A method for evaluating the influence of photovoltaic laying on grassland ecological system based on multi-source remote sensing data

Abstract

This invention discloses a method for assessing the impact of photovoltaic (PV) installations on grassland ecosystems based on multi-source remote sensing data. The method includes the following steps: constructing a multi-source remote sensing dataset of the PV installation area and surrounding grasslands; acquiring time series data of multi-dimensional ecological indicators such as vegetation, water, and energy; establishing multi-ring buffer zones based on the PV boundary, and extracting the variation characteristics of each ecological indicator within and outside the PV area and at different distances; and quantitatively assessing the intensity and spatial decay characteristics of the impact of PV installations on the grassland ecosystem through spatial difference analysis and comprehensive ecological impact indicators. This method enables long-term dynamic monitoring of the ecological effects of PV construction without ground-based observation, and has the advantages of high automation, strong scalability, and wide application range, providing scientific support for the ecological assessment and management of grassland renewable energy development.

Description

Technical Field

[0001] This invention belongs to the field of ecological environment remote sensing and ecological assessment technology, specifically involving a quantitative assessment method for the ecological effects of photovoltaic installation areas based on multi-source remote sensing data. It is applicable to the ecological monitoring and impact analysis of large-scale photovoltaic facilities in grasslands, desertification areas and ecologically sensitive areas. Background Technology

[0002] With the rapid development of the photovoltaic power generation industry, large-scale photovoltaic power plants have been constructed extensively in the arid and semi-arid grassland regions of northern my country. While photovoltaic panels alter the surface energy balance, heat transfer, and water cycle, they may also significantly impact vegetation growth, soil moisture, and surface temperature within grassland ecosystems. Existing research indicates that photovoltaic facilities can reduce evapotranspiration and mitigate drought stress to some extent through shading, but they can also lead to localized surface disturbance and vegetation degradation. Therefore, how to scientifically quantify the ecological impacts of photovoltaic installations based on objective, long-term observational data is a crucial issue in ecological management and sustainable energy development.

[0003] Traditional ecological impact assessments rely on ground surveys and sample plot monitoring, which have limited spatial coverage and insufficient temporal continuity, making them unsuitable for comprehensive assessments of ecological changes in large-scale photovoltaic (PV) fields and their surrounding areas. In recent years, the rapid development of remote sensing technology has provided continuous and objective spatial information support for ecological and environmental monitoring. Through the fusion of multi-source satellite data, multi-dimensional ecological parameters such as surface reflectance, vegetation index, surface temperature, soil moisture, and productivity can be obtained. However, a systematic remote sensing assessment method for the long-term impacts of PV facility construction and operation on grassland ecosystems is currently lacking.

