A regional drought remote sensing comprehensive index construction and monitoring method

Abstract

The application provides a regional drought remote sensing comprehensive index construction and monitoring method, and relates to the field of drought remote sensing monitoring, and the method comprises the following steps: based on time series remote sensing data, determining the ground temperature data of a target region, and then obtaining normalized ground temperature data; by determining the normalized vegetation index and the target section where the normalized vegetation index is located, estimating the vegetation coverage data of the target region; using the 1km resolution resampling data of the target region and a soil moisture estimation model, obtaining 1km resolution soil moisture data; based on the normalized ground temperature data, the vegetation coverage data and the 1km resolution soil moisture data, constructing a regional drought index model. The application selects different vegetation coverage estimation methods for different normalized vegetation indexes of different sections, so that the vegetation coverage data estimation is more accurate, and the combination of the 1km resolution soil moisture data makes the performance of the regional drought index model more optimal, and then the estimation accuracy of the drought index is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of drought remote sensing monitoring, and in particular to a regional drought remote sensing comprehensive index construction and monitoring method. BACKGROUND

[0002] The Euclidean distance algorithm can be used to reflect drought conditions, but there are uncertainties and limitations in estimating regional drought conditions. For example, using the same estimation method in different vegetation coverage areas will lead to inaccurate remote sensing estimation of vegetation coverage, and further lead to inaccurate drought estimation. SUMMARY

[0003] The present application provides a regional drought remote sensing comprehensive index construction and monitoring method to solve the technical problem of inaccurate drought estimation.

[0004] In a first aspect, the present application provides a regional drought remote sensing comprehensive index construction method, comprising:

[0005] Obtain time series remote sensing data of a target region, and based on the time series remote sensing data, determine land surface temperature data of the target region, normalize the land surface temperature data to obtain normalized land surface temperature data;

[0006] Based on the time series remote sensing data, determine the normalized vegetation index of the target region, and determine the target section where the normalized vegetation index is located, based on the pre-determined vegetation coverage estimation method of the target section, estimate the vegetation coverage data of the target region;

[0007] Based on the time series remote sensing data, determine the 1km resolution resampling data of the target region, and input the resampling data into a soil moisture estimation model to obtain the 1km resolution soil moisture data of the target region output by the soil moisture estimation model; the resampling data includes land surface temperature data, normalized vegetation index data, reflectivity data, digital elevation data and precipitation data;

[0008] Based on the normalized land surface temperature data, the vegetation coverage data and the 1km resolution soil moisture data, a regional drought index model is constructed.

[0009] According to the regional drought remote sensing comprehensive index construction method provided by the present application, the land surface temperature data is normalized to obtain normalized land surface temperature data, which comprises:

[0010] Based on the time series remote sensing data, determine the maximum temperature and the minimum temperature of the target region;

[0011] Based on the maximum temperature and the minimum temperature, the land surface temperature data is normalized to obtain the normalized land surface temperature data.

[0012] According to the regional drought remote sensing comprehensive index construction method provided by the application, the vegetation coverage data of the target region is estimated based on the vegetation coverage estimation method of the target section determined in advance, and the vegetation coverage estimation method comprises the following steps:

[0013] If the normalized difference vegetation index of the target region is greater than or equal to a first threshold value and less than or equal to a second threshold value, the minimum normalized difference vegetation index and the maximum normalized difference vegetation index of the target region are calculated;

[0014] Based on the first threshold value, the minimum normalized difference vegetation index and the maximum normalized difference vegetation index, the first vegetation coverage is calculated when the normalized difference vegetation index is equal to the first threshold value;

[0015] Based on the second threshold value, the minimum normalized difference vegetation index and the maximum normalized difference vegetation index, the second vegetation coverage is calculated when the normalized difference vegetation index is equal to the second threshold value;

[0016] Based on the normalized difference vegetation index, the minimum normalized difference vegetation index and the maximum normalized difference vegetation index, the vegetation coverage data of the target region is estimated.

[0017] According to the regional drought remote sensing comprehensive index construction method provided by the application, the vegetation coverage data of the target region is estimated based on the vegetation coverage estimation method of the target section determined in advance, and the vegetation coverage estimation method comprises the following steps:

[0018] If the normalized difference vegetation index of the target region is less than a first threshold value, the difference value vegetation index is determined according to the difference between the near-infrared band reflectance and the red band reflectance;

[0019] According to the difference value vegetation index, the minimum difference value vegetation index, the normalized difference vegetation index equal to the first threshold value and the first vegetation coverage, the vegetation coverage data of the target region is estimated.

[0020] According to the regional drought remote sensing comprehensive index construction method provided by the application, the vegetation coverage data of the target region is estimated based on the vegetation coverage estimation method of the target section determined in advance, and the vegetation coverage estimation method comprises the following steps:

[0021] If the normalized difference vegetation index of the target region is greater than a second threshold value, an enhanced vegetation index is determined;

[0022] According to the second vegetation coverage, the enhanced vegetation index, the enhanced vegetation index maximum value, and the normalized vegetation index value when the normalized vegetation index is the second threshold, the vegetation coverage data of the target area is estimated.

