A remote sensing integrated monitoring method for land degradation in coal mining subsidence areas

Abstract

The present invention relates to the technical field of land degradation monitoring, and in particular to a remote sensing integrated monitoring method for land degradation in coal mining subsidence areas. The method comprises: establishing a prediction model, using the prediction model to obtain a predicted soil erosion intensity sequence, a predicted primary productivity sequence, a predicted surface water content index sequence, and a predicted aridity remote sensing monitoring index sequence, and then respectively obtaining a first trend index, a second trend index, a third trend index, and a fourth trend index at the current moment; using the soil erosion intensity, primary productivity, surface water content index, aridity remote sensing monitoring index, a first trend index, a second trend index, a third trend index, and a fourth trend index at the current moment to obtain a first index, a second index, a third index, and a fourth index; based on the first index, the second index, the third index, and the fourth index at the current moment, obtaining a land degradation degree index of a coal mining subsidence area. The present invention can accurately judge the degree of land degradation in a coal mining subsidence area.

Description

Technical Field

[0001] The invention relates to the technical field of land degradation monitoring, and in particular to a remote sensing integrated monitoring method for land degradation in coal mining subsidence areas. Background Art

[0002] Land is one of the important natural resources for human survival and development, and is also an indispensable part of the entire ecosystem. The goaf caused by large-scale coal mining will not only cause land degradation, but also increase the occurrence of a series of geological disasters (such as ground collapse, ground fissures, collapse, landslides, mudslides, etc.), threatening ecological security. The impact of land degradation in coal mining subsidence areas is mainly reflected in two aspects. On the one hand, the surface subsidence caused by coal mining leads to reduced production of farmland or even uncultivated land, the fall or death of trees, and the degradation of grasslands; on the other hand, mineral development consumes a lot of water resources, affecting the water cycle in the region, leading to the decline or even death of the ecological function of surface vegetation. Therefore, it is necessary to monitor the land degradation in coal mining subsidence areas in real time.

[0003] Traditional land degradation monitoring has the disadvantages of being time-consuming, labor-intensive, field-risky, and unable to be traced back. In addition, most land degradation measurement methods in mining areas are based on remote sensing monitoring, which is used for external manifestations of land degradation such as vegetation degradation, land use change, and land quality degradation over a long time series and over a large area. This paper uses remote sensing technology combined with field surveys to analyze the impact of underground mining on vegetation in the Lingwu mining area, pointing out that the vegetation has experienced degradation, improvement, and initial recovery. Many studies have also explored the degree of land degradation by constructing an evaluation index system, but there is no unified measurement standard for land degradation in mining areas. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a remote sensing comprehensive monitoring method for land degradation in coal mining subsidence areas. The technical solutions adopted are as follows:

[0005] An embodiment of the present invention provides a remote sensing integrated monitoring method for land degradation in coal mining subsidence areas, the method comprising:

[0006] Calculate the soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment;

[0007] Establish a prediction model, and use the prediction model to predict soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index, and obtain the predicted soil erosion intensity series, predicted primary productivity series, predicted surface water content index series and predicted drought remote sensing monitoring index series;

[0008] According to the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence, the first trend index, the second trend index, the third trend index and the fourth trend index at the current moment are respectively obtained;

[0009] The first index, the second index, the third index and the fourth index are obtained by using the soil erosion intensity, the primary productivity, the surface water content index, the drought remote sensing monitoring index, the first trend index, the second trend index, the third trend index and the fourth trend index at the current moment;

[0010] The land degradation degree index of the coal mining subsidence area is obtained based on the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment.

[0011] Preferably, the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence are obtained, including:

[0012] The prediction model is TCN network, and its loss function is mean square error loss function. Soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index each correspond to a prediction model;

[0013] The historical soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought were obtained to form a historical data set; the historical data sets of soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought were input into the corresponding prediction models respectively to obtain the predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface moisture index sequence and predicted drought remote sensing monitoring index sequence.

[0014] Preferably, obtaining the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator at the current moment includes:

[0015] The predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence are respectively fitted with straight lines to obtain the first straight line, the second straight line, the third straight line and the fourth straight line; the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line are respectively obtained; according to the slope values ​​of the first, second third and fourth straight lines, they are respectively encoded, the slope value is negative, the encoding is -1, the slope value is positive, the encoding is 1, the slope value is 0, and the encoding values ​​corresponding to the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line are the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator, respectively.