[0004] To address this, this invention proposes a remote sensing-based method for assessing the impact of photovoltaic (PV) installations on grassland ecosystems. This method comprehensively utilizes multi-source remote sensing data, including optical, radar, and other sources, to construct a multi-source spatiotemporal ecological indicator system. Combined with spatial zonation analysis and time-series change detection techniques, it performs multi-scale quantitative assessments of the ecological response of PV areas and their surrounding regions. By establishing a multi-ring buffer structure and spatial attenuation model for PV boundaries, this invention can reveal the direct and indirect impacts of PV facilities on grassland ecosystems, providing a scientific basis for regional ecological security assessments and renewable energy planning. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing photovoltaic ecological assessment methods, such as poor spatial continuity, reliance on ground data, and lack of automated quantitative analysis. This invention proposes a method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data. This method fully utilizes multi-source satellite data, including optical and radar data, to construct a spatiotemporal ecological indicator system. Through spatial zoning and time-series change detection, it achieves quantitative assessment and spatial attenuation analysis of grassland ecological responses before and after photovoltaic construction.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: Step A: Construction of multi-source remote sensing dataset: Acquire multi-source remote sensing data of the photovoltaic installation area boundary and the surrounding grassland area, perform atmospheric correction and cloud mask preprocessing on the remote sensing data, and construct a multi-dimensional remote sensing data cube with unified spatial and temporal resolution; Step B: Calculation of multi-dimensional ecological indicators: Based on the aforementioned multidimensional remote sensing data structure, time series of ecological indicators related to vegetation, water, and energy are extracted. The ecological indicators include near-infrared vegetation reflectance index (NIRv), normalized difference in water volume index (NDWI), bare soil index (BSI), land surface temperature (LST), VV / VH ratio index (PRI), and photosynthetic productivity (GPP). Step C: Establish a spatial zonation structure based on photovoltaic location and extract index changes: Multiple annular buffer zones are generated based on the photovoltaic installation boundary. The statistical characteristics of each ecological indicator are extracted inside and outside the photovoltaic area and within the annular buffer zones at different distances. The differences of each ecological indicator before and after photovoltaic construction are calculated. Step D: Based on the year of photovoltaic installation, quantitatively assess the overall impact of photovoltaic installation on the grassland ecosystem using spatial difference analysis and the Integrated Ecological Impact Index (EII): The time periods before and after photovoltaic construction are determined, the rate of change of each ecological indicator in the photovoltaic area and each ring buffer zone is calculated, a spatial attenuation model of ecological change and distance is fitted, the direction and intensity of the impact are determined through significance testing, and finally the comprehensive ecological impact index (EII) is calculated to achieve quantitative assessment.

[0007] Preferably, in step A, the multi-source remote sensing data includes Sentinel-2 surface reflectance products, Landsat 8 / 9 surface reflectance images, MODISGPP products, MODIS surface temperature products, and Sentinel-1 radar data. Data preprocessing also includes generating cloud-free reflectance composite images from optical remote sensing data, performing radiometric calibration, terrain correction, and Lee filtering denoising on radar images, spatial registration and temporal alignment of MODIS products, and uniform resampling to 30m resolution, synthesizing data monthly or quarterly. The standardized multi-source data are then resampled, registered, and stacked according to spatial location and temporal order to construct a four-dimensional remote sensing data cube. , Where (x, y) are spatial coordinates, (t) is the time index, and (b) is the band or index type.

[0008] Preferably, in step B, the ecological indicators are calculated using the following methods: Near-infrared vegetation reflectance index (NIRv): , Where NIR is the reflectance in the near-infrared band and RED is the reflectance in the red band; Normalized Difference Water Index (NDWI): , Where SWIR is the reflectance in the short-wave infrared band; Bare Soil Index (BSI): , Where BLUE represents the reflectivity of the blue light band; VV / VH ratio index (PRI): , in, The backscattering intensity of radar waves under VV polarization conditions is expressed in dB. The backscattering intensity of radar waves under VH polarization conditions is expressed in dB. Indicates the spatial pixel position; Indicates the time of image acquisition; Land surface temperature (LST) was extracted from MODIS or Landsat thermal infrared products, and seasonal mean and interannual variation were calculated. Photosynthetic productivity (GPP) was extracted from MODIS or Landsat GPP products.

[0009] Preferably, in step C, the buffer radius of the annular buffer zone ranges from 0-50m, 50-200m, 200-500m, to 500-1000m; the statistical characteristics of the ecological indicators within each annular buffer zone include the mean, standard deviation, and time series trend slope, and an annular grouping data table is constructed. , Where ZoneID is the buffer number, Distance is the buffer distance, Indicator is the remote sensing indicator, Year is the year, Mean is the average value of each indicator, Std is the standard deviation of each indicator, and Trend is the time series trend slope of each indicator. The formula for calculating the difference between various ecological indicators before and after photovoltaic construction is as follows: , in For the ecological index values ​​after construction, These are the ecological indicator values ​​before construction.

[0010] Preferably, in step D, the time periods before and after construction are determined by photovoltaic installation year or time series breakpoint detection (BFAST algorithm); The formula for calculating the rate of change of ecological indicators is: , in To measure the difference in ecological indicators before and after construction, These are the ecological indicator values ​​before construction.

