A remote sensing evaluation method for ecological suitability of waste dumps along traffic routes disturbed by engineering

Abstract

The present invention discloses a remote sensing evaluation method for the ecological suitability of a waste dump along a traffic route disturbed by an engineering project, which belongs to the field of environmental monitoring technology, and includes: selecting ecological suitability evaluation factors that affect the site selection of the waste dump; collecting various remote sensing data of the study area, extracting indicators for various factors in the study area, and then classifying and assigning values ​​to obtain a grade spatial distribution map of each factor; randomly selecting a number of sample points in the study area, extracting the graded values ​​of each factor to the sample points, and obtaining a data set for random forest model training; using the random forest method to determine the relative importance of each factor in the comprehensive evaluation index system of ecological suitability based on the Gini index; and then linearly weighting and spatially superimposing each factor to obtain an evaluation result of the ecological suitability grade of the waste dump site selection. The present invention can identify and accurately evaluate the problems existing in the ecological suitability of regional waste dumps, greatly improving the efficiency and comprehensiveness of the evaluation.

Description

Technical Field

[0001] The invention relates to the technical field of environmental monitoring, and in particular to a remote sensing evaluation method for ecological suitability of a waste dump along a traffic line disturbed by an engineering project. Background Art

[0002] Ecological suitability evaluation is an important basic work for ecological protection and planning, and is widely used in the field of engineering construction planning and site selection. At present, ecological suitability evaluation research is mostly aimed at evaluating the suitability of various functional areas for different land use types in a small area, and often constructs its index system from the perspective of ecological sensitivity. For example, the patent application with publication number CN108090694A discloses a multi-factor comprehensive sensitivity evaluation method for regional ecosystems. Although its evaluation results have practical application significance, the limitation is that the spatial heterogeneity of a small area is relatively low and the evaluation index system is not comprehensive enough. Therefore, the current ecological suitability evaluation method cannot be extended to large-scale transportation areas.

[0003] At the same time, there is little research on the ecological suitability evaluation of waste dumps along transportation lines. With the rapid development of transportation engineering, the ecological suitability evaluation of waste dumps along transportation lines is an urgent problem to be solved. Especially under the engineering disturbance during the construction of transportation engineering, the ecological environment is easily damaged, especially the stacking of engineering waste will seriously affect the regional ecological environment. Therefore, conducting ecological suitability evaluation of waste dump site selection along transportation lines is of great significance to ecological environmental risk assessment.

[0004] At present, there are few studies on the ecological suitability evaluation methods for the site selection of engineering waste dumps, and they are limited to small-scale research. At the same time, there is an incomplete comprehensive evaluation index system. It is difficult to deal with the ecological suitability site selection of waste dumps in large-scale areas along transportation lines, complex and diverse terrains, and ecologically sensitive and fragile areas. Secondly, the existing evaluations mostly use manual data collection, which is time-consuming, labor-intensive, and inefficient. When coupling multiple factors for comprehensive evaluation, the hierarchical analysis method with human intervention is often used, which reduces the objectivity of the evaluation results. Furthermore, the existing research has not comprehensively considered the ecological environment characteristics along the transportation line, resulting in poor applicability of its ecological suitability evaluation method, and it is difficult to classify the ecological suitability level of waste dumps in practice, and it is impossible to provide technical support for the low-risk ecological site selection of waste dumps along the transportation line. Summary of the invention

[0005] The present invention aims to solve the problems of the current low-risk ecological site selection index system for waste dumps, such as imperfection, low efficiency, poor applicability, human intervention when coupling multiple factors for comprehensive evaluation, and lack of site selection for waste dumps along transportation lines. A remote sensing evaluation method for ecological suitability of waste dumps along transportation lines with engineering disturbance is proposed. The method extracts multiple factors affecting the ecological suitability of waste dumps through remote sensing technology, obtains the relative weight of each factor through random forest, and then uses spatial analysis to couple each factor to conduct a comprehensive and efficient ecological suitability evaluation of the site selection of waste dumps along transportation lines.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution of the present invention is as follows:

[0007] A remote sensing evaluation method for ecological suitability of waste dumps along traffic routes disturbed by engineering projects comprises the following steps:

[0008] Step S1, conduct field research in the study area and collect various basic data, and select ecological suitability evaluation factors that affect the site selection of the waste dump, including ecological sensitivity, soil nutrients, water conservation capacity, soil retention capacity, rock exposure rate, vegetation coverage and biological richness;