[0016] Preferably, obtaining the first indicator, the second indicator, the third indicator and the fourth indicator includes:

[0017] The soil erosion intensity at the current moment and the first trend index are weighted and summed to obtain the first index; the primary productivity at the current moment and the second trend index are weighted and summed to obtain the second index; the surface water content index at the current moment and the third trend index are weighted and summed to obtain the third index; the drought remote sensing monitoring index at the current moment and the fourth trend index are weighted and summed to obtain the fourth index;

[0018] The weights of soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index are all indicator weights with the same value; the weights of the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator are all trend weights with the same value.

[0019] Preferably, before obtaining the land degradation index of the coal mining subsidence area based on the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment, the method further includes:

[0020] Obtaining a first comprehensive weight, a second comprehensive weight, a third comprehensive weight, and a fourth comprehensive weight corresponding to the first indicator, the second indicator, the third indicator, and the fourth indicator, respectively;

[0021] The first indicator, the second indicator, the third indicator and the fourth indicator at the current moment of the first indicator are respectively normalized to obtain the normalized first indicator, the second indicator, the third indicator and the fourth indicator.

[0022] Preferably, obtaining the first comprehensive weight, the second comprehensive weight, the third comprehensive weight and the fourth comprehensive weight corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator respectively includes:

[0023] By using the hierarchical analysis method, the weights corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment are obtained, which are the first weight, the second weight, the third weight and the fourth weight respectively;

[0024] Using the entropy weight method, obtain the fifth weight, the sixth weight, the seventh weight and the eighth weight corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator respectively;

[0025] The average of the first weight and the fifth weight is the first comprehensive weight of the first indicator; the average of the second weight and the sixth weight is the second comprehensive weight of the second indicator; the average of the third weight and the seventh weight is the third comprehensive weight of the third indicator; the average of the fourth weight and the eighth weight is the fourth comprehensive weight of the fourth indicator.

[0026] Preferably, obtaining the normalized first index, second index, third index and fourth index includes:

[0027] According to the relationship between the first indicator, the second indicator, the third indicator, the fourth indicator and the severity of land degradation, the indicators are divided into positive indicators and negative indicators; positive indicators include the first indicator and the fourth indicator, and negative indicators include the second indicator and the third indicator;

[0028] The positive normalization formula and the negative normalization formula are used to perform positive normalization operations and negative normalization operations on the positive index and the reverse index respectively, so as to obtain the normalized first index, second index, third index and fourth index.

[0029] Preferably, obtaining the land degradation index of the coal mining subsidence area includes:

[0030] Based on the first comprehensive weight, second comprehensive weight, third comprehensive weight and fourth comprehensive weight corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator respectively, the normalized first indicator, the second indicator, the third indicator and the fourth indicator are weighted and summed to obtain the land degradation degree index of the coal mining subsidence area.

[0031] The embodiments of the present invention have at least the following beneficial effects: the present invention calculates the soil erosion intensity, primary productivity, surface moisture index and aridity remote sensing monitoring index of the coal mining subsidence area, analyzes the land degradation of the coal mining subsidence area from multiple directions, can make a more comprehensive judgment on the land degradation of the coal mining subsidence area, and improve the accuracy of monitoring; it also obtains the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface moisture index sequence and the predicted aridity remote sensing monitoring index sequence, and further analyzes and obtains the first trend index, the second trend index, the third trend index and the fourth trend index. The subsequent development and changes of each indicator are analyzed, which can monitor the subsequent land changes in coal mining subsidence areas and improve the comprehensiveness of monitoring; the soil erosion intensity, primary productivity, surface moisture index, drought remote sensing monitoring index are integrated with the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator, and the subsequent development and changes of each indicator are embedded in the indicator, which can take into account the current values ​​of each indicator and the subsequent changes; finally, the first indicator, the second indicator, the third indicator and the fourth indicator are integrated to comprehensively monitor the degree of land degradation in coal mining subsidence areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1A method flow chart of a method for comprehensive remote sensing monitoring of land degradation in coal mining subsidence areas provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, characteristics and effects of a remote sensing integrated monitoring method for land degradation in coal mining subsidence areas proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0035] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0036] The specific scheme of the remote sensing comprehensive monitoring method for land degradation in coal mining subsidence areas provided by the present invention is described in detail below with reference to the accompanying drawings.