[0023] According to the present application, a regional drought remote sensing comprehensive index construction method is provided, and the enhanced vegetation index is determined, including:

[0024] A first product is determined according to the red band reflectivity and a first parameter;

[0025] A second product is determined according to the blue band reflectivity and a second parameter;

[0026] A first sum is determined according to the near-infrared band reflectivity, the first product, the second product, and a preset constant;

[0027] The enhanced vegetation index is determined according to the quotient of the difference between the near-infrared band reflectivity and the red band reflectivity and the first sum.

[0028] According to the present application, a regional drought remote sensing comprehensive index construction method is provided, and the enhanced vegetation index is determined, including:

[0029] The resolution scale corresponding to the current land surface temperature data, the resolution scale corresponding to the current normalized vegetation index, the resolution scale corresponding to the current reflectivity, the resolution scale corresponding to the current digital elevation model data, and the resolution scale corresponding to the current precipitation data are uniformly improved to a first resolution scale corresponding to the current microwave radiometer data, so as to construct a soil moisture estimation model according to the current land surface temperature data, the current normalized vegetation index, the current reflectivity data, the current digital elevation model data, the current precipitation data, and the current microwave radiometer data after the resolution scale is improved;

[0030] The resampling is performed to a second resolution scale, and the resampled land surface temperature data, the resampled normalized vegetation index, the resampled reflectivity data, the resampled digital elevation model data, and the resampled precipitation data are input into the soil moisture estimation model to obtain soil monitoring data corresponding to the second resolution scale output by the soil moisture estimation model;

[0031] The soil moisture data is determined according to the soil monitoring data;

[0032] The first resolution scale is greater than the second resolution scale;

[0033] The second resolution scale is 1 km resolution.

[0034] The method for constructing a regional drought remote sensing comprehensive index provided by the application, the soil monitoring data includes soil moisture with maximum humidity in a monitoring area, soil moisture with minimum humidity in the monitoring area, and soil moisture monitoring data;

[0035] The soil moisture data is determined according to the soil monitoring data, including:

[0036] A first soil moisture difference value is determined according to the soil moisture monitoring data and the soil moisture with minimum humidity in the monitoring area;

[0037] A second soil moisture difference value is determined according to the soil moisture with maximum humidity in the monitoring area and the soil moisture with minimum humidity in the monitoring area;

[0038] The soil moisture data is determined according to the first soil moisture difference value and the second soil moisture difference value.

[0039] The method for constructing a regional drought remote sensing comprehensive index provided by the application, before time series remote sensing data of a target area is acquired, the method further includes:

[0040] According to the time series remote sensing data of any monitoring area, a crop identification index and a normalized vegetation index mean value of the monitoring area are determined;

[0041] In a case where the crop identification index is greater than or equal to a first preset threshold value and the normalized vegetation index mean value is greater than or equal to a second preset threshold value, the monitoring area is determined as the target area.

[0042] In a second aspect, a regional drought remote sensing comprehensive index monitoring method based on the method for constructing a regional drought remote sensing comprehensive index is provided, including:

[0043] Acquiring initial land surface temperature data, initial soil moisture data and initial vegetation reflectivity data of a region to be monitored;

[0044] Inputting the initial land surface temperature data, the initial soil moisture data and the initial vegetation reflectivity data to the regional drought index model to acquire a drought remote sensing comprehensive index of the region to be monitored output by the regional drought index model.

[0045] The application provides a regional drought remote sensing comprehensive index construction and monitoring method, which comprises the following steps: determining normalized ground temperature data according to time sequence remote sensing data of a target region; estimating vegetation coverage data of the target region based on a predetermined vegetation coverage estimation method of the target section; obtaining 1km resolution soil moisture data output by a soil moisture estimation model based on the time sequence remote sensing data; and constructing a regional drought index model based on the normalized ground temperature data, the vegetation coverage data and the 1km resolution soil moisture data. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0047] Figure 1 is one of the flowcharts of the regional drought remote sensing comprehensive index construction method provided by the present application;

[0048] Figure 2 is a flowchart for obtaining normalized ground temperature data provided by the present application;

[0049] Figure 3 is a flowchart for estimating vegetation coverage data of the target region provided by the present application;

[0050] Figure 4 is a flowchart for obtaining soil moisture data of the target region provided by the present application;

[0051] Figure 5 is the second flowchart of the regional drought remote sensing comprehensive index construction method provided by the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of the present application more clear, the technical solutions of the present application will be described clearly and completely in the following with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0053] At present, the methods for obtaining the regional drought information of space-time distribution include meteorological observation, hydrological observation and remote sensing method, wherein the meteorological observation, hydrological observation and the like mainly obtain the drought condition through statistical analysis by means of instrument observation of meteorological or hydrological change, and the method is difficult to effectively reflect the soil moisture, vegetation drought and the like under the influence of topography, terrain, water source and the like when facing large heterogeneous space drought monitoring.