[0009] Step S2, collecting various remote sensing data of the study area, extracting indicators of various factors in the study area based on the obtained remote sensing data and basic data, and then classifying and assigning values ​​to obtain a spatial distribution map of the levels of various factors;

[0010] Step S3, randomly select a number of sample points in the study area, combine the field drone image data with Google Earth high-definition satellite images, and use the site landscape pattern around the sample points for human-computer interaction to realize the ecological suitability level of the random sample points for the site selection of the waste dump; extract the values ​​of each factor after classification to the sample points to obtain the data set for random forest model training;

[0011] Step S4: train a random forest model based on the obtained data set, determine the relative importance of each factor in the comprehensive evaluation index system of ecological suitability based on the Gini index, and obtain the weight coefficient of each factor; then use the weight coefficient of each factor to linearly weight the spatial superposition of each factor to obtain the ecological suitability grade evaluation result of the waste dump site selection.

[0012] Furthermore, various types of remote sensing data including Landsat series data, DEM digital elevation model, land use data, soil properties and nutrient data and meteorological data were collected in the study area.

[0013] Furthermore, the spatial distribution map of the ecological sensitivity factors is obtained by extracting five secondary factors, namely elevation data, slope data, land use data, normalized vegetation index and water buffer zone, and assigning them values ​​respectively. Then, the analytic hierarchy process is used to determine the weight of each secondary factor. The ecological sensitivity index is obtained by spatially weighted superposition of each secondary factor, and then the ecological sensitivity index is obtained by graded division.

[0014] Furthermore, the extraction methods of the secondary factors of the ecological sensitivity evaluation index are as follows: the elevation data factor and the slope data factor are extracted in ArcGIS through the digital elevation model; the normalized vegetation index factor is extracted by using the GEE platform using the Landsat 8 image band operation, and the raster data downloaded from the GEE platform is imported into ArcGIS for reclassification to obtain the land use data factor; the land use data is used to screen the water bodies and wetland areas, and then the water buffer zone is established for the exported data to obtain the water buffer zone factor.

[0015] Furthermore, the spatial distribution map of soil nutrient factors is obtained as follows: the soil nutrient raster data of the National Earth System Science Data Center is used to extract organic matter, total nitrogen, total phosphorus, and total potassium data with a soil layer thickness of 0-30 cm. After normalization, soil nutrient indicators are obtained based on equal-weighted spatial superposition, and then graded to obtain the spatial distribution result map of soil nutrient grades.

[0016] Furthermore, the spatial distribution map of the level of water conservation factor is obtained by using the raster data of runoff rainfall, surface runoff and evapotranspiration provided by the National Earth Science Data Center platform, using the water balance equation to perform raster calculations to obtain the water conservation index, and then performing grade division to obtain the spatial distribution result map of the water conservation level.

[0017] Furthermore, the spatial distribution map of the soil conservation amount factor is obtained as follows: using the collected soil attribute data and digital elevation model data, the grid layers of rainfall erosion factor, soil erodibility factor, slope gradient and length factor, and vegetation coverage factor are extracted respectively; the modified soil loss equation is used to couple the secondary factors to obtain the soil conservation amount grid map, and then the soil conservation amount is graded based on the natural break point method and assigned values ​​from low to high according to the soil conservation amount to obtain the spatial distribution result map of the soil conservation amount grade.

[0018] Furthermore, the spatial distribution map of the rock exposure rate factor is obtained by: using the normalized bare rock index to calculate the rock exposure rate index according to the pixel binary model, classifying the rock exposure rate and assigning values ​​according to the high and low rock exposure rates, and obtaining the spatial distribution result map of the rock exposure rate level;

[0019] The evaluation thematic map of vegetation coverage factor is obtained by using the normalized vegetation index and calculating the vegetation coverage according to the pixel binary model, dividing the vegetation coverage into grades and assigning values ​​according to the size of vegetation coverage, and obtaining the spatial distribution result map of vegetation coverage grades.

[0020] Furthermore, the spatial distribution result map of the biological richness factor is obtained by assigning corresponding weights to various land use types according to land use data, and calculating the biological richness index in combination with the biological richness index formula, and then using the natural break point method to grade the obtained biological richness index, and then assigning values ​​according to the level of biological richness to obtain the spatial distribution result map of biological richness level.