[0037] Example:

[0038] The main application scenarios of the present invention are: coal mining strongly disturbs the ecological environment and is prone to cause serious land degradation. Land degradation in mining areas can be understood as: affected by coal mining disturbances and other human and natural factors and other profit-making factors, the overall structure and function of the land in the mining area are damaged and undergo quantitative and qualitative changes, and the core is the degradation of vegetation and soil. Degradation types include land degradation caused directly and indirectly by human activities, and land degradation caused by wind erosion and water erosion under natural erosion. The causes of degradation are mainly human activities and natural conditions. Since mining methods and regional differences will affect the degree of degradation, and because there are many influencing factors, there are certain difficulties in selecting influencing factors and quantifying the degree of influence of each factor. At present, remote sensing monitoring of land degradation in mining areas mostly explores the degree of land degradation in the research area by constructing an evaluation index system. Due to its complexity, there is no unified measurement standard. How to scientifically measure land degradation in mining areas is the current focus and difficulty.

[0039] See also Figure 1 , which shows a flow chart of a remote sensing integrated monitoring method for land degradation in coal mining subsidence areas provided by an embodiment of the present invention, the method comprising the following steps:

[0040] Step S1, respectively calculating the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment.

[0041] Determine the mining area that needs to be monitored and record it as the coal mining subsidence area; further obtain remote sensing image data, soil data, meteorological data, DEM data and other data.

[0042] The remote sensing image data selected are Landsat series satellite remote sensing image data. The implementer can select the data of the required period for analysis according to actual needs. The data comes from the United States Geological Survey (USGS) with an image resolution of 30m. To avoid the interference of clouds and snow, images with cloud cover less than 20% are selected. As the image basis for calculating the Normalized Difference Vegetation Index (NDVI) and land use classification, land use classification is one of the important factors of human activities in the analysis of the causes of land degradation. The Normalized Difference Vegetation Index is used to calculate soil erosion and primary productivity.

[0043] The soil data comes from the China 1:1 million database of the Nanjing Institute of Soil Science, Chinese Academy of Sciences. The database consists of a 1:1 million digital soil map of China and soil attribute data, including the organic matter content (%) of the main nutrient elements (N, P, K) in each soil type, which is used to calculate soil erosion.

[0044] Meteorological data comes from the National Meteorological Data Sharing Service Platform, which obtains the monthly average temperature (℃), monthly total precipitation (mm) and monthly total solar radiation (MJ·m-2) data of China's ground climate data monthly data set and China's radiation monthly data set, with a total of 197 meteorological stations. The data of each station are rasterized through Kriging interpolation to obtain temperature, rainfall and solar radiation for the calculation of soil erosion and soil and water conservation.

[0045] The DEM (Digital Elevation Model) data comes from the ASTER GDEM data of the Geospatial Data Cloud, which is jointly developed by the Ministry of Economy, Trade and Industry (METI) of Japan and the National Aeronautics and Space Administration (NASA) of the United States. The spatial resolution is 30m and is used to extract slope length and slope factors.

[0046] Other data include vector boundary data of coal mining subsidence areas and population density data. The vector boundary data of coal mining subsidence areas comes from the data set of the results of the third provincial land survey on the current land use survey and analysis of coal mining subsidence areas; the population density data comes from the World Pop Global Project Population Data, which is raster data with a spatial resolution of 1 km.

[0047] Furthermore, the acquired remote sensing data are unified in coordinates and resolution and cropped according to the scope of the coal mining subsidence area.

[0048] Based on the degradation monitoring data, the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment are calculated respectively.

[0049] For coal mining subsidence areas, soil erosion intensity is a very important indicator for degradation monitoring. In this invention, rainfall erosivity factor (R), soil erodibility factor (K), slope length and slope factor (LS), vegetation coverage factor (C) and soil and water conservation measures factor (P) are used as soil erosion influencing factors to calculate the soil erosion intensity in coal mining subsidence areas, specifically:

[0050] A=R·K·L·S·C·P,

[0051] Among them, A is the soil erosion intensity, that is, the amount of soil loss per unit area; R is the rainfall erosion factor; K is the soil erodibility factor; LS is the slope length and slope factor; C is the vegetation coverage factor, and P is the soil and water conservation measures factor.

