An ecosystem gross production value accounting method and device based on GIS spatial analysis and a storage medium

Abstract

The application relates to the technical field of value accounting, and discloses an ecosystem gross domestic product accounting method based on GIS spatial analysis, which comprises the following steps: accounting for regional basic data and satellite image data, constructing a two-way dynamic semantic mapping library corresponding to land change survey block classification-ecosystem type; obtaining Internet of Things monitoring data and statistical report data, performing spatial interpolation-boundary correction and sliding window smoothing space-time fusion to obtain a standardized data set; constructing a double-level accounting index system of urban areas and typical areas, and synchronously calculating physical quantity and value quantity; based on GIS, a model correction factor is called to quantize the index to a grid unit, neighborhood analysis is performed to obtain a result, and after the parameters are calibrated, a multi-year ecological value trend graph is generated in combination with GIS; customized results are output according to scenes such as ecological compensation, and after three-level quality control verification, the final results are output. The application can meet the needs of accounting accuracy and practicability of ecological compensation, EOD projects and the like.

Description

Technical Field

[0001] This application relates to the technical field of value accounting, and in particular to a method, device and storage medium for calculating the gross ecosystem product based on GIS spatial analysis. Background Technology

[0002] In the current accounting of Gross Ecosystem Product (GEP), there are semantic standard differences between land survey and ecological classification data, the static mapping relationship cannot be updated synchronously with the annual land data, and the lack of a unified fusion mechanism for multi-source data due to differences in spatiotemporal precision leads to large deviations in the basic accounting data.

[0003] At the same time, the existing indicator system is difficult to adapt to the differentiated needs of the entire city and special areas. Parameter calibration relies on manual methods, spatial expression is insufficient, and the ability to output scenarios is weak.

[0004] In summary, existing GEP accounting suffers from problems such as difficulty in data integration, poor indicator adaptability, lack of dynamic parameter calibration, and insufficient scenario-based output, making it difficult to meet the needs of ecological compensation, EOD projects, and other practical applications for accounting accuracy and practicality. As can be seen, how to meet the needs of ecological compensation, EOD projects, and other practical applications for accounting accuracy and practicality remains to be solved. Summary of the Invention

[0005] To meet the needs of ecological compensation, EOD projects and other practical applications for accounting accuracy and practicality, this application provides a method, equipment and storage medium for calculating the Gross Ecosystem Product based on GIS spatial analysis.

[0006] Firstly, this application provides a method for calculating the gross ecosystem product based on GIS spatial analysis, employing the following technical solution:

[0007] A method for calculating the gross ecosystem product based on GIS spatial analysis, comprising:

[0008] The accounting area's basic data and satellite imagery data are used to construct a two-way dynamic semantic mapping library corresponding to the land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes a primary mapping and a secondary mapping. The primary mapping associates the land and ecosystem primary categories, and the secondary mapping achieves the alignment of secondary category terms through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with the annual land data.

[0009] The corresponding multi-source data is acquired, including land change survey vector data, satellite remote sensing raster data of the required accuracy, IoT real-time monitoring data and statistical report data. The monitoring data includes soil moisture, air quality and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed through spatial interpolation-boundary correction and sliding window smoothing method to obtain the corresponding standardized dataset.

[0010] A two-tiered accounting indicator system is constructed, consisting of urban areas and typical districts. The urban area level includes three primary indicators: material products, regulatory services, and cultural services, as well as 14 secondary indicators. The typical district level is configured with exclusive secondary indicators for four types of characteristic areas, and these exclusive secondary indicators are enabled or disabled through dynamic switching. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model.

[0011] Based on GIS, pre-trained digital elevation models and 3D building models are retrieved, parameters are extracted to correct soil erosivity factor values ​​and vegetation transpiration, and the indicators are quantified to a preset precision raster unit. The raster-level ecological value results are obtained by processing the neighborhood analysis algorithm. The deviation rate between measured and calculated values ​​is calculated by accessing IoT monitoring data, and indicators exceeding the threshold are identified. The random forest algorithm is used to calibrate the calculation parameters and control the deviation rate within a preset range. Combined with GIS, a multi-year ecological value trend map is generated.