[0021] Furthermore, in step S4, the importance of the evaluation factor is measured by the percentage of the average Gini reduction value of the evaluation factor to the sum of the average Gini reduction values ​​of all evaluation factors; the weight coefficient of each evaluation factor is used to linearly weight the spatial superposition of each factor to obtain a comprehensive ecological suitability evaluation index; and the natural break point classification method is used to classify the comprehensive evaluation index to obtain the ecological suitability evaluation result of the waste dump site selection.

[0022] In summary, the present invention has the following advantages:

[0023] 1. The method of the present invention is applied to the site selection of engineering waste dumps along transportation lines. It evaluates ecological sensitivity based on the analytic hierarchy process, and combines the water balance equation, the modified soil loss model, the pixel binary model, etc. to comprehensively evaluate the ecological suitability factors affecting the site selection of waste dumps in the study area. It combines the ecological suitability site selection of waste dumps with raster management, which improves the accuracy, applicability and operability of the suitability evaluation. At the same time, the spatial analysis based on random forest weighting is used to couple the influencing factors, which greatly improves the objectivity of the ecological suitability evaluation. This method can identify and accurately evaluate the problems existing in the ecological suitability of regional waste dumps, greatly improving the efficiency and comprehensiveness of the evaluation;

[0024] 2. The present invention establishes an ecological suitability evaluation index system and method for the study area, which can evaluate the ecological suitability of the waste dump site along the transportation line. By drawing the spatial distribution result map of soil nutrient status, the consideration of the ecological resilience of the waste dump site area is strengthened, and the accuracy and applicability of the evaluation are improved. At the same time, the impact of other factors on the ecological suitability of the waste dump site is comprehensively considered to make the evaluation results more comprehensive;

[0025] 3. The present invention improves the efficiency, accuracy and objectivity of the study area evaluation by drawing the ecological suitability evaluation result map of each factor, and can conduct a comprehensive analysis of the waste dump site selection area;

[0026] 4. The present invention adopts the random forest method to determine the relative importance of each evaluation factor in the comprehensive evaluation system of ecological suitability based on the Gini index, and then performs spatial analysis on the coupling of each factor, which greatly avoids the influence of human factors and improves the objectivity of the evaluation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of ecological suitability evaluation for waste dump site selection in an embodiment of the present invention;

[0028] Figure 2 A result diagram of the spatial distribution of ecological sensitivity levels evaluated by the analytic hierarchy process provided in an embodiment of the present invention;

[0029] Figure 3 A result diagram of spatial distribution of soil nutrient status levels with equal weight spatial overlay provided by an embodiment of the present invention;

[0030] Figure 4 A result diagram of spatial distribution of water conservation levels of the water balance equation model provided in an embodiment of the present invention;

[0031] Figure 5 A result diagram of the spatial distribution of soil conservation levels calculated using the modified soil loss equation provided in an embodiment of the present invention;

[0032] Figure 6 A result diagram of the spatial distribution of rock exposure rate levels provided by an embodiment of the present invention;

[0033] Figure 7 A result diagram of the spatial distribution of vegetation coverage levels provided by an embodiment of the present invention;

[0034] Figure 8 A result diagram of the spatial distribution of biological richness levels provided by an embodiment of the present invention;

[0035] Fig. 9 The ecological suitability grade evaluation result of the waste dump site selection based on random forest weighting provided in the embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to explain the present invention more clearly, the present invention is further described below in conjunction with preferred embodiments and drawings. It should be understood by those skilled in the art that the content described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0037] The present invention proposes a remote sensing evaluation method for the ecological suitability of waste dumps along traffic routes disturbed by engineering. The ecological suitability evaluation process of waste dump site selection is as follows: Figure 1 shown.

[0038] The ecological suitability of the waste dump refers to the suitability of engineering waste dumping in the area based on the evaluation results of ecological sensitivity, ecosystem stability, and ecosystem recovery capacity, under the guidance of the principle of emphasizing the priority of ecological protection. Among them, ecological sensitivity refers to the degree of response of the ecosystem to human activity disturbance and natural environmental changes, indicating the difficulty and possibility of regional ecological and environmental problems, that is, the probability of regional ecological and environmental problems in each ecosystem under the same human activity influence or external force. Ecosystem stability refers to the ability of the ecosystem to maintain or restore its own relatively stable structure and function, and its stability can be measured by the ecosystem's water conservation and soil conservation capabilities. Ecosystem restoration is often achieved through vegetation restoration, which can be measured by the abundance of soil nutrients necessary for plant growth and the richness of organisms. Comprehensive ecological suitability evaluation of the site selection of waste dumps along transportation lines.

