A method to determine the optimal time scale for regional drought impacts

Abstract

The present invention relates to a method for determining the optimal time scale of regional drought impact. First, meteorological stations or grid meteorological data are screened according to the scope of the study area, and the average surface rainfall at different time scales in the region is calculated. Based on the regional multispectral satellite remote sensing data, the regional minimum time scale normalized vegetation index is calculated, and on this basis, the regional average normalized vegetation index consistent with the surface rainfall time scale is synthesized, and the first-order difference correlation relationship between the long-term average surface rainfall and the average normalized vegetation index at different time scales is calculated. The time scale with the largest correlation coefficient is the optimal time scale of the drought and water shortage affected area. The method of the present invention makes up for the deficiency of selecting the time scale for feature analysis based on experience or literature when conducting drought risk assessment, prediction and other drought spatiotemporal characteristics research involving time scales. After the optimal time scale is determined by this method, the drought characteristics can be analyzed more accurately.

Description

Technical Field

[0001] The invention relates to water conservancy engineering, disaster emergency response, hydrometeorology, ecological environment and their interdisciplinary fields, and in particular to a method for determining an optimal time scale for regional drought impact. Background Art

[0002] Drought is a natural phenomenon that occurs when regional water deficit reaches a certain level. It covers the entire process from the occurrence to the disappearance of drought in time. There is spatial heterogeneity in the degree of drought in space, and it has obvious temporal and spatial characteristics. When conducting drought research, drought is mainly identified based on various drought indices, and then characteristic analysis is conducted at different time scales. With the increase of data sources, especially the popularization of multidimensional geographic information, satellite remote sensing and other spatial data, and the emergence and development of multidimensional spatiotemporal data analysis methods, drought characteristic analysis has gradually expanded from the earliest site-based spatiotemporal analysis to global spatiotemporal analysis, realizing the leap from point to surface in drought risk assessment, prediction and other characteristic research.

[0003] Although current drought research has made great progress, the time scale for analyzing drought spatiotemporal characteristics is still mainly based on personal experience or literature reference. For example, the time scales of days, ten-days, months, seasons, and years are selected to analyze drought characteristics identified based on different drought indices. The results reflect the drought characteristics at different time scales, but cannot characterize the best time scale for drought-affected areas. For example, when analyzing drought characteristics at the annual time scale, drought events at the ten-day time scale may be ignored. Therefore, whether it is drought identification or drought risk analysis and prediction, the first step is to determine the best time for drought impact. On this basis, the drought spatiotemporal characteristics further analyzed are more in line with the actual situation.

[0004] As a water shortage event, drought directly or indirectly affects the regional ecological environment quality and social and economic conditions. According to the law of plant growth, vegetation growth in the flood period is better than that in the water shortage period. Related studies have also shown that the partial correlation between rainfall and vegetation index is good, and the two are positively correlated. Therefore, from the perspective of rainfall affecting vegetation growth, based on the correlation between long-term series rainfall and vegetation index at different time scales, the optimal time for drought-affected areas is determined. The analysis method and calculation process have clear physical meanings, which can not only provide a scientific optimization method for determining the time scale in the analysis of drought spatiotemporal characteristics, but also provide reference for the optimization of time scales in other scientific studies involving the relationship between rainfall and vegetation. Summary of the invention

[0005] The problem to be solved by the present invention is to provide a method for determining the optimal time scale for regional drought impact, so as to make up for the deficiency that different time scales are mainly selected for feature analysis based on personal experience or references in existing drought risk assessment, prediction and other drought spatiotemporal feature studies, so as to avoid the inability to reflect local characteristics of drought due to too large a time scale, and the inability to reflect global characteristics of drought due to too small a time scale.