[0011] Preferably, in step D, the expression for the spatial attenuation model is: , in The initial effect magnitude, The spatial half-life of ecological impact. This is the error term.

[0012] Preferably, in step D, the significance test is performed using the Mann-Kendall test or the t-test, and the main impact direction and sensitive variables of photovoltaic installation on the grassland ecosystem are determined by Pearson correlation analysis or Spearman correlation analysis.

[0013] Preferably, in step D, the formula for calculating the Integrated Ecological Impact Index (EII) is as follows: , in, These are the standardized values ​​of the rate of change for each indicator. To assign corresponding weights, the main ecological variables (such as NIRv or GPP) are used as references, and the Pearson correlation coefficients of each indicator are used for calculation. EII > 0 indicates that the grassland ecosystem in the photovoltaic installation area has improved, and EII < 0 indicates that the grassland ecosystem in the photovoltaic installation area has degraded.

[0014] More preferably, the output format of the multidimensional remote sensing data structure is NetCDF format or GeoTIFF sequence format; The ecological indicator time series output in step B is a time-series ecological indicator raster dataset, specifically: , It also outputs the annual synthetic product and trend sequence.

[0015] Preferably, in step A, the research scope of the grassland area surrounding the photovoltaic installation area is the grassland area extending 0.5-5 km beyond the photovoltaic installation area; the method is applicable to the impact assessment of large-scale photovoltaic facilities on grassland ecosystems in grasslands, desertified areas and ecologically sensitive areas.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention integrates optical, radar, and thermal infrared multi-source remote sensing data to achieve multi-dimensional monitoring of ecological structure and function; it employs spatiotemporal data reconstruction and ring-zone statistical algorithms to automatically generate zoning indicators and time-series change results; it achieves large-scale quantitative assessment of ecological effects under pure remote sensing conditions without relying on ground sampling points; and it reveals the direct and indirect impact range of photovoltaic facilities on grassland ecosystems through a spatial attenuation model, supporting scientific planning and environmental management. Attached Figure Description

[0017] Figure 1 This is a flowchart of the overall process of the method of the present invention.

[0018] Figure 2 This is a graph showing the differences in ecological indicators.

[0019] Figure 3 Distribution map of the comprehensive ecological impact index before and after photovoltaic installation. Detailed Implementation

[0020] The following description, in conjunction with the accompanying drawings, further illustrates the implementation process of a method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data, as proposed in this invention. The described embodiments are merely some, not all, of the embodiments of this invention.

[0021] Example In this embodiment, a typical large-scale photovoltaic power station in Qinghai Province is taken as the research object. The research scope covers the photovoltaic facility area and its buffer zone of 5km. Generally, the microclimate and disturbance changes caused by photovoltaics usually have a strong impact within the site and at the edge for hundreds of meters and 1 to 2km beyond. However, in order to avoid missing the weak far-field effects of downwind or terrain channels, a larger reference range is set here, namely the above-mentioned 5km. If the location to be evaluated is changed, adaptive settings can be made according to the data resolution used, the photovoltaic installation range, etc.

[0022] like Figure 1 As shown, step A, construction of multi-source remote sensing dataset: First, boundary data (Shapefil or GeoJSON format) of the grassland area extending 5-10km beyond the photovoltaic installation area and multi-source remote sensing data of the surrounding grassland area were acquired. The multi-source remote sensing data included Sentinel-2 surface reflectance products (spatial resolution 10m), Landsat 8 / 9 surface reflectance images (spatial resolution 30m), MODISGPP products (MOD17A2H, 500m, 8-day temporal resolution), MODIS surface temperature products (MOD11A2, 1km, 8-day temporal resolution), and Sentinel-1 radar data (VV / VH polarization, 10m). Then, atmospheric correction and cloud masking preprocessing are performed on the optical remote sensing data. Atmospheric correction uses the Sen2Cor or LEDAPS algorithm, and cloud masking uses the CFMASK algorithm to remove cloud and shadow pixels. Radiometric calibration, terrain correction, and Lee filtering are performed on the radar imagery. The MODIS products are spatially registered and temporally aligned, and uniformly resampled to 30m resolution, and synthesized monthly.