[0039] The present invention proposes a remote sensing evaluation method for the ecological suitability of waste dumps along traffic routes disturbed by engineering projects. The method aims to conduct a comprehensive evaluation of the ecological suitability of waste dump sites by coupling seven factors: ecological sensitivity, soil nutrient status, water conservation capacity, soil retention capacity, rock exposure rate, vegetation coverage, and biological richness. The method can systematically analyze various problems existing in the ecological suitability of the study area.

[0040] Refer to Figure 1 The ecological suitability assessment process for the site selection of waste dumps is shown in the figure. The implementation steps of this method are as follows:

[0041] Step 1: Based on various limiting factors such as the rationality and economy of the transportation distance of engineering waste slag along the traffic line, the site selection range of the waste slag site is preliminarily delineated as the research area range for ecological suitability evaluation; field investigation and data collection are conducted on the basic ecological conditions of the study area, and the factors affecting the ecological suitability of the site selection of the waste slag site are screened. A comprehensive evaluation index system for ecological suitability is constructed using multi-factor indicators, as shown in Table 1.

[0042] Table 1 Comprehensive evaluation index system of ecological suitability

[0043]

[0044] Step 2: Collect remote sensing data such as Landsat series data, DEM digital elevation model, LUCC land use data, soil properties and nutrient data, and meteorological data in the study area. The specific data sources are shown in Table 2 below:

[0045] Table 2 Sources of remote sensing data for the study area

[0046]

[0047] Step 3: Use the collected remote sensing data and various basic data to extract indicators for each factor in the region to obtain evaluation thematic maps for each factor.

[0048] Figure 2 This is a result diagram of the spatial distribution of ecological sensitivity levels evaluated by the hierarchical analysis method in the embodiment of the present invention. The acquisition method is detailed as follows:

[0049] According to the main ecological sensitivity characteristics of the study area, the factors of ecological sensitivity evaluation indicators were screened, and finally five secondary factors, namely elevation, slope, land use, NDVI, and water buffer zone, were selected to evaluate its ecological sensitivity. The specific extraction methods of each factor are as follows:

[0050] The elevation and slope data were extracted in ArcGIS through the digital elevation model. The NDVI factor was extracted using the GEE platform using the Landsat 8 image band operation. The land use data used the ESA10m resolution raster data downloaded from the GEE platform and imported into ArcGIS for reclassification. The water buffer factor was extracted using land use data to screen water bodies and wetland areas, and then exported, and then the exported data was used to establish the water buffer. Finally, the above-mentioned secondary factors were graded and assigned using the following Table 3, and the sensitivity level spatial distribution map of each secondary factor was obtained.

[0051] Table 3 Ecological sensitivity evaluation index system

[0052]

[0053] The above five secondary factors are used to construct a judgment matrix using the 1-9 scaling method. The judgment matrix is ​​shown in Table 4 below:

[0054] Table 4 Ecological sensitivity assessment judgment matrix

[0055]

[0056] After constructing the above judgment matrix, the five factors are weighted and the random consistency ratio of the matrix is ​​judged. In this embodiment, the CR value is 0.015, and it is judged to be credible according to CR<0.1, indicating that the above matrix consistency test has passed and can be used as an evaluation weight.

[0057] The obtained relative weight values ​​were coupled with each sensitivity factor for spatial analysis, and then the obtained ecological sensitivity grid map was reclassified using the natural break point method, and divided into extremely low sensitivity, low sensitivity, medium sensitivity, high sensitivity and extremely high sensitivity from low to high, and values ​​1-5 were assigned from high to low according to sensitivity. Finally, the spatial distribution result map of ecological sensitivity level was obtained ( Figure 2 ).

[0058] Figure 3 This is a result map of the spatial distribution of soil nutrient status levels with equal weighted spatial overlay. The acquisition method is described in detail as follows:

[0059] Taking into account the characteristics of ecological restoration, we selected the nutrients necessary for plants to evaluate the ecological restoration capacity of the region, and finally selected four factors, namely soil organic matter, soil total nitrogen, soil total phosphorus, and soil total potassium, to measure soil nutrient indicators. The specific steps are as follows:

[0060] The density data of soil organic matter, total nitrogen, total phosphorus, and total potassium in the study area were downloaded from the National Earth System Science Data Center. The downloaded raster data were selected for the cultivated layer soil data within a thickness of 0-30 cm and calculated using the ArcGIS raster calculator to obtain the density raster maps of each soil nutrient factor.