[0006] The present invention adopts the following technical solution: a method for determining the optimal time scale of regional drought impact, comprising the following steps:

[0007] Step S1, according to the spatial range S of the study area, screening the site type rainfall data or grid type rainfall data covering the study area, wherein the spatial range interpolated from the site type rainfall data and the spatial range of the grid type rainfall data both need to cover the entire study area;

[0008] Step S2, arranging the site type rainfall data or grid type rainfall data into a time scale T data set of a long time series, using the Thiessen polygon method based on the site type rainfall data or using the average value method based on the grid type rainfall data to calculate the time scale T regional average surface rainfall of the long time series in the study area;

[0009] Step S3, based on various multispectral satellite remote sensing data covering the study area, the normalized vegetation index of different periods is calculated, and the normalized vegetation index of the minimum time scale is synthesized using the average method to synthesize the normalized vegetation index data set of the time scale T of the long series, and then the average method is used to calculate the regional average normalized vegetation index of the time scale T of the long series;

[0010] Step S4, based on the station type rainfall data and the grid type rainfall data, respectively calculating the first order difference of the average surface rainfall in the time scale T region and the first order difference of the average normalized vegetation index in the time scale T region of the long time series, and performing correlation analysis;

[0011] Step S5, based on the correlation coefficient value obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area.

[0012] Furthermore, in step S1, the spatial range of the interpolated station-type rainfall data and the spatial range of the grid-type rainfall data both need to cover the entire study area; specifically:

[0013] When selecting station-type rainfall data, Thiessen polygon interpolation is performed according to the spatial position of the station. The Thiessen polygon interpolated by the minimum number of stations includes the spatial range S of the study area. The number of stations including the spatial range S of the study area is the minimum number of stations required. When selecting raster-type rainfall data, the raster data range based on the minimum grid unit includes the spatial range S of the study area, and the minimum grid is a square grid.

[0014] Furthermore, in step S2, the Thiessen polygon method is used to calculate the average surface rainfall of the time scale T in the study area based on the rainfall data of the station type. The specific steps are as follows:

[0015] Step S21: according to the discretely distributed sites Automatically construct Delaunay triangulation for sites And the corresponding triangles are numbered;

[0016] Step S22, draw perpendicular bisectors for the three sides of the triangle respectively, the point where the perpendicular bisectors intersect is the center of the circumscribed circle of the triangle, and connect the center of the circumscribed circle with the midpoints of the three sides;

[0017] Step S23, executing step S22 for each triangle in the Delaunay triangulation network one by one according to the number, finding the center of the circumscribed circle of each triangle, and drawing a line connecting the center of the circumscribed circle and the midpoints of the three sides of the corresponding triangle;

[0018] Step S24: The polygon formed by the lines connecting the centers of all circumscribed circles and the midpoints of the three sides of the triangle is the Thiessen polygon. The centers of each circumscribed circle are the inflection points of the Thiessen polygon. Each Thiessen polygon contains a unique corresponding site. , number the Thiessen polygons according to the site numbers;

[0019] Step S25, using the spatial range of the study area to clip the Thiessen polygons, according to the proportion of the area of ​​each Thiessen polygon in the spatial range of the study area to the total area of ​​the study area, the weight coefficient of each Thiessen polygon in calculating the regional average surface rainfall is obtained, and the rainfall data of each station type is multiplied by the weight coefficient and then summed to obtain the regional average surface rainfall of time scale T. The specific calculation formula is:

[0020] (1);

[0021] (2);

[0022] (3);

[0023] In the formula, The weight coefficient for calculating the average rainfall in the region for the ith Thiessen polygon, is the area of ​​the ith Thiessen polygon in the study area, is the total area of ​​the study area, n is the number of Thiessen polygons in the study area, ; is the average surface rainfall in area T at time t, is the rainfall value at the kth station at time scale T at time t.

[0024] Furthermore, in step S2, the average rainfall of the time scale T in the study area is calculated based on the grid-type rainfall data using the average method; see formula (4) for details:

[0025] (4);

[0026] In the formula, is the average surface rainfall in area T at time t, is the rainfall value of the g-th grid at time scale T at time t, and m is the number of grid cells in the study area.

[0027] Furthermore, in step S3, the vegetation index dataset is normalized, specifically:

[0028] The normalized vegetation index of all days in a week or ten days is averaged to obtain the normalized vegetation index of the weekly scale or ten days. The normalized vegetation index of all days, weeks or ten days in a month is averaged to obtain the normalized vegetation index of the monthly scale. The normalized vegetation index of all days, weeks, ten days or months in a season is averaged to obtain the normalized vegetation index of the seasonal scale. The normalized vegetation index of all days, weeks, ten days, months or seasons in a year is averaged to obtain the normalized vegetation index of the annual scale. The synthesis process of normalized vegetation index of other larger time scales is similar.