[0023] After standardization, the above data is used to construct a multidimensional remote sensing data cube with unified spatial and temporal resolution (in NetCDF or GeoTIFF format). , Where (x, y) are spatial coordinates, (t) is the time index, and (b) is the band or index type.

[0024] Step B: Calculation of multi-dimensional ecological indicators: Based on the multidimensional remote sensing data cube, time series of ecological indicators related to vegetation, water and energy are extracted. The ecological indicators include near-infrared vegetation reflectance index (NIRv), normalized difference in water volume index (NDWI), bare soil index (BSI), land surface temperature (LST), VV / VH ratio index (PRI), and photosynthetic productivity (GPP). The calculation methods for ecological indicators include: Near-infrared vegetation reflectance index (NIRv): , Where NIR is the reflectance in the near-infrared band and RED is the reflectance in the red band; Normalized Difference Water Index (NDWI): , Where SWIR is the reflectance in the short-wave infrared band; Bare Soil Index (BSI): , Where BLUE represents the reflectivity of the blue light band; VV / VH ratio index (PRI): , in, The backscattering intensity of radar waves under VV polarization conditions is expressed in dB. The backscattering intensity of radar waves under VH polarization conditions is expressed in dB. Indicates the spatial pixel position; Indicates the time of image acquisition; Land surface temperature (LST) was extracted from MODIS or Landsat thermal infrared products, and seasonal mean and interannual variation were calculated. Photosynthetic productivity (GPP) was extracted from MODIS or Landsat GPP products.

[0025] The above ecological indicator time series is a time-series ecological indicator grid dataset, specifically: , It also outputs the annual synthetic product and trend sequence.

[0026] Step C: Establish a spatial zonation structure based on photovoltaic location and extract index changes: Multiple annular buffer zones are generated based on the photovoltaic installation boundary. The statistical characteristics of each ecological indicator are extracted inside and outside the photovoltaic area and within the annular buffer zones at different distances. The differences of each ecological indicator before and after photovoltaic construction are calculated. The buffer radii of the aforementioned annular buffer zones range from 0-50m, 50-200m, 200-500m, and 500-1000m. The statistical characteristics of ecological indicators within each annular buffer zone include mean, standard deviation, and time-series trend slope. A grouped data table of annular zones is constructed. , Where ZoneID is the buffer number, Distance is the buffer distance, Indicator is the remote sensing indicator, Year is the year, Mean is the average value of each indicator, Std is the standard deviation of each indicator, and Trend is the time series trend slope of each indicator. The formulas for calculating the differences in various ecological indicators before and after photovoltaic construction are as follows: , in For the ecological index values ​​after construction, These are the ecological indicator values ​​before construction.

[0027] Step D: Based on the year of photovoltaic installation, quantitatively assess the overall impact of photovoltaic installation on the grassland ecosystem using spatial difference analysis and the Integrated Ecological Impact Index (EII): By using photovoltaic installation year or time series breakpoint detection (BFAST algorithm) to determine the time period before and after construction, the rate of change of each ecological indicator in the photovoltaic area and each ring buffer zone is calculated, a spatial decay model of ecological change and distance is fitted, the direction and intensity of the impact are determined through significance test, and finally the comprehensive ecological impact index (EII) is calculated to achieve quantitative assessment.

[0028] The formula for calculating the rate of change of the above ecological indicators is as follows: , in To measure the difference in ecological indicators before and after construction, Ecological indicator values ​​before construction; The expression for the above spatial attenuation model is: , in The initial effect magnitude, The spatial half-life of ecological impact. This is the error term; The significance tests mentioned above used the Mann-Kendall test or the t-test, and the main impact direction and sensitive variables of photovoltaic installation on grassland ecosystems were determined by Pearson correlation analysis or Spearman correlation analysis. The formula for calculating the comprehensive ecological impact index (EII) is as follows: , in, These are the standardized values ​​of the rate of change for each indicator. To assign corresponding weights, the main ecological variables (such as NIRv or GPP) are used as references, and the Pearson correlation coefficients of each indicator are used for calculation. EII > 0 indicates that the grassland ecosystem in the photovoltaic installation area has improved, and EII < 0 indicates that the grassland ecosystem in the photovoltaic installation area has degraded.