[0052] Among them, for the rainfall erosivity factor, rainfall is the main driving force of soil erosion. Rainfall erosivity characterizes the potential ability of soil erosion caused by rainfall factors, and is calculated using the monthly and annual precipitation in the coal mining subsidence area.

[0053] Soil erodibility factor refers to the potential possibility of soil erosion. The larger the K value, the greater the potential risk of soil erosion under the same conditions, and vice versa. It should be noted that the units when calculating the K value are American units and need to be converted to the metric K value.

[0054] The slope length and slope factor LS, also known as the terrain factor, is set as a meaningful mathematical parameter or index, which can quantitatively describe the characteristic surface of the landform. It is calculated through the DEM data of the coal mining subsidence area and can reflect the effect of regional topographic and geomorphic characteristics on soil erosion.

[0055] The comprehensive effect of the vegetation coverage factor C in the coal mining subsidence area is the influencing factor of different ground vegetation coverage conditions on soil erosion. The present invention is calculated from the vegetation coverage of the coal mining subsidence area, where the vegetation coverage is calculated based on the vegetation coverage model of the homogenized NDVI square. The vegetation coverage value is distributed between 0 and 1, 0 means that the surface is not covered with vegetation, 1 means that the surface is completely covered with vegetation, and the higher the value, the better the vegetation coverage.

[0056] The soil and water conservation measures factor P reflects the ratio between the amount of soil loss after taking water conservation measures (such as contour farming, terrace construction, etc.) and the amount of soil loss when planting along the slope. The value range of P is 0 to 1, where 0 represents an area with good prevention and control measures and basically no erosion, and 1 represents an area where no control measures are taken. Therefore, it is necessary to classify the land in the coal mining subsidence area and obtain the land use type classification results.

[0057] Specifically, Landsat images are obtained and screened in remote sensing data, and the image quality band (QA) value is obtained based on the CFMask algorithm. Clouds and cloud shadows are removed to eliminate the influence of clouds in Landsat images. Then, the image data set is synthesized and cropped according to the area of ​​coal mining subsidence to obtain image data of each year. After completing the preprocessing of Landsat image data, the data processing capabilities of the GEE cloud platform are used to realize the land use classification work in the Fenhe River Basin. To classify land use, it is necessary to construct a data set with classification characteristics. First, the spectral bands are added. In order to improve the accuracy of classification, spectral index features and terrain features are further added to construct a classification feature data set. The sample point data is constructed on the Landsat8 image, and the sample point data of different land use types in the study area are selected in combination with the high-resolution images provided by Google Earth Pro. The sample point data sets of other years are based on the sample point data, and are adjusted according to the characteristics of land object changes and prior knowledge in the study area to construct sample point data of different years. Then, the sample data of each year are divided into training samples and verification samples in a ratio of 7:3, and the land use classification of the study area is completed using the random forest classification algorithm. Finally, the land use classification data set of each year in the study area was obtained. The classification accuracy was evaluated by the overall classification accuracy and Kappa coefficient. Specifically, the land in the coal mining subsidence area was divided into six types: cultivated land, forest land, grassland, water area, construction land and unused land.

[0058] Then, P is assigned a value based on the land type classification results, where cultivated land is 0.3, forest land is 1, grassland is 1, water area is 0, building land is 0, and unused land is 0.

[0059] Furthermore, the primary productivity of coal mining subsidence areas is calculated. One of the consequences of the decline in soil and water conservation capacity is the decline and loss of land biological or economic productivity, which is directly manifested in the decline of vegetation biomass. Primary productivity (NPP) is the amount of organic matter accumulated by green plants per unit area and per unit time, which can characterize the quality of terrestrial ecosystems.

[0060] Based on the classification of land use over the years, the remote sensing estimation model of biomass of different vegetation types was applied to obtain the coal mining subsidence area, that is, the spatiotemporal distribution map of biomass of different vegetation types in the coal mining subsidence area. The improved CASA model is based on the view of resource balance and believes that the size of primary productivity is affected by the most scarce factor of vegetation photosynthesis. It is a driving model determined by the light energy utilization rate ε photosynthetically active radiation. The calculation formula is as follows:

[0061] NPP(x,t)=APAR(x,t)*ε(x,t)

[0062] Among them, NPP(x, t) represents the primary productivity in the coal mining subsidence area; APAR(x, t) is photosynthetically active radiation, which represents the effective photosynthetic radiation absorbed by vegetation per unit pixel x in time period t, and ε(x, t) represents the actual light energy utilization rate per unit pixel x after time period t. The primary productivity of the coal mining subsidence area at the current moment is recorded as NPP.