[0029] The NDVI dataset of time scale T is synthesized based on the NDVI of the minimum time scale using the average method, and then the regional average NDVI of time scale T is calculated using the average method. The specific steps are as follows:

[0030] Step S31, calculate the normalized vegetation index of the minimum time scale a, the calculation formula is as follows:

[0031] (5);

[0032] In the formula, is the normalized vegetation index at the minimum time scale a, It is the near infrared band in multispectral satellite remote sensing data. It is the red band in multispectral satellite remote sensing data;

[0033] Step S32, using the average method to synthesize the normalized vegetation index of time scale b based on the normalized vegetation index of the minimum time scale a, the calculation formula is as follows:

[0034] (6);

[0035] In the formula, is the normalized vegetation index at time scale b, is the Xth normalized vegetation index of the minimum time scale a, X is the Xth time scale a within the range of time scale b, p is the number of data of the minimum time scale a contained in time scale b, and b>a;

[0036] Step S33, using the average method to calculate the regional average normalized vegetation index at time t and time scale T, the calculation formula is:

[0037] (7);

[0038] In the formula, is the regional average normalized vegetation index value at time scale T at time t, is the normalized vegetation index value of the xth smallest grid at time scale T at time t, and q is the number of grids of the regional normalized vegetation index.

[0039] Furthermore, in step S4, the average normalized difference vegetation index refers to the average value of all grids in the region at a certain moment; the first-order difference value is calculated by subtracting the difference between the value at the next moment and the value at the previous moment on a long time series of a certain time scale.

[0040] Furthermore, in step S4, the specific steps are as follows:

[0041] Step S41, respectively calculate the first-order difference value of the regional average surface rainfall and the first-order difference value of the regional average normalized vegetation index, and the calculation formula is:

[0042] (8);

[0043] (9);

[0044] In the formula, is the first-order difference value of the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t-1; is the first-order difference value of the regional average normalized vegetation index at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t-1;

[0045] Step S42, the Pearson correlation coefficient of the first-order difference series of the average surface rainfall in the time scale T region and the first-order difference series of the average normalized vegetation index in the time scale T region is calculated by the formula:

[0046] (10);

[0047] In the formula, is the Pearson correlation coefficient between the first-order difference series of average surface rainfall in the time scale T region and the first-order difference series of average normalized vegetation index in the time scale T region; t is the t-th moment in the long time series, and u is the total number of t moments in the long time series; is the first-order difference mean of the average surface rainfall in the time scale T of the long series, is the mean of the first-order difference of the normalized vegetation index averaged over the time scale T of the long series;

[0048] Step S43, the first-order difference mean of the average surface rainfall in the time scale T region of the long time series and the first-order difference mean of the average normalized vegetation index in the time scale T region of the long time series are calculated by the following formulas:

[0049] (11);

[0050] (12).

[0051] Further, in step S5, based on the value of the correlation coefficient obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area; the specific steps are:

[0052] Step S51, based on the Pearson correlation coefficient of the first-order difference series of average surface rainfall in the time scale T region of the long time series obtained in step S42 and the first-order difference series of average normalized vegetation index in the time scale T region of the long time series, compare the size of the Pearson correlation coefficient;

[0053] Step S52, determining the optimal time scale of drought impact in the study area, that is, the time scale corresponding to the maximum Pearson correlation coefficient value is the optimal time scale of drought impact in the study area;

[0054] Step S53, compare and analyze the Pearson correlation coefficients between the long-term series of regional average surface rainfall first-order difference series and the long-term series of regional average normalized vegetation index first-order difference series at different time scales in the study area. The time scale with the maximum Pearson correlation coefficient is the optimal time scale. The judgment process is:

[0055] , (13);

[0056] In the formula, is the maximum value among all Pearson correlation coefficients. , , …, is the corresponding time scale T 1 、T 2 ,…,T w The Pearson correlation coefficient between the long-term series of the first-order difference series of the regional average surface rainfall and the long-term series of the first-order difference series of the regional average normalized vegetation index, This is the optimal time scale.