[0012] Customized results are output based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios, and are verified through a three-level quality control mechanism: data level to complete missing data, calculation level to intercept abnormal parameters, and result level to output the final result after passing the sampling verification of third-party measured data.

[0013] Optionally, the method further includes:

[0014] Obtain historical ecological data for the accounting area, including annual climate change data, vegetation cover change data, and land use transformation data.

[0015] An ecological value decay coefficient model is constructed, and historical ecological data is compared with current accounting data over time to calculate the value decay coefficient for different ecosystem types.

[0016] Based on the value decay coefficient, the current accounting results are corrected over time to generate comparable ecosystem GDP accounting results across years.

[0017] Optionally, the method further includes:

[0018] Identify ecologically sensitive areas within the accounting area, including nature reserves, wetland core areas, and habitats of endangered species;

[0019] A unique accounting weighting factor is assigned to ecologically sensitive areas, and the weighting factor is determined based on the protection level and the scarcity of ecosystem services in the ecologically sensitive areas.

[0020] By embedding special accounting weighting factors into a two-level accounting indicator system, the accounting results for ecologically sensitive areas are adjusted with weights.

[0021] Optionally, the method further includes:

[0022] Collect human activity data for the accounting area, including population density distribution, industrial layout data, and transportation network density;

[0023] Construct a human activity intensity index model to quantify human activity data into a standardized intensity index;

[0024] Spatial corrections are made to the ecological value quantification results based on the human activity intensity index to generate a heat map of activity impacts.

[0025] Optionally, the method further includes:

[0026] Obtain historical data on natural disasters in the accounting area, including flood frequency, drought intensity, and geological erosion degree;

[0027] Establish a natural disaster risk assessment matrix and convert historical natural disaster data into ecological value loss risk coefficients;

[0028] By incorporating the risk coefficient into the spatiotemporal premium parameter correction mechanism, an ecosystem GDP accounting result containing a risk premium is generated.

[0029] Optionally, the method further includes:

[0030] Construct an ecosystem service function flow network model to identify the supply areas, transmission paths and beneficiary areas of ecosystem services within the accounting region;

[0031] Calculate the spatial flow volume and value transfer efficiency of ecosystem services based on spatial flow analysis algorithms;

[0032] The cross-regional ecosystem service value is listed separately in the accounting results, and a spatial correlation map of ecosystem services is generated.

[0033] Optionally, the method further includes:

[0034] An ecosystem resilience assessment dimension is introduced to construct a three-dimensional resilience index that includes resistance, resilience, and adaptability;

[0035] By coupling the resilience index with traditional ecological value indicators, a comprehensive ecosystem value assessment model is established.

[0036] Based on the integrated ecosystem value assessment model, the output of the ecosystem gross product accounting results including resilience characteristics is used to generate an ecosystem health grading map.

[0037] Secondly, this application provides an ecosystem gross product accounting device based on GIS spatial analysis, which adopts the following technical solution:

[0038] An ecosystem gross product accounting device based on GIS spatial analysis, comprising:

[0039] The semantic mapping library construction module calculates regional basic data and satellite imagery data to construct a two-way dynamic semantic mapping library corresponding to land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes primary mapping and secondary mapping. The primary mapping associates land and ecosystem primary categories, and the secondary mapping achieves secondary category term alignment through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with annual land data.

[0040] The multi-source data fusion module acquires corresponding multi-source data, including land change survey vector data, satellite remote sensing raster data of the required accuracy, IoT real-time monitoring data, and statistical report data. The monitoring data includes soil moisture, air quality, and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed using spatial interpolation-boundary correction and sliding window smoothing methods to obtain the corresponding standardized dataset.