[0063] APAR (x, t) is the solar radiation with a wavelength range of 0.38 to 0.76 mm absorbed and utilized by the vegetation canopy, which depends on the total solar radiation in the coal mining subsidence area and the physiological characteristics of different vegetation. The light energy utilization rate of vegetation ε (x, t) refers to the ratio of the chemical energy contained in the organic matter produced per unit area per unit time to the photosynthetically active radiation energy in the same area. This parameter is affected by the external environment and the characteristics of different vegetation, and is closely related to temperature and precipitation.

[0064] Next, the surface water content index is calculated. The coal mining subsidence area is the coal mining subsidence area. The soil layer structure in the subsidence area is destroyed, which interferes with the normal transportation of water. Under the long-term effect of collapse, the soil water holding capacity will be reduced, and the soil water content will be reduced. The present invention takes into account that the surface water content index (SWCI) reflects the comprehensive surface water content information in a unified and intuitive way, and selects SWCI to reflect the water content status.

[0065] Specifically, the calculation formula of the surface moisture index is:

[0066]

[0067] Among them, SWCI represents the surface water content index calculated from LandSat8 images; b6 represents the reflectance value of the 6th band, with a wavelength of 1.57-1.65 μm; b7 represents the reflectance value of the 7th band, with a wavelength of 2.11-2.29 μm.

[0068] The SWCI value is between -1 and 1. The larger the value, the wetter the surface, the greater the soil moisture content, the better the irrigation and drainage conditions, and the higher the quality of the arable land. Conversely, the smaller the soil moisture content, the lower the arable land quality.

[0069] Finally, the Normalized Difference Vegetation Index (NDVI) and surface albedo (Albedo) of the coal mining subsidence area were inverted using MODIS and Landsat data to construct the drought remote sensing monitoring index (DDI).

[0070] In order to eliminate the dimension difference between NDVI and Albedo in the two-dimensional feature space, the two were normalized respectively, and the values ​​of 0.5 and 99.5 of the surface albedo and vegetation index of the five phases of data in the study area were counted as the maximum and minimum values ​​for normalization, specifically:

[0071]

[0072] Construct NDVI-Albedo feature space using normalized NDVI and Albedo

[0073] At the upper boundary of the NDVI-Albedo feature space, Albedo and NDVI show a significant linear negative correlation:

[0074] Albedo = aNDVI + b, indicating that as the degree of drought increases, NDVI gradually decreases, while the surface albedo gradually increases. Therefore, the aridification process and changes in its surface characteristics can be clearly and intuitively reflected in the NDVI-Albedo feature space.

[0075] By dividing the Albedo-NDVI feature space in the vertical direction of the drought change trend, different drought-affected lands can be effectively distinguished and expressed by a simple binary linear polynomial in the Albedo-NDVI feature space, that is, the drought remote sensing monitoring index (DDI):

[0076] DDI = mN-A,

[0077] Among them, m is the first coefficient, which is the negative reciprocal of a, that is, m = -1 / a; N is the normalized vegetation index; A is the normalized surface albedo.

[0078] So far, based on the collected degradation monitoring data of the coal mining subsidence area, the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment have been obtained.

[0079] It should be noted that when obtaining the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment, they are obtained at preset intervals. Preferably, in the embodiment of the present invention, three months is the preset time interval, and the implementer can select a suitable preset time interval according to actual conditions; when calculating the soil erosion intensity, the land type classification results are needed to assign a value to P. Since the degree of change of P does not change significantly with time, it is necessary to obtain the P value of the coal mining subsidence area at annual time intervals. For example, to calculate the soil erosion intensity of the coal mining subsidence area in 2020, the P value used is the P value at the end of 2019.

[0080] Step S2, establish a prediction model, use the prediction model to predict soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index, and obtain a predicted soil erosion intensity sequence, a predicted primary productivity sequence, a predicted surface moisture index sequence and a predicted drought remote sensing monitoring index sequence.

