Ecological environment quality assessment method, system and equipment
By using Sentinel-2A remote sensing data and random forest classification methods, combined with the land use transfer matrix and ecological environment quality index model, the difficult problem of dynamic monitoring of ecological environment quality in cold temperate drinking water source basins was solved, and accurate assessment and dynamic monitoring of ecological environment quality were achieved.
Patent Information
- Application Number
- CN202510780728.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-12
- Publication Date
- 2025-09-26
AI Technical Summary
Current technology makes it difficult to accurately obtain dynamic monitoring results of ecological and environmental quality in cold temperate drinking water source basins, especially on long-term time scales, and there is a lack of effective assessment methods.
Sentinel-2A remote sensing data combined with random forest supervised classification method were used to obtain land use classification results and construct a land use transfer matrix. An ecological environment quality index (RESI) model was constructed using multiple categories of remote sensing ecological indicators, and the model was adjusted in combination with the landscape pattern index to achieve ecological environment quality assessment.
It has achieved accurate dynamic monitoring of the ecological environment quality of the cold temperate drinking water source basin, revealed the impact of land use changes and landscape pattern evolution on ecological environment quality, and provided scientific evaluation methods and system support.
Smart Images
Figure CN120706690A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of ecological protection technology, and in particular to an ecological environment quality assessment method, system and equipment. Background Art
[0002] The ecological and environmental quality (RESI) of drinking water sources is a key indicator for ensuring regional water security and human health. In the process of global urbanization, land use change (LUCC) and landscape fragmentation (LPC) have become core drivers of reservoir ecosystem degradation by altering the hydrological cycle, biological habitats, and carbon and nitrogen fluxes. Due to their unique seasonal freeze-thaw characteristics and ecological fragility, cold temperate basins respond significantly differently to human disturbances than temperate and tropical regions. However, approximately 83% of cold temperate reservoirs worldwide currently lack long-term ecological monitoring data, seriously restricting the achievement of regional sustainable development goals.
[0003] With its 10-meter spatial resolution and red-edge band advantages, the Sentinel-2A satellite can accurately identify small-scale land cover changes in the cold temperate zone (such as wetland fragmentation and expansion of construction land). The random forest (RF) algorithm is widely used in land use classification due to its ability to process high-dimensional data. The application of landscape indices (patch density PD, aggregation index AI) can quantitatively characterize landscape fragmentation and ecological connectivity. In addition, the RESI model that integrates NDVI (normalized difference vegetation index), NDWI (normalized water index) and LST (land surface temperature) can comprehensively evaluate vegetation productivity, water stress and thermal environment effects. However, current research is mostly limited to a single time section or spatial scale, and the long-term dynamic evolution process of RESI in cold temperate basins still needs to be revealed. Summary of the Invention
[0004] The present invention provides an ecological environment quality assessment method, system, and device to solve the above-mentioned problem existing in the prior art, namely, how to accurately obtain dynamic monitoring results of watersheds in cold temperate drinking water sources. The present invention provides an ecological environment quality assessment method, which includes:
[0005] Obtain Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured;
[0006] The random forest supervised classification method was used to classify the Sentinel-2A remote sensing data and determine the land use classification results;
[0007] Based on the land use classification results, a land use transfer matrix is constructed, and land use change data is obtained through the land use transfer matrix;
[0008] Based on multiple types of remote sensing ecological indicators, an ecological environment quality index RESI model is constructed to obtain the initial ecological environment quality index RESI. The initial ecological environment quality index RESI is adjusted through land use change data and landscape pattern index to obtain the ecological environment quality assessment results of the tested area.
[0009] Optionally, the acquisition of land use change data specifically includes:
[0010] The following formula is used to obtain land use change data:
[0011]
[0012] Where C is the rate of land use change, At1 and At2 represent the area of the first type of land use in the initial and target years, respectively, and t2-t1 is the year interval.
[0013] Optionally, constructing an ecological environment quality index (RESI) model based on multiple remote sensing ecological indicators to obtain an initial ecological environment quality index (RESI) specifically includes:
[0014] The following formula is used to obtain the ecological environment quality index:
[0015] RESI=w1NDVI+w2SAVI+w3LST+w4NDWI
[0016] Among them, NDVI is the normalized difference vegetation index, SAVI is the soil adjusted vegetation index, LST is the land surface temperature, NDWI is the normalized difference water index, and w1, w2, w3, and w4 are the index weights corresponding to NDVI, SAVI, LST, and NDWI, respectively.