[0054] The remote sensing method has the advantages of objectivity, rapidness, low cost and the like, and the common methods are soil moisture observation method, vegetation canopy water observation method, temperature observation method, vegetation index observation method and temperature-vegetation index observation method, however, it is generally considered that the factors such as vegetation, temperature, soil moisture, precipitation and albedo can more comprehensively reflect the drought condition, and the method of adopting the Euclidean distance method is considered to be a more effective method.

[0055] However, when the Euclidean distance method is adopted, the soil moisture data involved is generally low in spatial resolution, for example, for a large area, the soil moisture inversion product with a resolution of 25km is adopted, the vegetation index obtained by inversion is easy to saturate, or for a low vegetation coverage area, it is not sensitive enough, and thus there is great uncertainty.

[0056] In order to solve the above technical problems, the present application provides a regional drought remote sensing comprehensive index construction and monitoring method, considering the uncertainty and limitation in the estimation of regional drought by the Euclidean distance method, the present application comprehensively considers the feasibility of the soil moisture downscaling and improved vegetation index method, and proposes a soil moisture down-sampling method by random forest and a vegetation coverage remote sensing estimation method of improved vegetation index, so as to achieve higher spatial resolution and more accurate estimation of drought, specifically, Figure 1 is one of the flowcharts of the regional drought remote sensing comprehensive index construction method provided by the present application, comprising:

[0057] Step 101, acquiring time series remote sensing data of a target region, and determining surface temperature data of the target region based on the time series remote sensing data, normalizing the surface temperature data to obtain normalized surface temperature data;

[0058] Step 102, determining normalized vegetation index of the target region based on the time series remote sensing data, and determining a target section where the normalized vegetation index is located, estimating vegetation coverage data of the target region based on a pre-determined vegetation coverage estimation method of the target section;

[0059] Step 103, determining 1km resolution resampling data of the target region based on the time series remote sensing data, and inputting the resampling data into a soil moisture estimation model to obtain 1km resolution soil moisture data of the target region output by the soil moisture estimation model; the resampling data includes land surface temperature data, normalized vegetation index data, reflectivity data, digital elevation data, and precipitation data;

[0060] Step 104, constructing a regional drought index model based on the normalized land surface temperature data, the vegetation coverage data, and the 1km resolution soil moisture data.

[0061] In step 101, Figure 2 is a flowchart for obtaining normalized land surface temperature data provided by the present application, as Figure 2 shown, the land surface temperature data is normalized to obtain normalized land surface temperature data, including:

[0062] Based on the time series remote sensing data, the maximum temperature and the minimum temperature of the target region are determined;

[0063] Based on the maximum temperature and the minimum temperature, the land surface temperature data is normalized to obtain the normalized land surface temperature data.

[0064] Optionally, according to the maximum temperature LST max of the target region and the actual land surface temperature LST max determined by remote sensing, the first temperature difference LST max -LST is determined; according to the maximum temperature LST min of the target region and the minimum temperature LST max of the target region, the second temperature difference LST min -LST is determined; according to the first temperature difference and the second temperature difference, the land surface temperature index, i.e. the normalized land surface temperature index, is determined, which can refer to the following formula:

[0065]

[0066] In formula (1), NLST is the land surface temperature index; LST max is the maximum temperature of the target region; LST min is the minimum temperature of the target region; LST is the actual land surface temperature determined by remote sensing.

[0067] In step 102, based on the time series remote sensing data, a normalized vegetation index of the target region is determined, and a target section where the normalized vegetation index is located is determined, based on a predetermined vegetation coverage estimation method of the target section, vegetation coverage data of the target region is estimated.

[0068] The determination of the normalized vegetation index of the target region based on the time series remote sensing data comprises: determining the current normalized vegetation index according to the sum and the difference of the near-infrared band reflectivity and the red band reflectivity.

[0069] Optionally, the sum of the near-infrared band reflectivity and the red band reflectivity is R nir +R r , and the difference between the near-infrared band reflectivity and the red band reflectivity is R nir -R r , and the determination of the current normalized vegetation index is based on the following formula:

[0070]

[0071] In formula (2), NDVI is the current normalized vegetation index, R nir is the near-infrared band reflectivity, and R r is the reflectivity of the red band.

[0072] Optionally, all vegetation coverage estimation methods are determined according to different coverage degrees of vegetation, each vegetation coverage estimation method corresponds to each section, the vegetation coverage estimation method of the target section is determined, and each section is determined after the value range of the normalized vegetation index is divided, that is, the target section where the current normalized vegetation index is located is first determined, and then based on the predetermined vegetation coverage estimation method of the target section, the vegetation coverage data of the target region is estimated.