[0041] The accounting indicator system configuration and calculation module constructs a two-tier accounting indicator system for urban areas and typical districts. The urban area level includes three primary indicators (material products, regulatory services, and cultural services) and 14 secondary indicators. The typical district level is configured with exclusive secondary indicators for four types of characteristic areas, and these exclusive secondary indicators are enabled or disabled through dynamic switches. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model.

[0042] The GIS spatial calculation and parameter calibration module retrieves pre-trained digital elevation models and 3D building models based on GIS, extracts parameters to correct soil erosivity factor values ​​and vegetation transpiration, quantifies the indicators to preset precision raster cells, and obtains raster-level ecological value results through neighborhood analysis algorithms. It also connects IoT monitoring data to calculate the deviation rate between measured and calculated values, identifies indicators exceeding thresholds, uses random forest algorithms to calibrate the calculation parameters, controls the deviation rate within a preset range, and combines GIS to generate a multi-year ecological value trend map.

[0043] The scenario-based output and quality control module outputs customized results based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios. It is verified through a three-level quality control mechanism: data level completes missing data, calculation level intercepts abnormal parameters, and result level outputs the final result after passing the sampling verification of third-party measured data.

[0044] Thirdly, this application provides an ecosystem gross product accounting device based on GIS spatial analysis, which adopts the following technical solution:

[0045] An ecosystem gross product accounting device based on GIS spatial analysis includes a processor, wherein the processor runs a program for the ecosystem gross product accounting method based on GIS spatial analysis as described above.

[0046] Fourthly, this application provides a storage medium, which adopts the following technical solution:

[0047] A storage medium storing a program for the ecosystem gross product accounting method based on GIS spatial analysis as described in any one of the above.

[0048] In summary, this application includes at least one of the following beneficial technical effects:

[0049] Multi-dimensional optimization enhances accounting accuracy: a two-way dynamic semantic mapping library is constructed to address the semantic inconsistency between land and ecological data; spatial interpolation-boundary correction and sliding window smoothing methods are combined to achieve standardized fusion of multi-source data; ecological factors are corrected by extracting topographic and building parameters based on GIS technology; the random forest algorithm is introduced to dynamically calibrate accounting parameters, keeping the deviation rate within a preset range; and a three-level quality control mechanism ensures data integrity and result reliability, providing high-precision data support for ecological compensation, EOD projects, and other projects.

[0050] In terms of improving practicality, a two-tiered indicator system of urban areas and typical districts is adopted. By dynamically switching on and off, it adapts to the differentiated needs of characteristic areas. Combined with the spatiotemporal premium model and regional correction coefficient, it enhances the scenario adaptability of value and physical quantity calculation. Customized results are output for scenarios such as ecological compensation, EOD projects, and ecological audits. Simultaneously, multi-year ecological value trend maps and cross-regional service correlation maps are generated to intuitively reflect changes in ecological value and spatial correlations, meeting the diverse needs of accurate measurement, trend prediction, and scenario application in practice. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating an ecosystem gross product accounting method based on GIS spatial analysis, according to an exemplary embodiment.

[0052] Figure 2 This is a structural block diagram of an ecosystem gross product accounting device based on GIS spatial analysis, according to an exemplary embodiment. Detailed Implementation

[0053] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0054] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0055] This application discloses an ecosystem gross product accounting method based on GIS spatial analysis, referring to... Figure 1 ,include:

[0056] S100 uses regional basic data and satellite imagery data to construct a two-way dynamic semantic mapping library corresponding to land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes primary mapping and secondary mapping. The primary mapping associates land and ecosystem primary categories, and the secondary mapping aligns secondary category terms through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with annual land data.

[0057] The specific execution process of S100 is as follows:

[0058] Step 1: First, acquire two types of core data: Basic data of the accounting area: Focus on collecting annual land change survey data (including attribute information such as land parcel classification, ownership, and current use status) released by the natural resources department and ecosystem type survey data (such as the distribution, area, and functional level data of ecosystems such as forests, wetlands, and farmland) from the ecological environment department, to ensure that the data covers the entire accounting area and that the time dimension matches the accounting base year;

[0059] Satellite imagery data: Select satellite remote sensing images that meet the accuracy requirements of the calculation (such as Landsat series images with a resolution of 30 meters or high-resolution images with a sub-meter resolution) to help verify the spatial matching relationship between land parcels and ecosystem types. For example, the actual vegetation cover of forest land parcels can be confirmed through image interpretation to provide spatial reference for subsequent semantic mapping.