[0017] Optionally, the landscape pattern index specifically includes:
[0018] Patch density PD, Shannon diversity index SHDI, mean patch area MPS and aggregation index AI.
[0019] Optionally, preprocessing the Sentinel-2A remote sensing data may include:
[0020] Sentinel-2A remote sensing data is processed for radiation correction, atmospheric correction, terrain correction, geometric registration, and cloud and shadow removal.
[0021] The present invention provides an ecological environment quality assessment system, comprising:
[0022] Acquisition module for Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured;
[0023] The classification module is used to classify Sentinel-2A remote sensing data using the random forest supervised classification method to determine the land use classification results;
[0024] The land use change data determination module is used to construct a land use transfer matrix based on the land use classification results, and obtain land use change data through the land use transfer matrix;
[0025] The ecological environment quality assessment result determination module is used to construct an ecological environment quality index RESI model based on multiple types of remote sensing ecological indicators, obtain the initial ecological environment quality index RESI, adjust the initial ecological environment quality index RESI through land use change data and landscape pattern index, and obtain the ecological environment quality assessment results of the tested area.
[0026] The present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned ecological environment quality assessment method is implemented.
[0027] Compared with the existing technology, the beneficial effects of the present invention are as follows: the present invention provides an ecological environment quality assessment method, which obtains Sentinel-2A multi-temporal remote sensing images and adopts a supervised classification method to accurately extract the spatial distribution of land use types over a period of time, and analyzes the dynamic change characteristics of each land use type through the land use transfer matrix, identifying the dominant trend and key conversion path of land use structure adjustment in the study area; by obtaining multiple types of landscape pattern indicators, the landscape pattern evolution characteristics of the measured basin can be obtained; according to the land use classification, landscape pattern analysis and RESI assessment results, the land use changes, landscape pattern evolution and their impact on RESI in the Taoshan Lake Reservoir basin during the study period are revealed. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0029] Figure 1 A flowchart of an ecological environment quality assessment method provided by an embodiment of the present invention;
[0030] Figure 2 A Sankey diagram of land use provided by an embodiment of the present invention;
[0031] Figure 3 A schematic diagram of the impact of land use changes on the reservoir ecological environment provided by an embodiment of the present invention;
[0032] Figure 4Schematic diagram of computer equipment for the ecological environment quality assessment method provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0033] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.
[0034] The following describes in detail the technical solution of the present invention and how the technical solution of the present invention solves the above-mentioned technical problems using specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The following embodiments of the present invention are described in conjunction with the accompanying drawings.
[0035] Figure 1 This is a flow chart of an ecological environment quality assessment method provided by an embodiment of the present invention. Figure 1 As shown, this embodiment shows an ecological environment quality assessment method, including:
[0036] S1: Obtain Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured.
[0037] Optionally, the Sentinel-2A remote sensing data can be processed for radiometric correction, atmospheric correction, terrain correction, geometric registration, and cloud and shadow removal.
[0038] For example, the landscape pattern index may include patch density PD, Shannon diversity index SHDI, mean patch area MPS and aggregation index AI.
[0039] Exemplarily, in order to ensure the accuracy and temporal consistency of remote sensing data, the present invention performs standardized preprocessing on Sentinel-2A images from 2017 to 2023. First, the FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes) model is used to perform radiation correction and atmospheric correction to eliminate the influence of atmospheric scattering and aerosols on remote sensing data, thereby enhancing the comparability of reflectivity data. Secondly, terrain correction is performed based on the high-precision SRTMDEM (Shuttle Radar Topography Mission Digital Elevation Model) to reduce the geometric distortion caused by terrain undulations. Then, high-resolution remote sensing images (such as Landsat-8 and Google Earth images) are used for geometric registration, and the images are uniformly projected to ensure the spatial consistency of multi-temporal data. In addition, MAJA (MACCS-ATCOR Joint Algorithm) is used to remove the influence of clouds and shadows, and interpolation is performed in combination with time series data to reduce the impact of missing data on subsequent analysis.
[0040] S2: Random forest supervised classification method is used to classify Sentinel-2A remote sensing data and determine the land use classification results.