[0073] Optionally, all the methods in the vegetation coverage estimation method take near-infrared reflectivity and red band reflectivity as processing objects, and finally obtain the target vegetation coverage, but due to different methods, the calculation formulas between different processing methods are different, and the present application determines different vegetation coverage estimation methods according to different coverage degrees of vegetation, for example, taking difference vegetation index (DVI), normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) as the vegetation index used for reconstructing the vegetation coverage, considering that the difference vegetation index DVI is mainly suitable for low-coverage vegetation area, the normalized difference vegetation index NDVI is suitable for medium-coverage vegetation area, and the enhanced vegetation index EVI is suitable for medium-high-coverage vegetation area, so that the estimation of the target vegetation coverage is more accurate.

[0074] In step 103, based on the time series remote sensing data, 1km resolution resampling data of the target area is determined, and the resampling data is input into a soil moisture estimation model to obtain 1km resolution soil moisture data of the target area output by the soil moisture estimation model; the resampling data includes land surface temperature data, normalized vegetation index data, reflectivity data, digital elevation model data and precipitation data.

[0075] Optionally, the land surface temperature data, the normalized vegetation index data, the reflectivity data, the digital elevation model data and the precipitation data are down-scaled, and combined with the microwave radiometer data to obtain down-scaled data, and the soil moisture data of the target area is determined based on the down-scaled data, the present application unifies the resolution scale of different data in the time series remote sensing data to the first resolution scale corresponding to the current microwave radiometer data, so as to construct the soil moisture estimation model according to the data after the scale is raised, down-scale the land surface temperature data, the normalized vegetation index data, the reflectivity data, the digital elevation model data and the precipitation data, and combine the microwave radiometer data to obtain the 1km resolution soil moisture data of the target area output by the soil moisture estimation model, and the first resolution scale is greater than the second resolution scale. Optionally, the soil moisture data of the target area determined based on the down-scaled data can combine the spatial down-sampling of the random forest to realize the purpose of accurately identifying regional drought.

[0076] In step 104, based on the normalized land surface temperature data, the vegetation coverage data and the soil moisture data, a regional drought index model is constructed, and the present application considers the three dimensions of the normalized land surface temperature data, the vegetation coverage data and the soil moisture data to determine a remote sensing drought index, and a regional drought index model is constructed, which can refer to the following formula:

[0077]

[0078] In formula (3), MTVMDI is the modified remote sensing drought index, NLST is the normalized land surface temperature data, NVCI is the vegetation coverage data, and NSM is the soil moisture data.

[0079] The present application provides a regional drought remote sensing comprehensive index construction and monitoring method, according to the time series remote sensing data of the target region, the normalized land surface temperature data is determined, based on the pre-determined vegetation coverage estimation method of the target section, the vegetation coverage data of the target region is estimated, based on the time series remote sensing data, the soil moisture data of 1km resolution output by the soil moisture estimation model is obtained; based on the normalized land surface temperature data, the vegetation coverage data and the 1km resolution soil moisture data, a regional drought index model is constructed, the present application selects different coverage estimation algorithms for different coverage areas, so that the target vegetation coverage estimation is more accurate, and the estimation accuracy of the drought index is improved.

[0080] Optionally, the present application segments and processes different coverage vegetation regions according to the normalized vegetation index, obtains different sections, determines different vegetation coverage estimation methods according to different coverage vegetation regions, each vegetation coverage estimation method corresponds to each section, and according to the normalized vegetation index of the target region, the target section where the normalized vegetation index is located is determined.

[0081] The present application determines the maximum value and the minimum value of the normalized vegetation index in the monitoring region according to all the normalized vegetation indexes obtained in the monitoring region, takes the maximum value and the minimum value of the normalized vegetation index as the value range, segments and processes different coverage vegetation regions, and obtains different sections.

[0082] Considering that the difference vegetation index DVI is mainly suitable for low-coverage vegetation regions, the normalized vegetation index NDVI is suitable for medium-coverage vegetation regions, and the enhanced vegetation index EVI is suitable for medium-high-coverage vegetation regions, the present application corresponds each vegetation coverage estimation method to each section, and different sections correspond to different coverage vegetation regions, so that different vegetation coverage estimation methods are determined according to different coverage vegetation regions.

[0083] According to the normalized vegetation index of the target region, a target section where the normalized vegetation index is located is determined, and the normalized vegetation index of the target region is taken as a judgment basis for determining the target section, so that the vegetation coverage estimation method adopted is more in line with the actual vegetation coverage, and the estimation is more accurate.

[0084] Figure 3 is a flowchart for estimating vegetation coverage data of the target region, and the vegetation coverage estimation method includes a difference vegetation index estimation method, a normalized vegetation index estimation method, and an enhanced vegetation index estimation method.

[0085] The vegetation coverage estimation method based on the target section is used to estimate the vegetation coverage data of the target region, including:

[0086] Step 301, if the normalized vegetation index of the target region is greater than or equal to a first threshold value and less than or equal to a second threshold value, the normalized vegetation index minimum value and the normalized vegetation index maximum value of the target region are calculated.

[0087] Step 302, based on the first threshold value, the normalized vegetation index minimum value and the normalized vegetation index maximum value, the first vegetation coverage when the normalized vegetation index takes the first threshold value is calculated.