[0081] After obtaining the current soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought, further analysis is needed to obtain the degree of soil degradation. However, it is one-sided to obtain the degree of soil degradation only by analyzing the current soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought. It does not take into account the subsequent changes in the coal mining subsidence area, which is not conducive to the subsequent land degradation analysis. The subsequent gradual deterioration or alleviation of land degradation cannot be manifested at the current moment. Therefore, it is necessary to combine historical data to predict soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought, and conduct subsequent analysis based on the predicted data and combined with current data.

[0082] A prediction model is constructed, wherein the prediction model in the embodiment of the present invention is a TCN network, and its loss function is a mean square error loss function, wherein each indicator corresponds to a prediction model; historical soil erosion intensity, primary productivity, surface moisture index and aridification remote sensing monitoring index are obtained to form a historical data set; the historical data sets of soil erosion intensity, primary productivity, surface moisture index and aridification remote sensing monitoring index are respectively input into corresponding prediction models to obtain a predicted soil erosion intensity sequence, a predicted primary productivity sequence, a predicted surface moisture index sequence and a predicted aridification remote sensing monitoring index sequence.

[0083] At this point, the predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface moisture index sequence and predicted drought remote sensing monitoring index sequence corresponding to the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index at the current moment can be obtained.

[0084] It should be noted that the time series length of each prediction sequence obtained needs to be determined according to the actual situation, and the implementer can determine the time series length of each prediction sequence according to the specific actual situation.

[0085] Step S3, respectively obtaining the first trend index, the second trend index, the third trend index and the fourth trend index at the current moment according to the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface moisture index sequence and the predicted drought remote sensing monitoring index sequence.

[0086] After obtaining the predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface moisture index sequence and predicted drought remote sensing monitoring index sequence, it is necessary to analyze the predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface moisture index sequence and predicted drought remote sensing monitoring index sequence to obtain the subsequent changes in soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index.

[0087] Specifically, linear fitting is performed on the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence to obtain the first straight line, the second straight line, the third straight line and the fourth straight line; the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line are obtained respectively; encoding is performed according to the slope values ​​of the first, second, third and fourth straight lines, the slope value is negative, the encoding is -1, the slope value is positive, the encoding is 1, and the slope value is 0, the encoding is zero; the encoding values ​​corresponding to the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line are the first trend index, the second trend index, the third trend index and the fourth trend index respectively. The purpose of encoding the slope value is to reduce the error. In the process of prediction and linear fitting, both will produce certain errors. If encoding is not performed, the trend change index will have a great impact on the calculated index itself, resulting in an increase in the error of the subsequent results.

[0088] So far, the first trend index, the second trend index, the third trend index and the fourth trend index corresponding to the soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index at the current moment are obtained respectively.

[0089] Step S4, using the soil erosion intensity, primary productivity, surface water content index, drought remote sensing monitoring index, first trend indicator, second trend indicator, third trend indicator and fourth trend indicator at the current moment to obtain the first indicator, the second indicator, the third indicator and the fourth indicator.

[0090] After obtaining the first trend indicator, second trend indicator, third trend indicator and fourth trend indicator corresponding to the soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index at the current moment, they need to be integrated to obtain the corresponding first indicator, second indicator, third indicator and fourth indicator.

[0091] Specifically, the soil erosion intensity at the current moment and the first trend index are weighted and summed to obtain the first index; the primary productivity at the current moment and the second trend index are weighted and summed to obtain the second index; the surface water content index at the current moment and the third trend index are weighted and summed to obtain the third index; the drought remote sensing monitoring index at the current moment and the fourth trend index are weighted and summed to obtain the fourth index; and the weights of the soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index are all indicator weights, and the weights of the first trend index, the second trend index, the third trend index and the fourth trend index are all trend weights. Preferably, in the embodiment of the present invention, the value of the indicator weight is 0.8, and the index of the trend weight is 0.2. The implementer can set the values ​​of the indicator weight and the trend weight according to the actual situation.

[0092] At this point, the first index, the second index, the third index, and the fourth index at the current moment are obtained.

[0093] Step S5, obtaining a land degradation index of the coal mining subsidence area based on the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment.

[0094] In step S4, the first indicator, the second indicator, the third indicator and the fourth indicator of the coal mining subsidence area at the current moment are obtained. However, it is difficult to judge the degree of land degradation using these indicators alone. Therefore, it is necessary to integrate them to obtain comprehensive indicators to characterize the land degradation situation in the coal mining subsidence area.