[0057] Furthermore, the time scale T is day, week, ten-day period, month, season, year, etc. The normalized vegetation index has spatial characteristics, and the average value method is used to calculate the normalized vegetation index of different spatial positions at different times as the average normalized vegetation index of the study area at different times.

[0058] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0059] (1) Starting from the perspective of rainfall affecting vegetation growth, the present invention determines the optimal time for drought-affected areas based on the correlation between long-term rainfall series and vegetation indices at different time scales. The analysis method and calculation process have clear physical meanings, which can not only provide a scientific optimization method for determining the time scale in drought spatiotemporal characteristic analysis, but also provide a reference for the optimization of time scales in other scientific studies involving the relationship between rainfall and vegetation.

[0060] (2) The present invention can eliminate the abnormal fluctuations of the surface rainfall and the normalized vegetation index on the time axis by taking the first-order difference of the long-term series of rainfall and vegetation index and then performing correlation analysis. By evaluating the correlation between the changes, the mechanism is consistent with the change in vegetation growth caused by rainfall changes, so that the obtained time scale is more in line with the actual situation and can more accurately reflect the time scale of the drought-affected area. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is a flow chart of a method for determining an optimal drought impact time scale for a region according to the present invention;

[0062] Figure 2 It is a schematic diagram of Thiessen polygons and a study area interpolated based on sites in an embodiment of the present invention;

[0063] Figure 3 It is a schematic diagram of raster data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0064] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the application is further elaborated in detail below in conjunction with the accompanying drawings. The described embodiments are only a part of the embodiments involved in the present invention. All non-innovative embodiments of other researchers in the field on this embodiment belong to the protection scope of the present invention. At the same time, for the step numbering in the embodiment of the present invention, it is only set for the convenience of explanation, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0065] like Figure 1 As shown, the present invention adopts the following technical solution: a method for determining the optimal time scale of regional drought impact, comprising the following steps:

[0066] Step S1, according to the spatial range S of the study area, screening the site type rainfall data or grid type rainfall data covering the study area, wherein the spatial range interpolated from the site type rainfall data and the spatial range of the grid type rainfall data both need to cover the entire study area;

[0067] Step S2, arranging the site type rainfall data or grid type rainfall data into a time scale T data set of a long time series, using the Thiessen polygon method based on the site type rainfall data or using the average value method based on the grid type rainfall data to calculate the time scale T regional average surface rainfall of the long time series in the study area;

[0068] Step S3, based on various multispectral satellite remote sensing data covering the study area, the normalized vegetation index of different periods is calculated, and the normalized vegetation index of the minimum time scale is synthesized using the average method to synthesize the normalized vegetation index data set of the time scale T of the long series, and then the average method is used to calculate the regional average normalized vegetation index of the time scale T of the long series;

[0069] Step S4, based on the station type rainfall data and the grid type rainfall data, respectively calculating the first order difference of the average surface rainfall in the time scale T region and the first order difference of the average normalized vegetation index in the time scale T region of the long time series, and performing correlation analysis;

[0070] Step S5, based on the correlation coefficient value obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area.

[0071] Furthermore, in step S1, the spatial range of the interpolated station-type rainfall data and the spatial range of the grid-type rainfall data both need to cover the entire study area; specifically:

[0072] When selecting station-type rainfall data, Thiessen polygon interpolation is performed according to the spatial position of the station. The Thiessen polygon interpolated by the minimum number of stations includes the spatial range S of the study area. The number of stations including the spatial range S of the study area is the minimum number of stations required. When selecting raster-type rainfall data, the raster data range based on the minimum grid unit includes the spatial range S of the study area, and the minimum grid is a square grid.