[0041] For example, the present invention uses a Random Forest (RF) supervised classification method to classify land use in remote sensing imagery. The Random Forest algorithm, which relies on voting among multiple decision trees, offers high classification accuracy and robustness, making it particularly suitable for land use classification using multi-source, multi-temporal remote sensing data. Training samples are selected based on high-resolution imagery, historical land use data, and field survey data to ensure representative and accurate classification.
[0042] In terms of feature selection, in addition to traditional ecological indicators (such as NDVI, NDWI, and SAVI), texture features (such as mean and standard deviation) and topographic features (such as slope and elevation) were introduced to enhance classification discriminability. The classification system includes six land use types: cropland, forestland, grassland, waterbodies, built-up land, and unused land. The RF model was trained using the Google Earth Engine (GEE) platform, with 500 decision trees and 10-fold cross-validation to optimize model parameters. Classification accuracy was evaluated using the confusion matrix, overall accuracy (OA), and Kappa coefficient to ensure the reliability of the classification results.
[0043] S3: Based on the land use classification results, a land use transfer matrix is constructed, and land use change data is obtained through the land use transfer matrix.
[0044] For example, the following formula can be used to obtain land use change data:
[0045]
[0046] Where C is the rate of land use change, At1 and At2 represent the area of the first type of land use in the initial and target years, respectively, and t2-t1 is the year interval.
[0047] ArcGIS can be used to map land use changes and analyze the conversion patterns of different land use types to identify key change areas with greater impact on ecological and environmental quality.
[0048] S4: Based on multiple types of remote sensing ecological indicators, an ecological environment quality index RESI model is constructed to obtain the initial ecological environment quality index RESI. The initial ecological environment quality index RESI is adjusted through land use change data and landscape pattern index to obtain the ecological environment quality assessment results of the tested area.
[0049] For example, the RESI model, which integrates the NDVI (normalized difference vegetation index), the soil-adjusted vegetation index (SAVI), the NDWI (normalized difference water index), and the LST (land surface temperature), can be used to comprehensively evaluate vegetation productivity, water stress, and thermal environment effects. The following formula can be used to obtain the ecological environment quality index:
[0050] RESI=w1NDVI+w2SAVI+w3LST+w4NDWI
[0051] Among them, NDVI is the normalized difference vegetation index, SAVI is the soil-adjusted vegetation index, LST is the land surface temperature, and NDWI is the normalized water index. w1, w2, w3, and w4 are the index weights corresponding to NDVI, SAVI, LST, and NDWI, respectively. The index weights can be calculated using the entropy weight method (EWM) to reduce the influence of human subjective factors.
[0052] For example, changes in landscape pattern are an important manifestation of the impact of land use change on the spatial structure and function of ecosystems. After obtaining the landscape pattern index, this application conducts analysis at three levels: patch level, class level, and landscape level, focusing on the following key indices: Patch Density (PD): characterizes the degree of landscape fragmentation; Shannon's Diversity Index (SHDI): measures the diversity and uniformity of the landscape; Mean Patch Size (MPS): reflects landscape connectivity; Aggregation Index (AI): measures the degree of spatial aggregation of landscape types. A decrease in the index indicates increased landscape fragmentation.
[0053] Based on the temporal changes in landscape indices from 2017 to 2023, we compared landscape pattern characteristics across different years and analyzed the temporal trends in landscape fragmentation, ecological connectivity, and landscape diversity. Combined with land use change data, we explored the contribution of different land use types to landscape pattern changes and identified ecologically sensitive areas.
[0054] The evolution of land use patterns directly reflects the dynamics of watershed ecosystems, directly impacting hydrological processes, ecological connectivity, and ecosystem services in water source areas. Based on classification results from Sentinel-2A remote sensing imagery from 2017 to 2023, the main land use types in the tested watershed include cropland, forestland, grassland, water bodies, built-up land, and unused land. The results indicate that the watershed's land use pattern underwent significant changes during the experimental period, exhibiting characteristics of urbanization-driven land use transformation.
[0055] For example, land use change (LUCC) directly alters land cover and heat and water conditions, which in turn changes NDVI, LST, and NDWI values, and RESI rises or falls accordingly without the need for additional intermediate quantities. The effects of land use change (LUCC) on RESI are shown in Table 1.