[0095] The present invention uses the analytic hierarchy process to obtain the weights corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment, which are the first weight W1, the second weight W2, the third weight W3 and the fourth weight W4 respectively; because the analytic hierarchy process is greatly affected by the subjectivity of experts, the present invention also uses the entropy weight method to obtain the weights corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator, namely the fifth weight W1′, the sixth weight W2′, the seventh weight W3′ and the eighth weight W4′; the average of the first weight and the fifth weight is the first comprehensive weight W1″ of the first indicator; the average of the second weight and the sixth weight is the second comprehensive weight W2″ of the second indicator; the average of the third weight and the seventh weight is the third comprehensive weight W3″ of the third indicator; the average of the fourth weight and the eighth weight is the fourth comprehensive weight W4″ of the fourth indicator. When the entropy weight method is used to obtain the weights, it is necessary to calculate the first indicator, the second indicator, the third indicator and the fourth indicator in the previous history, and then use the entropy weight method to obtain the weight corresponding to each indicator.

[0096] Since the first, second, third and fourth indicators at the current moment have dimensional inconsistency, they need to be preprocessed, that is, normalized. From the analysis of the previous indicators, it can be seen that the indicators are divided into two categories according to the relationship between the first, second, third and fourth indicators and the severity of land degradation: one is that as the indicator factor value increases, the greater the degree of land degradation is likely to occur, which is called a positive indicator, including the first and fourth indicators, and a positive normalization operation is performed on it; the other is that as the indicator factor value increases, the less likely the degree of land degradation is likely to occur, which is called a reverse indicator, including the second and third indicators, and a reverse normalization operation is performed on it. The specific normalization formula includes a positive normalization formula and a negative normalization formula. The positive normalization formula is:

[0097]

[0098] The negative normalization formula is:

[0099]

[0100] Among them, X scale and X scale - are the results of single factor forward and reverse normalization, X is the original data; X max is the maximum value in the data; X minIt is the minimum value in the data. It should be noted that when performing planning operations, it is necessary to calculate the first, second, third and fourth indicators of the history, and then normalize the first, second, third and fourth indicators at the current moment. In this way, the first, second, third and fourth indicators obtained will increase as the degree of possible land degradation increases. The normalized first, second, third and fourth indicators are recorded as C1, C2, C3 and C4 respectively.

[0101] Finally, the comprehensive index model was used to perform weighted sum calculations on each indicator, with LDI representing the land degradation index, specifically:

[0102] LDI=W1″×C1+W2″×C2+W3″×C3+W4″×C4,

[0103] The larger the LDI value, the more serious the land degradation, and vice versa. It should be noted that the first comprehensive weight, the second comprehensive weight, the third comprehensive weight and the fourth comprehensive weight need to be normalized before weighted combination.

[0104] It should be noted that the embodiment of the present invention obtains the land degradation index corresponding to the coal mining subsidence area in units of preset intervals, and the implementer can select a suitable time unit for analysis according to actual conditions.

[0105] The degree of land degradation is divided into five levels: degradation (0≤LDI<0.35), mild degradation (0.35≤LDI<0.45), moderate degradation (0.45≤LDI<0.55), high degradation (0.55≤LDI<0.65) and severe degradation (LDI≥0.65).

[0106] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0107] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A remote sensing integrated monitoring method for land degradation in coal mining subsidence areas, characterized in that: The method includes: Calculate the soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index of the coal mining subsidence area at the current moment; Establish a prediction model, and use the prediction model to predict soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index, and obtain the predicted soil erosion intensity series, predicted primary productivity series, predicted surface water content index series and predicted drought remote sensing monitoring index series; According to the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence, the first trend index, the second trend index, the third trend index and the fourth trend index at the current moment are respectively obtained; The first index, the second index, the third index and the fourth index are obtained by using the soil erosion intensity, the primary productivity, the surface water content index, the drought remote sensing monitoring index, the first trend index, the second trend index, the third trend index and the fourth trend index at the current moment; Obtaining a land degradation index of the coal mining subsidence area based on the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment; The obtaining of the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator at the current moment includes: Perform straight line fitting on the predicted soil erosion intensity sequence, the predicted primary productivity sequence, the predicted surface water content index sequence and the predicted drought remote sensing monitoring index sequence, respectively, to obtain the first straight line, the second straight line, the third straight line and the fourth straight line; respectively obtain the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line; respectively encode according to the slope values ​​of the first, second third and fourth straight lines, the slope value is negatively encoded as 0, the slope value is positively encoded as 1, and the encoding values ​​corresponding to the slopes of the first straight line, the second straight line, the third straight line and the fourth straight line are the first trend index, the second trend index, the third trend index and the fourth trend index, respectively; The obtaining of the first index, the second index, the third index and the fourth index comprises: The soil erosion intensity at the current moment and the first trend index are weighted and summed to obtain the first index; the primary productivity at the current moment and the second trend index are weighted and summed to obtain the second index; the surface water content index at the current moment and the third trend index are weighted and summed to obtain the third index; the drought remote sensing monitoring index at the current moment and the fourth trend index are weighted and summed to obtain the fourth index; The weights of soil erosion intensity, primary productivity, surface moisture index and drought remote sensing monitoring index are all indicator weights with the same value; the weights of the first trend indicator, the second trend indicator, the third trend indicator and the fourth trend indicator are all trend weights with the same value.