[0073] Further, such as Figure 2 As shown, in step S2, the Thiessen polygon method is used to calculate the average surface rainfall of the time scale T area in the study area based on the rainfall data of the station type; the specific steps are as follows:

[0074] Step S21: according to the discretely distributed sites Automatically construct Delaunay triangulation for sites And the corresponding triangles are numbered;

[0075] Step S22, draw perpendicular bisectors for the three sides of the triangle respectively, the point where the perpendicular bisectors intersect is the center of the circumscribed circle of the triangle, and connect the center of the circumscribed circle with the midpoints of the three sides;

[0076] Step S23, executing step S22 for each triangle in the Delaunay triangulation network one by one according to the number, finding the center of the circumscribed circle of each triangle, and drawing a line connecting the center of the circumscribed circle and the midpoints of the three sides of the corresponding triangle;

[0077] Step S24: The polygon formed by the lines connecting the centers of all circumscribed circles and the midpoints of the three sides of the triangle is the Thiessen polygon. The centers of each circumscribed circle are the inflection points of the Thiessen polygon. Each Thiessen polygon contains a unique corresponding site. , number the Thiessen polygons according to the site numbers;

[0078] Step S25, using the spatial range of the study area to clip the Thiessen polygons, according to the proportion of the area of ​​each Thiessen polygon in the spatial range of the study area to the total area of ​​the study area, the weight coefficient of each Thiessen polygon in calculating the regional average surface rainfall is obtained, and the rainfall data of each station type is multiplied by the weight coefficient and then summed to obtain the regional average surface rainfall of time scale T. The specific calculation formula is:

[0079] (1);

[0080] (2);

[0081] (3);

[0082] In the formula, The weight coefficient for calculating the average rainfall in the region for the ith Thiessen polygon, is the area of ​​the ith Thiessen polygon in the study area, is the total area of ​​the study area, n is the number of Thiessen polygons in the study area, ; is the average surface rainfall in area T at time t, is the rainfall value at the kth station at time scale T at time t.

[0083] Further, in step S2, if Figure 3 As shown in the figure, the average rainfall in the study area with a time scale of T is calculated based on the grid-type rainfall data using the average method; see formula (4) for details:

[0084] (4);

[0085] In the formula, is the average surface rainfall in area T at time t, is the rainfall value of the g-th grid at time scale T at time t, and m is the number of grid cells in the study area.

[0086] Furthermore, in step S3, the vegetation index dataset is normalized, specifically:

[0087] The normalized vegetation index of all days in a week or ten days is averaged to obtain the normalized vegetation index of the weekly scale or ten days. The normalized vegetation index of all days, weeks or ten days in a month is averaged to obtain the normalized vegetation index of the monthly scale. The normalized vegetation index of all days, weeks, ten days or months in a season is averaged to obtain the normalized vegetation index of the seasonal scale. The normalized vegetation index of all days, weeks, ten days, months or seasons in a year is averaged to obtain the normalized vegetation index of the annual scale. The synthesis process of normalized vegetation index of other larger time scales is similar.

[0088] The NDVI dataset of time scale T is synthesized based on the NDVI of the minimum time scale using the average method, and then the regional average NDVI of time scale T is calculated using the average method. The specific steps are as follows:

[0089] Step S31, calculate the normalized vegetation index of the minimum time scale a, the calculation formula is as follows:

[0090] (5);

[0091] In the formula, is the normalized vegetation index at the minimum time scale a, It is the near infrared band in multispectral satellite remote sensing data. It is the red band in multispectral satellite remote sensing data;

[0092] Step S32, using the average method to synthesize the normalized vegetation index of time scale b based on the normalized vegetation index of the minimum time scale a, the calculation formula is as follows:

[0093] (6);

[0094] In the formula, is the normalized vegetation index at time scale b, is the Xth normalized vegetation index of the minimum time scale a, X is the Xth time scale a within the range of time scale b, p is the number of data of the minimum time scale a contained in time scale b, and b>a;

[0095] Step S33, using the average method to calculate the regional average normalized vegetation index at time t and time scale T, the calculation formula is:

[0096] (7);

[0097] In the formula, is the regional average normalized vegetation index value at time scale T at time t, is the normalized vegetation index value of the xth smallest grid at time scale T at time t, and q is the number of grids of the regional normalized vegetation index.

[0098] Furthermore, in step S4, the average normalized difference vegetation index refers to the average value of all grids in the region at a certain moment; the first-order difference value is calculated by subtracting the difference between the value at the next moment and the value at the previous moment on a long time series of a certain time scale.