[0061] The density grid data of each soil nutrient factor were normalized to eliminate the dimension effect, and then the normalized grid layers of the four factors were equally weighted and superimposed. Then, the natural break point classification method was used to classify them and assign values ​​of 1-5 from low to high, and the spatial distribution result map of soil nutrient status was obtained ( Figure 3 ).

[0062] Figure 4 This is a result diagram of the spatial distribution of water conservation levels of the water balance equation model provided in the embodiment of the present invention. The acquisition method is described in detail as follows:

[0063] Water conservation is the process by which ecosystems (such as forests and grasslands) interact with water through their unique structures, intercept, infiltrate and accumulate precipitation, and regulate water flow and water circulation through evapotranspiration. It is mainly manifested in alleviating surface runoff, replenishing groundwater, slowing seasonal fluctuations in river flow, retaining floodwater and replenishing drought, and ensuring water quality. Through the evaluation of this indicator, key areas that can currently and in the future assume water conservation functions can be identified. Using water conservation capacity or water conservation volume as a measurement indicator, this indicator can reflect the degree of stability of the ecosystem. The specific calculation model and steps of this indicator are as follows:

[0064] ;

[0065] ——Water conservation capacity, m3 / a;

[0066] ——runoff-generating rainfall, mm / a;

[0067] ——surface runoff, mm / a;

[0068] ——Evaporation rate, mm / a.

[0069] In this embodiment, the flow-generating rainfall, surface runoff and evapotranspiration are directly downloaded from the raster data provided by the National Earth Science Data Center platform. The raster data is imported into ArcGIS for raster calculation to obtain a spatial distribution raster map of water conservation capacity, and the natural break point classification method is used to classify it and assign values ​​1-5 from low to high according to the water conservation capacity, and finally a spatial distribution result map of water conservation capacity is obtained ( Figure 4 ).

[0070] Figure 5 The modified soil loss equation (RUSLE) model is used to extract the spatial distribution result map of soil conservation level provided in the embodiment of the present invention. The acquisition method is described in detail as follows:

[0071] Soil conservation is the role of ecosystems (such as forests, grasslands, etc.) in reducing soil erosion caused by water erosion through their structure and processes. Through the evaluation of this indicator, key areas that currently and in the future will bear soil conservation functions can be identified. Soil conservation functions are mainly related to climate, soil, topography and vegetation. Using soil conservation (the difference between potential soil erosion and actual soil erosion) as an evaluation indicator can be used to reflect the stability of the regional ecosystem. This indicator can be extracted using the modified soil and water loss equation (RUSLE) model. The specific calculation model and steps are as follows:

[0072] ;

[0073] Where: ——Soil conservation capacity, t / (km 2 *a);

[0074] , so that the Q value is converted from the American unit to the Chinese unit;

[0075] ——Rainfall erosivity factor, 100ft*sht*in / (ac*h*a);

[0076] ——Soil erodibility factor, sht*h / (100ft*sht*in);

[0077] ——slope length factor, dimension 1;

[0078] ——slope length factor, dimension 1;

[0079] ——Vegetation cover factor, dimension 1.

[0080] The specific calculation methods of each factor also include:

[0081] Rainfall erosivity factor Based on the monthly average rainfall data and annual average rainfall data provided by the hydrological stations, according to the following formula:

[0082] ;

[0083] Where: is the monthly rainfall (mm), is the annual rainfall (mm).

[0084] Calculate the rainfall erosivity , and then the Kriging interpolation method is used to obtain the spatial distribution of rainfall erosivity in the study area, that is, the rainfall erosivity factor grid map is obtained.

[0085] Soil erodibility factor Based on the soil organic carbon and the distribution content of sand, silt and clay in the soil texture in the HWSD world large-scale soil spatial distribution dataset, according to the following formula:

[0086] ;

[0087] Where: Sa is sand, Si is silt, Cl is clay, C is soil organic carbon content, Calculate the soil erodibility .

[0088] Slope length and slope factor ( LS ) is processed in ArcGIS based on the digital elevation model. The specific formula is as follows:

[0089] ;

[0090] Where: is the slope factor, is the slope;

[0091] ;

[0092] ;

[0093] ;

[0094] Where: is the slope length factor, is the slope length, is the slope length coefficient, is the slope, is the slope correction value.