2. A remote sensing comprehensive monitoring method for land degradation in coal mining subsidence areas according to claim 1, characterized in that: The obtained predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface water content index sequence and predicted drought remote sensing monitoring index sequence include: The prediction model is TCN network, and its loss function is mean square error loss function. Soil erosion intensity, primary productivity, surface water content index and drought remote sensing monitoring index each correspond to a prediction model; The historical soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought were obtained to form a historical data set; the historical data sets of soil erosion intensity, primary productivity, surface moisture index and remote sensing monitoring index of drought were input into the corresponding prediction models respectively to obtain the predicted soil erosion intensity sequence, predicted primary productivity sequence, predicted surface moisture index sequence and predicted drought remote sensing monitoring index sequence.

3. The remote sensing comprehensive monitoring method for land degradation in coal mining subsidence areas according to claim 1 is characterized in that: Before obtaining the land degradation index of the coal mining subsidence area based on the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment, the method further includes: Obtaining a first comprehensive weight, a second comprehensive weight, a third comprehensive weight, and a fourth comprehensive weight corresponding to the first indicator, the second indicator, the third indicator, and the fourth indicator, respectively; The first indicator, the second indicator, the third indicator and the fourth indicator at the current moment of the first indicator are respectively normalized to obtain the normalized first indicator, the second indicator, the third indicator and the fourth indicator.

4. A remote sensing comprehensive monitoring method for land degradation in coal mining subsidence areas according to claim 3, characterized in that: The obtaining of the first comprehensive weight, the second comprehensive weight, the third comprehensive weight and the fourth comprehensive weight respectively corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator comprises: By using the hierarchical analysis method, the weights corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator at the current moment are obtained, which are the first weight, the second weight, the third weight and the fourth weight respectively; Using the entropy weight method, obtain the fifth weight, the sixth weight, the seventh weight and the eighth weight corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator respectively; The average of the first weight and the fifth weight is the first comprehensive weight of the first indicator; the average of the second weight and the sixth weight is the second comprehensive weight of the second indicator; the average of the third weight and the seventh weight is the third comprehensive weight of the third indicator; the average of the fourth weight and the eighth weight is the fourth comprehensive weight of the fourth indicator.

5. The remote sensing integrated monitoring method for land degradation in coal mining subsidence areas according to claim 3 is characterized in that: The normalized first index, second index, third index and fourth index include: According to the relationship between the first indicator, the second indicator, the third indicator, the fourth indicator and the severity of land degradation, the indicators are divided into positive indicators and negative indicators; positive indicators include the first indicator and the fourth indicator, and negative indicators include the second indicator and the third indicator; The positive normalization formula and the negative normalization formula are used to perform positive normalization operations and negative normalization operations on the positive index and the reverse index respectively, so as to obtain the normalized first index, second index, third index and fourth index.

6. The remote sensing integrated monitoring method for land degradation in coal mining subsidence areas according to claim 1 is characterized in that: The method of obtaining the land degradation index of the coal mining subsidence area includes: Based on the first comprehensive weight, second comprehensive weight, third comprehensive weight and fourth comprehensive weight corresponding to the first indicator, the second indicator, the third indicator and the fourth indicator respectively, the normalized first indicator, the second indicator, the third indicator and the fourth indicator are weighted and summed to obtain the land degradation degree index of the coal mining subsidence area.

Images