[0099] Furthermore, in step S4, the specific steps are as follows:

[0100] Step S41, respectively calculate the first-order difference value of the regional average surface rainfall and the first-order difference value of the regional average normalized vegetation index, and the calculation formula is:

[0101] (8);

[0102] (9);

[0103] In the formula, is the first-order difference value of the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t-1; is the first-order difference value of the regional average normalized vegetation index at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t-1;

[0104] Step S42, the Pearson correlation coefficient of the first-order difference series of the average surface rainfall in the time scale T region and the first-order difference series of the average normalized vegetation index in the time scale T region is calculated by the formula:

[0105] (10);

[0106] In the formula, is the Pearson correlation coefficient between the first-order difference series of average surface rainfall in the time scale T region and the first-order difference series of average normalized vegetation index in the time scale T region; t is the t-th moment in the long time series, and u is the total number of t moments in the long time series; is the first-order difference mean of the average surface rainfall in the time scale T of the long series, is the mean of the first-order difference of the normalized vegetation index averaged over the time scale T of the long series;

[0107] Step S43, the first-order difference mean of the average surface rainfall in the time scale T region of the long time series and the first-order difference mean of the average normalized vegetation index in the time scale T region of the long time series are calculated by the following formulas:

[0108] (11);

[0109] (12).

[0110] Further, in step S5, based on the value of the correlation coefficient obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area; the specific steps are:

[0111] Step S51, based on the Pearson correlation coefficient of the first-order difference series of average surface rainfall in the time scale T region of the long time series obtained in step S42 and the first-order difference series of average normalized vegetation index in the time scale T region of the long time series, compare the size of the Pearson correlation coefficient;

[0112] Step S52, determining the optimal time scale of drought impact in the study area, that is, the time scale corresponding to the maximum Pearson correlation coefficient value is the optimal time scale of drought impact in the study area;

[0113] Step S53, compare and analyze the Pearson correlation coefficients between the long-term series of regional average surface rainfall first-order difference series and the long-term series of regional average normalized vegetation index first-order difference series at different time scales in the study area. The time scale with the maximum Pearson correlation coefficient is the optimal time scale. The judgment process is:

[0114] , (13);

[0115] In the formula, is the maximum value among all Pearson correlation coefficients. , , …, is the corresponding time scale T 1 、T 2 ,…,T w The Pearson correlation coefficient between the long-term series of the first-order difference series of the regional average surface rainfall and the long-term series of the first-order difference series of the regional average normalized vegetation index, This is the optimal time scale.

[0116] Furthermore, the time scale T is day, week, ten-day period, month, season, year, etc. The normalized vegetation index has spatial characteristics, and the average value method is used to calculate the normalized vegetation index of different spatial positions at different times as the average normalized vegetation index of the study area at different times.

[0117] In summary, a method for determining the optimal time scale of regional drought impact in this embodiment, from the perspective of the impact of drought and water shortage on the growth of various vegetation in the ecosystem, based on the correlation coefficient between the first-order difference series of regional average surface rainfall of different time scales and the first-order difference series of regional average normalized vegetation index of the long time series, determines the optimal time scale of the drought and water shortage research area, and the time scale corresponding to the maximum correlation coefficient is the optimal time scale of regional drought impact, indicating that the impact of drought and water shortage on the regional ecosystem at this time scale is significantly greater than the impact of drought and water shortage on the regional ecosystem at other time scales. The analysis method and calculation process used in the present invention have clear physical meanings and are consistent with the actual situation. At the same time, the method for determining the optimal regional drought impact time scale proposed in the method of the present invention can be used for the selection of time scales in the study of drought spatiotemporal characteristics analysis, such as the determination of time scales in the study of drought risk assessment, drought characteristics analysis, drought risk prediction, etc., and can also provide reference for the optimization of time scales in other scientific studies involving the relationship between rainfall and vegetation, and further meet the scientific research and application needs in hydrology, water resources, ecology, meteorology, and emergency management.