[0088] Step 303, based on the second threshold value, the normalized vegetation index minimum value and the normalized vegetation index maximum value, the second vegetation coverage when the normalized vegetation index takes the second threshold value is calculated.

[0089] Step 304, based on the normalized vegetation index, the normalized vegetation index minimum value and the normalized vegetation index maximum value, the vegetation coverage data of the target region is estimated.

[0090] In step 301, the first threshold value is a1, and the second threshold value is a2, wherein a 1＜ a2, if the normalized vegetation index of the target region is greater than or equal to a1 and less than or equal to a2, the normalized vegetation index minimum value NDVI min and the normalized vegetation index maximum value NDVI max are calculated.

[0091] In step 302, based on the first threshold value, the normalized vegetation index minimum value and the normalized vegetation index maximum value, the first vegetation coverage when the normalized vegetation index takes the first threshold value is calculated, which can be referred to as the following formula:

[0092]

[0093] In formula (4), C a1 is the first vegetation coverage, NDVI a1 is the value of the normalized vegetation index when the first threshold value is reached, the NDVI min is the minimum value of the vegetation index, NDVI max is the maximum value of the vegetation index.

[0094] In step 303, based on the second threshold value, the minimum value of the normalized vegetation index, and the maximum value of the normalized vegetation index, the second vegetation coverage when the value of the normalized vegetation index reaches the second threshold value is calculated, which can be referred to as formula (5):

[0095]

[0096] In formula (5), C a2 is the second vegetation coverage, NDVI a2 is the value of the normalized vegetation index when the second threshold value is reached, the NDVI min is the minimum value of the vegetation index, NDVI max is the maximum value of the vegetation index.

[0097] In step 304, based on the normalized vegetation index, the minimum value of the normalized vegetation index, and the maximum value of the normalized vegetation index, the vegetation coverage data of the target area is estimated, which can be referred to as formula (6):

[0098]

[0099] In formula (6), C is the vegetation coverage data of the target area, NDVI is the current vegetation index, the NDVI min is the minimum value of the vegetation index, NDVI a1-a2 is the vegetation index when the value is between a1 and a2.

[0100] Optionally, the vegetation coverage estimation method is set to three vegetation coverage estimation methods, and correspondingly, three sections need to be set, i.e., the first vegetation coverage estimation method corresponding to the first section, the second vegetation coverage estimation method corresponding to the second section, and the third vegetation coverage estimation method corresponding to the section. In other embodiments, four vegetation coverage estimation methods or five vegetation coverage estimation methods can be set, and the corresponding sections are then given accordingly.

[0101] In the case that the normalized vegetation index of the target area is located in the first section, i.e., the current normalized vegetation index is located in the first section 〔0, a1〕, the vegetation coverage data of the target area is estimated based on the difference vegetation index.

[0102] If the normalized difference vegetation index of the target region is located in the second section [a1, a2], it is determined that the vegetation coverage data of the target region is estimated by using the normalized difference vegetation index.

[0103] If the normalized difference vegetation index of the target region is located in the third section [a2, ∞], wherein, if the current normalized difference vegetation index is located in the section [a2, ∞], it is determined that the vegetation coverage data of the target region is estimated by using the enhanced vegetation index.

[0104] Optionally, the vegetation coverage estimation method based on the predetermined target section comprises:

[0105] If the normalized difference vegetation index of the target region is less than the first threshold value, the difference value vegetation index is determined according to the difference between the near-infrared band reflectivity and the red band reflectivity.

[0106] The vegetation coverage data of the target region is estimated according to the difference value vegetation index, the minimum value of the difference value vegetation index, the difference value vegetation index when the normalized difference vegetation index is the first threshold value, and the first vegetation coverage.

[0107] Optionally, the second difference value determined according to the difference between the near-infrared band reflectivity and the red band reflectivity can be used to determine the difference value vegetation index, which can refer to the following formula:

[0108] DVI = R nir - R r (7)

[0109] In formula (7), the DVI is the difference value vegetation index, R nir is the near-infrared band reflectivity, and R r is the reflectivity of the red band.

[0110] Optionally, the vegetation coverage data of the target region is estimated according to the difference value vegetation index, the minimum value of the difference value vegetation index, the difference value vegetation index when the normalized difference vegetation index is the first threshold value, and the first vegetation coverage, which can refer to the following formula:

[0111]

[0112] In formula (8), C is the vegetation coverage data of the target region, DVI is the difference value vegetation index, DVI min is the minimum value of the difference value vegetation index, DVI a1 is the difference value vegetation index when the normalized difference vegetation index is the first threshold value, and C a1 is the first vegetation coverage.

[0113] Optionally, the vegetation coverage estimation method based on the predetermined target section includes:

[0114] If the normalized difference vegetation index of the target region is greater than a second threshold value, an enhanced vegetation index is determined.

[0115] According to the second vegetation coverage, the enhanced vegetation index, the maximum enhanced vegetation index, and the enhanced vegetation index when the normalized difference vegetation index is equal to the second threshold value, vegetation coverage data of the target region is estimated.