[0056] Table 1 Effect of land use change LUCC on RESI
[0057]
[0058] The effect of landscape pattern index LPI on RESI is shown in Table 2.
[0059] Table 2 Effect of landscape pattern index LPI on RESI
[0060]
[0061] The effect pathways of the landscape pattern index LPI on RESI include pattern improvement (AI↑, MPS↑) → more continuous vegetation, greater transpiration → increased NDVI and NDWI, decreased LST → increased RESI; and vice versa.
[0062] Therefore, land use changes first change vegetation, water and thermal conditions; the landscape pattern index describes whether these changes are continuous or fragmented; the four remote sensing indicators rise and fall accordingly, and RESI is just their weighted summary, so "LUCC+LPI→indicator change→RESI value change" directly determines the ecological environment quality assessment result, that is, the final ecological quality level.
[0063] like Figure 2 The following is a Sankey diagram of land use change, which mainly shows how land use changes and from which type to which type of land use; Figure 3 (a) to Figure 3 Panel (c) shows seasonal changes in RESI ecological and environmental quality across different sub-basins within the basin. Spatially, in 2017, forestland and cultivated land dominated the basin's land use, accounting for X% and Y% of the total area, respectively. Water area remained relatively stable, with a zonal distribution along the reservoir perimeter. However, with economic development and increased human activity, by 2023, construction land had significantly expanded, while forestland and cultivated land areas had decreased. Land-use shifts were particularly dramatic in the lower reaches of the basin and along the river, as shown in Table 3. Water area fluctuated slightly during the study period, likely influenced by factors such as precipitation, land development, and hydrological regulation.
[0064] Table 3 Land use statistics
[0065] 2017 2018 2019 2020 2021 2022 2023 water bodies 34.09 35.25 38.51 37.37 30.33 27.1 31.48 woodland 922.11 943.67 954.66 978.53 975.77 963.6 949.99 grassland 6.72 5.82 7.53 7.8 8.35 9.34 9.64 wetlands 1.82 1.55 1.23 0.92 3.78 7.89 3.63 arable land 1007.25 985.27 971.03 952.7 959.42 966.59 980.67 construction land 67.98 70.77 70.49 66.09 65.47 68.73 67.79 bare land 3.98 1.63 0.49 0.56 0.83 0.71 0.77
[0066] By analyzing the land use transfer matrix, we can further reveal the conversion trend of various land use types, as shown in Table 4. The results show that the conversion of cultivated land to construction land is the most significant. Between 2017 and 2023, a total of X km 2 of cultivated land was converted to construction land, accounting for Y% of the total cultivated land area. This conversion was mainly concentrated in the downstream areas of reservoirs and around transportation corridors, indicating that urbanization and infrastructure expansion were the main drivers of cultivated land reduction.
[0067] Table 4 Land use transfer matrix
[0068]
[0069]
[0070]
[0071] In addition, the conversion of forest land to other land use types was also obvious. 2 The conversion of forestland into grassland or construction land is particularly pronounced in the hilly areas of the western and southeastern parts of the basin, resulting in significant vegetation destruction and ecological landscape fragmentation. Because forestland plays an important role in water conservation and purification, its reduction could lead to increased soil erosion, impacting reservoir water quality and ecological stability.
[0072] In contrast, the water area remained relatively stable during the study period, with only B km 2 Seasonal changes in reservoir storage may be the primary factor driving fluctuations in water area. Furthermore, some grasslands and unused lands experienced relatively small changes, but their spatial distribution did undergo some adjustments, potentially influenced by policy interventions (such as ecological restoration projects) or natural processes.
[0073] This application examines land use change and landscape pattern evolution within the reservoir basin, using Sentinel-2A multi-temporal remote sensing data from 2017 to 2023, and their impact on ecological and environmental quality (RESI). The study will employ supervised classification, landscape index analysis, Remote Sensing Ecological Index (RESI) evaluation, and correlation analysis to quantitatively assess the mechanisms by which different land use types and landscape pattern characteristics influence ecological and environmental quality. The results of this study will not only reveal evolving trends in ecological and environmental quality within the basin but also provide scientific support for water source ecological protection, land use optimization, and regional sustainable development.