[0118] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for determining the optimal time scale for regional drought impacts, characterized by: The steps include: Step S1, according to the spatial range S of the study area, screening the site type rainfall data or grid type rainfall data covering the study area, wherein the spatial range interpolated from the site type rainfall data and the spatial range of the grid type rainfall data both need to cover the entire study area; Step S2, arranging the site type rainfall data or grid type rainfall data into a time scale T data set of a long time series, using the Thiessen polygon method based on the site type rainfall data or using the average value method based on the grid type rainfall data to calculate the time scale T regional average surface rainfall of the long time series in the study area; Step S3, based on various multispectral satellite remote sensing data covering the study area, the normalized vegetation index of different periods is calculated, and the normalized vegetation index of the minimum time scale is synthesized using the average method to synthesize the normalized vegetation index data set of the time scale T of the long series, and then the average method is used to calculate the regional average normalized vegetation index of the time scale T of the long series; Step S4, based on the station type rainfall data and the grid type rainfall data, respectively calculating the first order difference of the average surface rainfall in the time scale T region and the first order difference of the average normalized vegetation index in the time scale T region of the long time series, and performing correlation analysis; Step S5, based on the value of the correlation coefficient obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area; In step S4, the average normalized vegetation index refers to the average value of all grids in the region at a certain time; the first-order difference value is calculated by subtracting the difference between the value at the next moment and the value at the previous moment on a long time series of a certain time scale; In step S4, the specific steps are as follows: Step S41, respectively calculate the first-order difference value of the regional average surface rainfall and the first-order difference value of the regional average normalized vegetation index, and the calculation formula is: （8）； （9）； In the formula, is the first-order difference value of the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t, is the regional average surface rainfall at time scale T at time t-1; is the first-order difference value of the regional average normalized vegetation index at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t, is the regional average normalized vegetation index value at time scale T at time t-1; Step S42, the Pearson correlation coefficient of the first-order difference series of the average surface rainfall in the time scale T region and the first-order difference series of the average normalized vegetation index in the time scale T region is calculated by the formula: （10）； In the formula, is the Pearson correlation coefficient between the first-order difference series of average surface rainfall in the time scale T region and the first-order difference series of average normalized vegetation index in the time scale T region; t is the t-th moment in the long time series, and u is the total number of t moments in the long time series; is the first-order difference mean of the average surface rainfall in the time scale T of the long series, is the mean of the first-order difference of the normalized vegetation index averaged over the time scale T of the long series; Step S43, the first-order difference mean of the average surface rainfall in the time scale T region of the long time series and the first-order difference mean of the average normalized vegetation index in the time scale T region of the long time series are calculated by the following formulas: （11）； （12）； Step S5, based on the value of the correlation coefficient obtained by the correlation analysis in step S4, the optimal time scale of drought impact is identified, and the time scale corresponding to the maximum value of the correlation coefficient is the optimal time scale of drought impact in the study area; the specific steps are: Step S51, based on the Pearson correlation coefficient of the first-order difference series of average surface rainfall in the time scale T region of the long time series obtained in step S42 and the first-order difference series of average normalized vegetation index in the time scale T region of the long time series, compare the size of the Pearson correlation coefficient; Step S52, determining the optimal time scale of drought impact in the study area, that is, the time scale corresponding to the maximum Pearson correlation coefficient value is the optimal time scale of drought impact in the study area; Step S53, compare and analyze the Pearson correlation coefficients between the long-term series of regional average surface rainfall first-order difference series and the long-term series of regional average normalized vegetation index first-order difference series at different time scales in the study area. The time scale with the maximum Pearson correlation coefficient is the optimal time scale. The judgment process is: ， (13)； In the formula, is the maximum value among all Pearson correlation coefficients. , , …, is the corresponding time scale T1, T2, ..., T w The Pearson correlation coefficient between the long-term series of the first-order difference series of the regional average surface rainfall and the long-term series of the first-order difference series of the regional average normalized vegetation index, This is the optimal time scale.

2. A method for determining the optimal time scale of regional drought impact according to claim 1, characterized in that: In step S1, the spatial range of the interpolated station-type rainfall data and the spatial range of the grid-type rainfall data both need to cover the entire study area; specifically: When selecting station-type rainfall data, Thiessen polygon interpolation is performed according to the spatial position of the station. The Thiessen polygon interpolated by the minimum number of stations includes the spatial range S of the study area. The number of stations including the spatial range S of the study area is the minimum number of stations required. When selecting raster-type rainfall data, the raster data range based on the minimum grid unit includes the spatial range S of the study area, and the minimum grid is a square grid.