[0116] Optionally, the following formula can be referred to:

[0117]

[0118] In formula (9), EVI is the enhanced vegetation index, EVI a2 is the enhanced vegetation index when the normalized difference vegetation index is equal to the second threshold value, EVI max is the maximum enhanced vegetation index, C a2 is equal to the second vegetation coverage, the target vegetation coverage of the present application is divided into three sections by determining the NDVI, and different vegetation index estimation algorithms are selected, i.e. the coverage is divided into three intervals of 0-C a1 , C a1 -C a2 , C a2 , and the target vegetation coverage is obtained by estimation.

[0119] Optionally, the determination of the enhanced vegetation index includes:

[0120] A first product is determined according to the red band reflectance and a first parameter;

[0121] A second product is determined according to the blue band reflectance and a second parameter;

[0122] A first sum is determined according to the near-infrared band reflectance, the first product, the second product, and a preset constant;

[0123] The enhanced vegetation index is determined according to the quotient of the difference between the near-infrared band reflectance and the red band reflectance and the first sum.

[0124] Optionally, the determination of the enhanced vegetation index can refer to the following formula:

[0125]

[0126] In formula (10), EVI is the enhanced vegetation index; G is 2.5; R r is the red band reflectance; Rb is the reflectivity of the blue band; L is a preset constant, which is optionally 1; C1 is a first parameter, which is optionally 6; C2 is a second parameter, which is optionally 7.5; R nir is the reflectivity of the near-infrared band, R r is the reflectivity of the red band.

[0127] Figure 4 is a flowchart of a process for obtaining soil moisture data of a target area, the process comprising: determining 1km resolution resampling data of the target area based on time series remote sensing data; inputting the resampling data into a soil moisture estimation model to obtain 1km resolution soil moisture data output by the soil moisture estimation model, including:

[0128] Step 401, unify the resolution scale corresponding to the current land surface temperature data, the resolution scale corresponding to the current normalized vegetation index, the resolution scale corresponding to the current reflectivity, the resolution scale corresponding to the current digital elevation model data, and the resolution scale corresponding to the current precipitation data to a first resolution scale corresponding to the current microwave radiometer data, to construct a soil moisture estimation model according to the current land surface temperature data, the current normalized vegetation index, the current reflectivity data, the current digital elevation model data, the current precipitation data, and the current microwave radiometer data after the resolution scale is unified;

[0129] Step 402, resample to a second resolution scale, input resampled land surface temperature data, resampled normalized vegetation index, resampled reflectivity data, resampled digital elevation model data, and resampled precipitation data into the soil moisture estimation model to obtain soil monitoring data corresponding to the second resolution scale output by the soil moisture estimation model;

[0130] Step 403, determine soil moisture data according to the soil monitoring data;

[0131] The first resolution scale is greater than the second resolution scale;

[0132] The second resolution scale is 1km resolution.

[0133] In the present application, if the spatial resolution of the soil moisture data used in calculating the drought index is relatively low, the method of steps 401 to 403 can be used to realize spatial down-sampling of the random forest.

[0134] In step 401, the first resolution scale can be 25km, in the case that the resolution scale corresponding to the current land surface temperature data, the resolution scale corresponding to the current normalized vegetation index, the resolution scale corresponding to the current reflectivity, the resolution scale corresponding to the current digital elevation model data and the resolution scale corresponding to the current precipitation data are all different, the resolution scale corresponding to the current land surface temperature data, the resolution scale corresponding to the current normalized vegetation index, the resolution scale corresponding to the current reflectivity, the resolution scale corresponding to the current digital elevation model data and the resolution scale corresponding to the current precipitation data are unified to the first resolution scale corresponding to the current microwave radiometer data, and a random forest network model is constructed.

[0135] In step 402, resampling to the second resolution scale, inputting the resampled land surface temperature data, the resampled normalized vegetation index, the resampled reflectivity, the resampled digital elevation model data and the resampled precipitation data into the network model, and obtaining the soil moisture data corresponding to the second resolution scale output by the network model, which can be 1km. The present application uses the upscaled data and the current microwave radiometer data to establish a network model, and evaluates and analyzes the network model. The trained model is applied to the data resampled to 1km resolution to generate 1km resolution soil moisture data.

[0136] In step 403, the soil moisture index is determined according to the soil moisture data. The present application verifies the downscaling results of the network model with the original soil moisture data, and compares them with other downscaling methods.

[0137] Optionally, the soil moisture data includes the soil moisture with the maximum humidity in the monitoring area, the soil moisture with the minimum humidity in the monitoring area, and soil moisture monitoring data.

[0138] Optionally, the soil monitoring data includes the soil moisture with the maximum humidity in the monitoring area, the soil moisture with the minimum humidity in the monitoring area, and soil moisture monitoring data.

[0139] The soil moisture data is determined according to the soil monitoring data, including:

[0140] A first soil moisture difference value is determined according to the soil moisture monitoring data and the soil moisture with the minimum humidity in the monitoring area.