[0095] Vegetation cover factor Based on the Landsat8 remote sensing image, NDVI was calculated using band operations, and then the pixel binary model was used to extract the annual maximum vegetation coverage of the study area. The vegetation coverage factor was calculated according to the following formula.

[0096] ;

[0097] Where c is the vegetation coverage.

[0098] Using the collected soil attribute data, digital elevation model and other data, the rainfall erosivity factor, soil erodibility factor, slope length factor and vegetation coverage factor were extracted respectively, and the raster layers of each factor were obtained respectively. The modified water and soil loss equation (RUSLE) was used to couple the above factors to obtain the soil conservation amount raster map, and then it was graded based on the natural break point method and assigned values ​​1-5 from low to high according to the soil conservation amount, and the spatial distribution result map of soil conservation amount was obtained ( Figure 5 ).

[0099] Figure 6 The spatial distribution result diagram of rock exposure rate provided by the embodiment of the present invention is obtained as follows:

[0100] In addition to vegetation and bare soil, the surface cover types in the area where the waste dump is located along the transportation line also have widely distributed bare bedrock. The rock exposure rate can directly reflect the degree of ecological vulnerability of the region. The higher the exposure rate, the more fragile the ecology is, and vice versa. The normalized bare rock index is used to The rock exposure rate is calculated based on the pixel binary model. The calculation formula is as follows:

[0101] ;

[0102] Where: , They are the corresponding bands of Lantsat8 OLI remote sensing images.

[0103] According to the pixel binary model, a pixel NDRI The value consists of two parts: exposed rock information and non-exposed rock information. NDRI The formula for calculating rock exposure rate is:

[0104] ;

[0105] Where: When it is composed of completely exposed rocks NDRI value; When it is composed of no exposed rock NDRIThe rock exposure rate index is calculated. The rock exposure rate is then classified according to Table 5 and the rock exposure rate is assigned values ​​from 1 to 5 from high to low to obtain the spatial distribution result of the rock exposure rate ( Figure 6 ).

[0106] Figure 7 The spatial distribution result diagram of vegetation coverage level provided by the embodiment of the present invention. The acquisition method is described in detail as follows:

[0107] The vegetation coverage index can accurately reflect the land cover status and also reflect the stability of the regional ecosystem from the side. Generally speaking, the higher the vegetation coverage, the stronger its self-regulation ability and the more stable the regional ecology. Therefore, the vegetation coverage index should also be taken into consideration when selecting the site for engineering waste dumping. The pixel binary model can be used to invert the vegetation coverage index of the study area. The specific steps also include:

[0108] Normalized Difference Vegetation Index And according to the pixel binary model, Calculate vegetation coverage, ,in , They are the corresponding bands of Lantsat8 OLI remote sensing images. The pixel binary model assumes that a pixel can be divided into two parts: soil and vegetation. Then the information observed by the remote sensing sensor can be composed of vegetation component contribution rate and soil contribution rate. Based on their values, a model can be constructed to calculate vegetation coverage. Calculate vegetation cover , which is expressed as follows:

[0109] ;

[0110] In the formula, For areas with completely bare soil or no vegetation cover value, It means that the pixels covered by vegetation are Value, that is, the value of pure vegetation pixels value; According to the overview of the study area and previous research experience, and Generally, the maximum and minimum values ​​within a certain confidence range are taken to eliminate the noise in the remote sensing image. In this invention, the confidence level is 5%, that is, the value corresponding to 5% of the pixel value is , 95% pixel value corresponds to . The vegetation coverage index is calculated.

[0111] Then, the vegetation coverage is divided into different levels according to Table 5 and the values ​​are assigned from 1 to 5 from small to large. The spatial distribution result of the vegetation coverage level is obtained ( Figure 7 ).

[0112] Table 5 Classification of rock exposure rate and vegetation coverage

[0113]

[0114] Figure 8 The method for obtaining the spatial distribution result diagram of biological richness level provided in the embodiment of the present invention is described in detail as follows:

[0115] The biological abundance index can be used to evaluate the number and variety of organisms in the study area. Therefore, the biological abundance index can be used to reflect the biodiversity of the study area, and can also invert the stability and recovery capacity of the regional ecosystem. Generally speaking, the higher the regional biological abundance index, the more stable the ecosystem and the stronger its recovery capacity after disturbance. The factors in the biological abundance index can be extracted using land use data. The calculation formula and weights of various indicators (Table 6) are as follows:

[0116] ;

[0117] Where: represents the biodiversity index, represents the normalized coefficient of biological richness.