[0029] After implementing the above methods, the following three types of results were obtained: (1) The difference plots and significance distribution results of each ecological indicator are used to identify areas of significant ecological change. Please refer to [the relevant documentation]. Figure 2 The figure shows the distribution of the difference between the NIRv index before and after photovoltaic installation. The index of regular photovoltaic fields increased significantly after photovoltaic installation compared with other areas, indicating that after the construction of photovoltaic facilities, the remote sensing vegetation index can preliminarily characterize the grassland status of the study area as showing a tendency to stabilize or slightly improve. (2) Spatial half-life distance table, used to quantify the characteristics of ecological impact decay. Please refer to Table 1. This table is a spatial half-life table of three indices. The results show that the comprehensive ecological improvement effect near the boundary of the photovoltaic area is the most significant, with a spatial half-life distance of about 500m. The positive improvement effect of NDVI and NDWI can be sustained to the range of 400-600m, while LST is concentrated within about 300m of the photovoltaic area and decays by half. This reflects that the photovoltaic facilities have a certain spatial diffusion improvement effect on vegetation growth, water retention and thermal environment during the operation phase. (3) Ecological Impact Comprehensive Index Distribution Map, describing the overall changes in the grassland ecosystem before and after photovoltaic construction. Please refer to [link / reference]. Figure 3The figure shows the spatial distribution of the ecological impact index before and after the installation of photovoltaic facilities. The area inside and adjacent to the photovoltaic field is dominated by positive and near-zero pixels, with obvious blue block structures in some areas. This indicates that the installation of photovoltaic facilities has not caused significant ecological degradation, but has instead created a certain positive ecological effect or buffering effect in some areas.

[0030] Table 1

Claims

1. A method for assessing the impact of photovoltaic (PV) installation on grassland ecosystems based on multi-source remote sensing data, characterized in that, Includes the following steps: Step A: Construction of multi-source remote sensing dataset: Acquire multi-source remote sensing data of the photovoltaic installation area boundary and the surrounding grassland area, perform atmospheric correction and cloud mask preprocessing on the remote sensing data, and construct a multi-dimensional remote sensing data cube with unified spatial and temporal resolution; Step B: Calculation of multi-dimensional ecological indicators: Based on the aforementioned multidimensional remote sensing data structure, time series of ecological indicators related to vegetation, water, and energy are extracted. The ecological indicators include near-infrared vegetation reflectance index (NIRv), normalized difference in water volume index (NDWI), bare soil index (BSI), land surface temperature (LST), VV / VH ratio index (PRI), and photosynthetic productivity (GPP). Step C: Establish a spatial zonation structure based on photovoltaic location and extract index changes: Multiple annular buffer zones are generated based on the photovoltaic installation boundary. The statistical characteristics of each ecological indicator are extracted inside and outside the photovoltaic area and within the annular buffer zones at different distances. The differences of each ecological indicator before and after photovoltaic construction are calculated. Step D: Based on the year of photovoltaic installation, quantitatively assess the overall impact of photovoltaic installation on the grassland ecosystem using spatial difference analysis and the Integrated Ecological Impact Index (EII): The time periods before and after photovoltaic construction are determined, the rate of change of each ecological indicator in the photovoltaic area and each ring buffer zone is calculated, a spatial attenuation model of ecological change and distance is fitted, the direction and intensity of the impact are determined through significance testing, and finally the comprehensive ecological impact index (EII) is calculated to achieve quantitative assessment.

2. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data as described in claim 1, characterized in that, In step A, the multi-source remote sensing data includes Sentinel-2 surface reflectance products, Landsat 8 / 9 surface reflectance images, MODISGPP products, MODIS surface temperature products, and Sentinel-1 radar data. Data preprocessing also includes generating cloud-free reflectance composite images from optical remote sensing data, performing radiometric calibration, terrain correction, and Lee filtering denoising on radar images, spatial registration and temporal alignment of MODIS products, and uniform resampling to 30m resolution, synthesizing data monthly or quarterly. The standardized multi-source data are then resampled, registered, and stacked according to spatial location and temporal order to construct a four-dimensional remote sensing data cube. ， Where (x, y) are spatial coordinates, (t) is the time index, and (b) is the band or index type.

3. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step B, the ecological indicators are calculated using the following methods: Near-infrared vegetation reflectance index (NIRv): ， Where NIR is the reflectance in the near-infrared band and RED is the reflectance in the red band; Normalized Difference Water Index (NDWI): ， Where SWIR is the reflectance in the short-wave infrared band; Bare Soil Index (BSI): ， Where BLUE represents the reflectivity of the blue light band; VV / VH ratio index (PRI): ， in, The backscattering intensity of radar waves under VV polarization conditions is expressed in dB. The backscattering intensity of radar waves under VH polarization conditions is expressed in dB. Indicates the spatial pixel position; Indicates the time of image acquisition; Land surface temperature (LST) was extracted from MODIS or Landsat thermal infrared products, and seasonal mean and interannual variation were calculated. Photosynthetic productivity (GPP) was extracted from MODIS or Landsat GPP products.

4. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step C, the buffer radius of the annular buffer zone ranges from 0-50m, 50-200m, 200-500m, to 500-1000m; the statistical characteristics of ecological indicators within each annular buffer zone include mean, standard deviation, and time series trend slope, and an annular grouping data table is constructed. ， Where ZoneID is the buffer number, Distance is the buffer distance, Indicator is the remote sensing indicator, Year is the year, Mean is the average value of each indicator, Std is the standard deviation of each indicator, and Trend is the time series trend slope of each indicator. The formula for calculating the difference between various ecological indicators before and after photovoltaic construction is as follows: ， in For the ecological index values ​​after construction, These are the ecological indicator values ​​before construction.

5. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step D, the time periods before and after construction are determined by photovoltaic installation year or time series breakpoint detection (BFAST algorithm); The formula for calculating the rate of change of ecological indicators is: ， in To measure the difference in ecological indicators before and after construction, These are the ecological indicator values ​​before construction.

6. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step D, the expression for the spatial attenuation model is: ， in The initial effect magnitude, The spatial half-life of ecological impact. This is the error term.

7. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step D, the significance test is performed using the Mann-Kendall test or the t-test, and the main impact direction and sensitive variables of photovoltaic installation on the grassland ecosystem are determined by Pearson correlation analysis or Spearman correlation analysis.

8. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step D, the formula for calculating the Integrated Ecological Impact Index (EII) is as follows: ， in, These are the standardized values ​​of the rate of change for each indicator. To assign corresponding weights, the main ecological variables (such as NIRv or GPP) are used as references, and the Pearson correlation coefficients of each indicator are used for calculation. EII > 0 indicates that the grassland ecosystem in the photovoltaic installation area has improved, and EII < 0 indicates that the grassland ecosystem in the photovoltaic installation area has degraded.

9. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 2, characterized in that, The output format of the multidimensional remote sensing data structure is either NetCDF format or GeoTIFF sequence format; The ecological indicator time series output in step B is a time-series ecological indicator raster dataset, specifically: ， It also outputs the annual synthetic product and trend sequence.

10. The method for assessing the impact of photovoltaic installation on grassland ecosystems based on multi-source remote sensing data according to claim 1, characterized in that, In step A, the research scope of the grassland area surrounding the photovoltaic installation area is the grassland area extending 0.5-5 km beyond the photovoltaic installation area; the method is applicable to the impact assessment of large-scale photovoltaic facilities on grassland ecosystems in grasslands, desertified areas and ecologically sensitive areas.

Images