[0141] A second soil moisture difference value is determined according to the soil moisture with the maximum humidity in the monitoring area and the soil moisture with the minimum humidity in the monitoring area.

[0142] The soil moisture data is determined according to the first soil moisture difference value and the second soil moisture difference value.

[0143] Optionally, the soil moisture index is determined according to the soil moisture data, and the formula is as follows:

[0144]

[0145] In formula (11), NSM is a normalized soil moisture index; SM max is the soil moisture with the maximum humidity in the monitoring area; SM min is the soil moisture with the minimum humidity in the monitoring area; and SM is soil moisture monitoring data.

[0146] Figure 5 Fig. 2 is a flowchart of a method for constructing a regional drought remote sensing comprehensive index according to the present application, and the method further comprises the following steps before acquiring time series remote sensing data of a target region:

[0147] Step 501: determining a crop recognition index and a normalized vegetation index mean of any monitoring region according to time series remote sensing data of the monitoring region.

[0148] Step 502: determining that the monitoring region is a target region when the crop recognition index is greater than or equal to a first preset threshold and the normalized vegetation index mean is greater than or equal to a second preset threshold.

[0149] In step 501, time series remote sensing data of multiple monitoring regions can be acquired, but not all monitoring regions are target monitoring regions. The present application aims to monitor the drought index of farmland regions, so it is necessary to distinguish which are farmland and which are non-farmland in the remote sensing data.

[0150] The present application distinguishes the coverage information of farmland and non-farmland by acquiring time series high temporal resolution remote sensing data, and the crop recognition index of the monitoring region can be determined according to the following formula:

[0151]

[0152] In formula (12), NDVI i is the NDVI value of the i-th data covering the same region; i is the i-th data of the remote sensing image participating in the calculation; n1 is the number of scenes of the remote sensing image; is the NDVI mean value of the time series.

[0153] In step 502, when CRI≥a0 and , it is determined that the monitoring region is a target monitoring region, i.e., a farmland region.

[0154] The application combines the vegetation coverage and soil moisture product resolution of the region, obtains the crop range and non-crop spatial range through the medium and high resolution remote sensing data, estimates the purpose of accurately identifying regional drought by using temperature, soil moisture and vegetation index, combining the spatial down-sampling of random forest and the method of index reconstruction.

[0155] The application also provides a regional drought remote sensing comprehensive index monitoring method based on the regional drought remote sensing comprehensive index construction method, comprising:

[0156] Obtaining initial land surface temperature data, initial soil moisture data and initial vegetation reflectivity data of a region to be monitored;

[0157] Inputting the initial land surface temperature data, the initial soil moisture data and the initial vegetation reflectivity data into the regional drought index model to obtain the drought remote sensing comprehensive index of the region to be monitored output by the regional drought index model.

[0158] The application can integrate the drought index monitoring method of the application into an existing monitoring system or software system or develop into an independent program by setting a remote sensing drought index software module, so that in the actual application process, firstly, the initial land surface temperature data, the initial soil moisture data and the initial vegetation reflectivity data of a region to be monitored are obtained, the current normalized vegetation index of the target monitoring region is obtained by using the method of steps 501 to 503, and the down-sampled soil moisture data and the synthetic coverage estimation data are obtained by using the random forest method; the independent program or the remote sensing drought index software module is used to carry out drought monitoring, the initial land surface temperature data, the initial soil moisture data and the initial vegetation reflectivity data are input into the regional drought index model, the results of farmland drought or non-farmland drought are obtained, the drought remote sensing comprehensive index of the region to be monitored output by the regional drought index model is obtained, and the drought grade is determined to clarify the drought impact degree.

[0159] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the application, and not to limit them; although the application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the application.

Claims

1. A method for constructing a comprehensive remote sensing index for regional drought, characterized in that, include: Acquire time-series remote sensing data of the target area, and determine the land surface temperature data of the target area based on the time-series remote sensing data. Normalize the land surface temperature data to obtain normalized land surface temperature data. Based on the time-series remote sensing data, the normalized vegetation index of the target area is determined, and the target segment where the normalized vegetation index is located is determined. Based on the pre-determined vegetation cover estimation method of the target segment, the vegetation cover data of the target area is estimated. Based on the time-series remote sensing data, resampled data at a resolution of 1 km for the target area are determined, and the resampled data is input into a soil moisture estimation model to obtain soil moisture data at a resolution of 1 km for the target area output by the soil moisture estimation model; the resampled data includes surface temperature data, normalized vegetation index data, reflectance data, digital elevation data, and precipitation data. Based on the normalized surface temperature data, the vegetation cover data, and the 1km resolution soil moisture data, a regional drought index model is constructed. The process of determining 1km resolution resampled data for the target area based on the time-series remote sensing data, and inputting the resampled data into a soil moisture estimation model to obtain 1km resolution soil moisture data for the target area output by the soil moisture estimation model includes: The resolution scales corresponding to the current surface temperature data, the current normalized vegetation index, the current reflectance, the current digital elevation model data, and the current precipitation data are uniformly upgraded to the first resolution scale corresponding to the current microwave radiometer data. A soil moisture estimation model is then constructed based on the upgraded current surface temperature data, current normalized vegetation index, current reflectance data, current digital elevation model data, current precipitation data, and current microwave radiometer data. Resample to the second resolution scale, input resampled surface temperature data, resampled normalized vegetation index, resampled reflectance data, resampled digital elevation model data and resampled precipitation data into the soil moisture estimation model, and obtain the soil monitoring data corresponding to the second resolution scale output by the soil moisture estimation model; Soil moisture data are determined based on the soil monitoring data. The first resolution scale is larger than the second resolution scale; The second resolution scale is 1km resolution.