3. A method for determining the optimal time scale of regional drought impact according to claim 2, characterized in that: In step S2, the Thiessen polygon method is used to calculate the average surface rainfall of the time scale T in the study area based on the rainfall data of the station type. The specific steps are as follows: Step S21: according to the discretely distributed sites Automatically construct Delaunay triangulation network for site And the corresponding triangles are numbered; Step S22, draw perpendicular bisectors for the three sides of the triangle respectively, the point where the perpendicular bisectors intersect is the center of the circumscribed circle of the triangle, and connect the center of the circumscribed circle with the midpoints of the three sides; Step S23, executing step S22 for each triangle in the Delaunay triangulation network one by one according to the number, finding the center of the circumscribed circle of each triangle, and drawing a line connecting the center of the circumscribed circle and the midpoints of the three sides of the corresponding triangle; Step S24: The polygon formed by the lines connecting the centers of all circumscribed circles and the midpoints of the three sides of the triangle is the Thiessen polygon. The centers of each circumscribed circle are the inflection points of the Thiessen polygon. Each Thiessen polygon contains a unique corresponding site. , number the Thiessen polygons according to the site numbers; Step S25, using the spatial range of the study area to clip the Thiessen polygons, according to the proportion of the area of ​​each Thiessen polygon in the spatial range of the study area to the total area of ​​the study area, the weight coefficient of each Thiessen polygon in calculating the regional average surface rainfall is obtained, and the rainfall data of each station type is multiplied by the weight coefficient and then summed to obtain the regional average surface rainfall of time scale T. The specific calculation formula is: （1）； （2）； （3）； In the formula, The weight coefficient for calculating the average rainfall in the region for the ith Thiessen polygon, is the area of ​​the ith Thiessen polygon in the study area, is the total area of ​​the study area, n is the number of Thiessen polygons in the study area, ; is the average surface rainfall in area T at time t, is the rainfall value at the kth station at time scale T at time t.

4. The method for determining the optimal time scale of regional drought impact according to claim 3, characterized in that: In step S2, the average rainfall of the time scale T in the study area is calculated based on the grid-type rainfall data using the average method; see formula (4) for details: （4）； In the formula, is the average surface rainfall in area T at time t, is the rainfall value of the g-th grid at time scale T at time t, and m is the number of grid cells in the study area.

5. A method for determining the optimal time scale of regional drought impact according to claim 4, characterized in that: In step S3, the vegetation index dataset is normalized, specifically: The normalized vegetation index of all days in a week or ten days is averaged to obtain the normalized vegetation index of the weekly scale or ten days. The normalized vegetation index of all days, weeks or ten days in a month is averaged to obtain the normalized vegetation index of the monthly scale. The normalized vegetation index of all days, weeks, ten days or months in a quarter is averaged to obtain the normalized vegetation index of the seasonal scale. The normalized vegetation index of all days, weeks, ten days, months or seasons in a year is averaged to obtain the normalized vegetation index of the annual scale. The NDVI dataset of time scale T is synthesized based on the NDVI of the minimum time scale using the mean value method, and then the regional average NDVI of time scale T is calculated using the mean value method. The specific steps are: Step S31, calculate the normalized vegetation index of the minimum time scale a, the calculation formula is as follows: （5）； In the formula, is the normalized vegetation index at the minimum time scale a, It is the near infrared band in multispectral satellite remote sensing data. It is the red band in multispectral satellite remote sensing data; Step S32, using the average method to synthesize the normalized vegetation index of time scale b based on the normalized vegetation index of the minimum time scale a, the calculation formula is as follows: （6）； In the formula, is the normalized vegetation index at time scale b, is the Xth normalized vegetation index of the minimum time scale a, X is the Xth time scale a within the range of time scale b, p is the number of data of the minimum time scale a contained in time scale b, and b>a; Step S33, using the average method to calculate the regional average normalized vegetation index at time t and time scale T, the calculation formula is: （7）； In the formula, is the regional average normalized vegetation index value at time scale T at time t, is the normalized vegetation index value of the xth smallest grid at time scale T at time t, and q is the number of grids of the regional normalized vegetation index.

Images