[0074] The above is an ecological environment quality assessment method provided in one or more embodiments of this specification. Based on the same idea, this specification also provides a corresponding ecological environment quality assessment system, including:
[0075] Acquisition module for Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured;
[0076] The classification module is used to classify Sentinel-2A remote sensing data using the random forest supervised classification method to determine the land use classification results;
[0077] The land use change data determination module is used to construct a land use transfer matrix based on the land use classification results, and obtain land use change data through the land use transfer matrix;
[0078] The ecological environment quality assessment result determination module is used to construct an ecological environment quality index RESI model based on multiple types of remote sensing ecological indicators, obtain the initial ecological environment quality index RESI, adjust the initial ecological environment quality index RESI through land use change data and landscape pattern index, and obtain the ecological environment quality assessment results of the tested area.
[0079] The specific limitations of the ecological environment quality assessment system can be found in the limitations of the ecological environment quality assessment method described above and will not be further elaborated here. Each module in the aforementioned ecological environment quality assessment system may be implemented in whole or in part through software, hardware, or a combination thereof. Each of the aforementioned modules may be embedded in or independent of a processor in a computer device in hardware form, or may be stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.
[0080] The present invention also provides Figure 4 The structural diagram of the computer equipment shown in FIG. Figure 4 As shown, at the hardware level, the computer device includes a processor, an internal bus, a network interface, memory, and non-volatile storage, and may also include hardware required for other services. The processor reads the corresponding computer program from the non-volatile storage into the memory and then runs it to implement the ecological environment quality assessment method provided in the above embodiment.
[0081] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present invention.
Claims
1. A method for evaluating ecological environment quality, characterized in that: include: Obtain Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured; The random forest supervised classification method was used to classify the Sentinel-2A remote sensing data and determine the land use classification results; Based on the land use classification results, a land use transfer matrix is constructed, and land use change data is obtained through the land use transfer matrix; Based on multiple types of remote sensing ecological indicators, an ecological environment quality index RESI model is constructed to obtain the initial ecological environment quality index RESI. The initial ecological environment quality index RESI is adjusted through land use change data and landscape pattern index to obtain the ecological environment quality assessment results of the tested area.
2. The ecological environment quality assessment method according to claim 1, characterized in that: The acquisition of the land use change data specifically includes: The following formula is used to obtain land use change data: Where C is the rate of land use change, At1 and At2 represent the area of the first type of land use in the initial and target years, respectively, and t2-t1 is the year interval.
3. The ecological environment quality assessment method according to claim 1, characterized in that: The aforementioned construction of an ecological environment quality index RESI model based on multiple types of remote sensing ecological indicators to obtain an initial ecological environment quality index RESI specifically includes: The following formula is used to obtain the ecological environment quality index: RESI=w1NDVI+w2SAVI+w3LST+w4NDWI Among them, NDVI is the normalized difference vegetation index, SAVI is the soil adjusted vegetation index, LST is the land surface temperature, NDWI is the normalized difference water index, and w1, w2, w3, and w4 are the index weights corresponding to NDVI, SAVI, LST, and NDWI, respectively.
4. The ecological environment quality assessment method according to claim 1, characterized in that: The landscape pattern index specifically includes: Patch density PD, Shannon diversity index SHDI, mean patch area MPS and aggregation index AI.
5. The ecological environment quality assessment method according to claim 1, characterized in that: Preprocessing the Sentinel-2A remote sensing data includes: Sentinel-2A remote sensing data is processed for radiation correction, atmospheric correction, terrain correction, geometric registration, and cloud and shadow removal.
6. An ecological environment quality assessment system, characterized in that: include: Acquisition module for Sentinel-2A remote sensing data, landscape pattern index and multiple remote sensing ecological indicators of the area to be measured; The classification module is used to classify Sentinel-2A remote sensing data using the random forest supervised classification method to determine the land use classification results; The land use change data determination module is used to construct a land use transfer matrix based on the land use classification results, and obtain land use change data through the land use transfer matrix; The ecological environment quality assessment result determination module is used to construct an ecological environment quality index RESI model based on multiple types of remote sensing ecological indicators, obtain the initial ecological environment quality index RESI, adjust the initial ecological environment quality index RESI through land use change data and landscape pattern index, and obtain the ecological environment quality assessment results of the tested area.
7. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for evaluating the ecological environment quality described in any one of claims 1 to 5 is implemented.