2. The method for constructing a regional drought remote sensing comprehensive index according to claim 1, characterized in that, The surface temperature data is normalized to obtain normalized surface temperature data, including: Based on the time-series remote sensing data, the maximum and minimum temperatures of the target area are determined; Based on the maximum temperature and the minimum temperature, the surface temperature data is normalized to obtain the normalized surface temperature data.

3. The method for constructing a regional drought remote sensing comprehensive index according to claim 1, characterized in that, The vegetation cover estimation method based on the predetermined target section estimates the vegetation cover data of the target area, including: If the normalized vegetation index of the target area is greater than or equal to the first threshold and less than or equal to the second threshold, then the minimum and maximum normalized vegetation index of the target area are calculated. Based on the first threshold, the minimum value of the normalized vegetation index, and the maximum value of the normalized vegetation index, calculate the first vegetation cover when the normalized vegetation index is equal to the first threshold. Based on the second threshold, the minimum value of the normalized vegetation index, and the maximum value of the normalized vegetation index, calculate the second vegetation cover when the normalized vegetation index is equal to the second threshold. The vegetation cover data of the target area is estimated based on the normalized vegetation index, the minimum value of the normalized vegetation index, and the maximum value of the normalized vegetation index.

4. The method for constructing a regional drought remote sensing comprehensive index according to claim 3, characterized in that, The vegetation cover estimation method based on the predetermined target section estimates the vegetation cover data of the target area, including: If the normalized vegetation index of the target area is less than the first threshold, the difference vegetation index is determined based on the difference between the near-infrared reflectance and the red reflectance. The vegetation cover data of the target area is estimated based on the difference vegetation index, the minimum difference vegetation index, the difference vegetation index when the normalized vegetation index is equal to the first threshold, and the first vegetation cover.

5. The method for constructing a regional drought remote sensing comprehensive index according to claim 3, characterized in that, The vegetation cover estimation method based on the predetermined target section estimates the vegetation cover data of the target area, including: If the normalized vegetation index of the target area is greater than the second threshold, then the enhanced vegetation index is determined. The vegetation cover data of the target area is estimated based on the second vegetation cover, the enhanced vegetation index, the maximum value of the enhanced vegetation index, and the enhanced vegetation index when the normalized vegetation index is set to the second threshold.

6. The method for constructing a regional drought remote sensing comprehensive index according to claim 5, characterized in that, The determination of the enhanced vegetation index includes: The first product is determined based on the red band reflectivity and the first parameter; The second product is determined based on the blue band reflectivity and the second parameter; The first sum is determined based on the near-infrared band reflectivity, the first product, the second product, and a preset constant; The enhanced vegetation index is determined by the quotient of the difference between the near-infrared band reflectance and the red band reflectance and the first sum.

7. The method for constructing a regional drought remote sensing comprehensive index according to claim 1, characterized in that, The soil monitoring data includes soil moisture with the highest humidity in the monitoring area, soil moisture with the lowest humidity in the monitoring area, and soil moisture monitoring data. The step of determining soil moisture data based on the soil monitoring data includes: The first soil moisture difference is determined based on the soil moisture monitoring data and the soil moisture with the lowest humidity in the monitoring area. The second soil moisture difference is determined based on the soil moisture with the highest humidity and the soil moisture with the lowest humidity in the monitored area. The soil moisture data is determined based on the first soil moisture difference and the second soil moisture difference.

8. The method for constructing a regional drought remote sensing comprehensive index according to claim 1, characterized in that, Before acquiring time-series remote sensing data of the target area, the method further includes: Based on the time-series remote sensing data of any monitoring area, determine the crop identification index and the mean normalized vegetation index of the monitoring area; If the crop identification index is greater than or equal to a first preset threshold and the mean normalized vegetation index is greater than or equal to a second preset threshold, the monitoring area is determined as the target area.

9. A method for monitoring a regional drought remote sensing comprehensive index based on the regional drought remote sensing comprehensive index construction method according to any one of claims 1-8, characterized in that, include: Acquire initial surface temperature data, initial soil moisture data, and initial vegetation reflectance data for the area to be monitored; Input the initial surface temperature data, the initial soil moisture data, and the initial vegetation reflectance data into the regional drought index model, and obtain the drought remote sensing comprehensive index of the area to be monitored, output by the regional drought index model.

Images