[0118] Table 6 Biodiversity index and its weight

[0119]

[0120] The specific acquisition steps are to import the land use data of the study area into ArcGIS, divide it into 1kmx1km grids using the fishing net tool, and then calculate the biological richness value of each grid according to the above indicator weight and calculation formula in each grid, normalize it to obtain the biological richness index, and then use the natural break point method to divide it into levels, and assign values ​​1-5 in order from low to high biological richness, and obtain the spatial distribution result map of biological richness level ( Figure 8 ).

[0121] Fig. 9 The distribution result diagram of ecological suitability evaluation level of waste dump site selection based on random forest weighting provided by the embodiment of the present invention. The acquisition method is described in detail as follows:

[0122] <1> Random forest dataset construction

[0123] 150 points were randomly selected within the study area, and the field drone image data and Google Earth high-definition satellite images were combined. The surrounding landscape pattern of the point was used for human-computer interaction to achieve the ecological suitability level of the 150 sample points for the site selection of the waste dump. At the same time, the multi-value extraction to point tool in ArcGIS was used to extract the graded values ​​of ecological sensitivity, soil nutrient status, water conservation, soil retention, rock exposure rate, vegetation coverage, and biological richness to points to obtain a data set for random forest training.

[0124] <2> Model training and weight calculation

[0125] The random forest model is trained based on the sklearn library in Python. 70% of the data in the dataset is used for training and 30% for validation. After the training is completed, the Gini index method is used to , where m, n, and t are the total number of evaluation factors, the number of classification trees, and the number of nodes in a single tree, respectively; is the Gini index reduction value of the kth evaluation factor at the jth node of the hth tree; is the importance of the kth evaluation factor among all evaluation factors. The relative weights of each factor are shown in Table 7.

[0126] Table 7 Relative weights of various factors in ecological suitability evaluation

[0127]

[0128] Based on the random forest objective weight coupling of each evaluation factor, each evaluation factor is superimposed based on the random forest objective weight space. The formula is as follows:

[0129] ;

[0130] In the formula, y is the comprehensive ecological suitability evaluation index; m is the number of evaluation factors; Is the evaluation factor The corresponding random forest weight coefficient is obtained. Finally, the quantitative results of ecological suitability of the waste dump site are obtained.

[0131] The natural break point classification method was then used to classify the comprehensive evaluation index into five levels: unsuitable, basically unsuitable, basically suitable, suitable, and very suitable. The ecological suitability evaluation results of the waste dump site were obtained ( Fig. 9 ).

[0132] The method of the present invention is applied to the site selection of engineering waste dumps along transportation lines. It evaluates ecological sensitivity based on the analytic hierarchy process, combines the water balance equation, the modified soil loss model, the pixel binary model, etc. to comprehensively evaluate the ecological suitability factors affecting the site selection of waste dumps in the study area, and combines the ecological suitability site selection of waste dumps with raster management, thereby improving the accuracy, applicability and operability of the suitability evaluation. At the same time, the spatial analysis based on random forest weighting is used to couple the various influencing factors, which also greatly improves the objectivity of the ecological suitability evaluation. It is able to identify and accurately evaluate the problems existing in the ecological suitability of regional waste dumps, greatly improving the efficiency and comprehensiveness of the evaluation.

