The construction process of the spatiotemporal distribution simulation capability evaluation model is as follows: Create a scoring model and use it to calculate the Taylor score corresponding to the global climate system model and interannual variability scores ; Among them, Taylor score The higher the value, the better the simulation effect of the corresponding global climate system model on the spatial distribution of WA values. The lower it is, the better the corresponding global climate system model simulates the interannual variability of WA values; Based on Taylor rating and interannual variability scores , using comprehensive scoring The global climate system model's ability to simulate the spatiotemporal distribution of WA values ​​is obtained, and the comprehensive score is The calculation formula is as follows: ； Where, The number of global climate system models that participated in the scoring, is the number of scoring methods, , The scoring method The order of descending order of preference is The larger the value, the stronger the corresponding global climate system model's ability to simulate the spatiotemporal distribution characteristics of WA.

3. The statistical correction method for terrestrial water resource availability simulation data according to claim 1, characterized in that: The steps for selecting and adapting the mode simulation trend saving method are as follows: Step 101: Obtain the same cumulative probability The quantile function value of the historical reference data and the quantile function value of historical simulation data , if the quantile function value and quantile function values If the signs of are the same, then step 102 is executed; otherwise, the additive simulation trend saving mode is selected; Step 102: If , then select the multiplication simulation trend saving mode, otherwise execute step 103; Step 103: If , then select the mixed simulation trend saving method, otherwise select the additive simulation trend saving method.

4. The statistical correction method for terrestrial water resource availability simulation data according to claim 1, characterized in that: Future simulation data based on time series The expression is: ,in, Indicates the year Future simulation data, is the empirical cumulative distribution function of future simulated data, Quantile function for bias-corrected future simulated data, Indicates the year Future simulation values ​​of the bias correction.

5. The statistical correction method for terrestrial water resource availability simulation data according to claim 3, characterized in that: The quantile mapping method of the additive simulation trend preservation method is as follows: Where, Year Historical reference data of yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data.

6. The statistical correction method for terrestrial water resource availability simulation data according to claim 3, characterized in that: The quantile mapping method of the multiplication simulation trend preservation method is as follows: Where, Year Historical observation data, yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data, .

7. The statistical correction method for terrestrial water resource availability simulation data according to claim 3, characterized in that: The quantile mapping method of the hybrid simulation trend preservation method is as follows: ;in, Year Historical observation data, yes The cumulative probability of Indicates the cumulative probability bias-corrected future simulation data, is a mixed correction operator.

8. The statistical correction method for terrestrial water resource availability simulation data according to claim 7, characterized in that: The hybrid correction operator The calculation formula is as follows: , where Indicates the cumulative probability of historical reference data The quantile function value of is the cumulative probability of historical simulation data The quantile function value of .

9. The statistical correction method for terrestrial water resource availability simulation data according to claim 1, characterized in that: The following steps are also included: The historical period is kept as the modeling period, and the future period is used as the correction period. A sliding window is used to select different correction periods to correct the future simulation data, forming a long time series of bias-corrected historical simulation data and future simulation data; The above correction processing is performed on the future simulation data of the global climate system on a grid-by-grid, monthly, and individual basis to form bias-corrected historical simulation data sets and future simulation data sets, thereby obtaining complete historical simulation data sets and future simulation data sets.

Images