[0133] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any simple modification or equivalent change made to the above embodiment based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A remote sensing evaluation method for ecological suitability of waste dumps along traffic routes disturbed by engineering, characterized in that: The steps include: Step S1, conduct field research in the study area and collect various basic data, and select ecological suitability evaluation factors that affect the site selection of the waste dump, including ecological sensitivity, soil nutrients, water conservation capacity, soil retention capacity, rock exposure rate, vegetation coverage and biological richness; Step S2, collecting various remote sensing data of the study area, extracting indicators of various factors in the study area based on the obtained remote sensing data and basic data, and then classifying and assigning values ​​to obtain a spatial distribution map of the levels of various factors; The spatial distribution map of ecological sensitivity factors is obtained by extracting five secondary factors, namely elevation data, slope data, land use data, normalized vegetation index and water buffer zone, and assigning values ​​to them respectively. Then, the weight of each secondary factor is determined by using the analytic hierarchy process. The ecological sensitivity index is obtained by spatially weighting and superimposing each secondary factor. Then, the spatial distribution map of ecological sensitivity is obtained by performing grade division. The spatial distribution map of soil nutrient factors is obtained by using the soil nutrient grid data of the National Earth System Science Data Center to extract the organic matter, total nitrogen, total phosphorus, and total potassium data of the 0-30cm soil layer thickness, and after normalization, the soil nutrient index is obtained based on equal-weighted spatial superposition, and then the soil nutrient level spatial distribution result map is obtained by level division; The spatial distribution map of the water conservation capacity factor is obtained by using the raster data of runoff rainfall, surface runoff and evapotranspiration provided by the National Earth Science Data Center platform, using the water balance equation to perform raster calculation to obtain the water conservation capacity index, and then performing grade division to obtain the spatial distribution result map of the water conservation capacity grade; The spatial distribution map of soil conservation factor levels is obtained by using the collected soil attribute data and digital elevation model data to extract the grid layers of rainfall erosion factor, soil erodibility factor, slope length factor and vegetation coverage factor respectively; the modified soil loss equation is used to couple the secondary factors to obtain the soil conservation grid map, and then the soil conservation grid map is graded based on the natural break point method and assigned values ​​from low to high according to the soil conservation amount to obtain the spatial distribution result map of soil conservation level; The spatial distribution map of the rock exposure rate factor is obtained by using the normalized bare rock index to calculate the rock exposure rate index according to the pixel binary model, classifying the rock exposure rate and assigning values ​​according to the high and low rock exposure rates, and obtaining the spatial distribution result map of the rock exposure rate level; The evaluation thematic map of vegetation coverage factor is obtained by using the normalized vegetation index and calculating the vegetation coverage according to the pixel binary model, dividing the vegetation coverage into grades and assigning values ​​according to the size of vegetation coverage, and obtaining the result map of the spatial distribution of vegetation coverage grades; The method for obtaining the spatial distribution result map of the biological richness factor is as follows: according to the land use data, various land use types are assigned corresponding weights, and the biological richness index is calculated in combination with the biological richness index formula, and then the biological richness index is graded using the natural break point method, and then the biological richness level is assigned according to the biological richness level to obtain the spatial distribution result map of the biological richness level; Step S3, randomly select a number of sample points in the study area, combine the field drone image data with Google Earth high-definition satellite images, and use the site landscape pattern around the sample points for human-computer interaction to realize the ecological suitability level of the random sample points to be used as the site selection of the waste dump; extract the values ​​of each factor after classification to the sample points to obtain the data set for random forest model training; Step S4: train a random forest model based on the obtained data set, determine the relative importance of each factor in the comprehensive evaluation index system of ecological suitability based on the Gini index, and obtain the weight coefficient of each factor; then use the weight coefficient of each factor to linearly weight the spatial superposition of each factor to obtain the ecological suitability grade evaluation result of the waste dump site selection.

2. The method for evaluating the ecological suitability of waste dumps along traffic routes disturbed by engineering according to claim 1 is characterized in that: Various types of remote sensing data collected in the study area include Landsat series data, DEM digital elevation model, land use data, soil properties and nutrient data, and meteorological data.

3. The method for remote sensing evaluation of ecological suitability of waste dumps along traffic routes disturbed by engineering according to claim 1, characterized in that: The extraction methods of the secondary factors of the ecological sensitivity evaluation index are as follows: the elevation data factor and slope data factor are extracted in ArcGIS through the digital elevation model; the normalized vegetation index factor is extracted by using the GEE platform and the Landsat 8 image band operation, and the raster data downloaded from the GEE platform is imported into ArcGIS for reclassification to obtain the land use data factor; the land use data is used to screen water bodies and wetland areas, and then the water buffer zone is established for the exported data to obtain the water buffer zone factor.

4. The method for remote sensing evaluation of ecological suitability of waste dumps along traffic routes disturbed by engineering according to claim 1, characterized in that: In step S4, the importance of the evaluation factor is measured by the percentage of the average Gini reduction value of the evaluation factor to the sum of the average Gini reduction values ​​of all evaluation factors; the weight coefficient of each evaluation factor is used to linearly weight the spatial superposition of each factor to obtain a comprehensive ecological suitability evaluation index; The natural break point classification method is then used to classify the comprehensive evaluation index to obtain the ecological suitability evaluation results of the waste dump site selection.

Images