[0060] Step 2: Based on the above data, design mapping rules in two layers to form a complete semantic mapping system:

[0061] First-level mapping construction: Using "Land Change Survey First-level Category" and "Ecosystem First-level Category" as matching objects, a direct association is established. For example, "arable land" in the land first-level category is associated with "farmland ecosystem" in the ecosystem first-level category, "forest land" is associated with "forest ecosystem", and "water area and water conservancy facility land" is associated with "wetland / water ecosystem", ensuring semantic consistency between the two types of data at the highest classification level;

[0062] Second-level mapping construction: To address the semantic differences between the second-level categories of land and ecology, an ontology-based semantic matching algorithm is introduced to achieve accurate alignment. First, a "semantic ontology library" is constructed for land parcel classification and ecosystem types. Attribute features of the two data second-level categories are defined (e.g., the attribute of the land second-level category "paddy field" is "artificially planted, seasonally waterlogged," and the attribute of the ecology second-level category "artificial wetland - paddy field wetland" is "artificial intervention, possessing both production and ecological functions"). Then, the attribute similarity is calculated using an algorithm to complete semantic matching. For example, the land second-level category "paddy field" is matched to the ecology second-level category "artificial wetland - paddy field wetland," "forested land" is matched to "arbor forest ecosystem," and "urban residential land" is matched to "urban ecosystem - residential green space subsystem."

[0063] Step 3: To avoid lag in mapping relationships, establish an automatic update logic that links with annual land data:

[0064] Set "Annual Trigger Conditions": Use the annual land change survey update data released by the natural resources department as the trigger signal. When the new data is released, the system will automatically extract the classification information of newly added / changed land parcels (such as newly added ecological restoration land and adjusted urban construction land boundaries).

[0065] Automatically update mapping relationships: Based on preset semantic matching rules, newly added / changed land classification units are automatically matched to generate corresponding ecosystem type mapping results. At the same time, special cases that require manual review (such as the new land use type "photovoltaic composite land") are marked to ensure that the mapping library is updated synchronously with the land data and does not need to be manually reconstructed.

[0066] By unifying the two types of data classification systems through bidirectional semantic mapping, a foundation is laid for the integration of multi-source data. At the same time, relying on the dynamic update mechanism, the mapping relationship is synchronized with the annual land data iteration, ensuring the timeliness of the data to meet the needs of long-term dynamic accounting. Furthermore, satellite imagery is combined to verify the spatial matching between land parcels and ecosystems, improving the spatial correlation of data and providing semantic-spatial dual basis for subsequent S400 spatial calculations, ensuring the accuracy of accounting.

[0067] S200 acquires corresponding multi-source data, including land change survey vector data, satellite remote sensing raster data of appropriate accuracy, IoT real-time monitoring data, and statistical report data. The monitoring data includes soil moisture, air quality, and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed using spatial interpolation-boundary correction and sliding window smoothing methods to obtain the corresponding standardized dataset.

[0068] The specific execution process of S200 is as follows:

[0069] Step 1: First, collect four types of core data according to data type and complete preliminary preprocessing:

[0070] Land change survey vector data: The latest annual vector dataset of the accounting area is obtained from the natural resources department. It includes attribute information such as land parcel boundaries (area features), classification codes, and current land use status. During preprocessing, it is uniformly converted to the 2000 geodetic coordinate system, and duplicate or ambiguous land parcel features are removed to ensure the spatial accuracy and attribute integrity of the vector data.

[0071] Satellite remote sensing raster data: Select satellite images (such as Landsat series and Gaofen series images) that meet the calculation accuracy (usually 30 meters to sub-meter level), extract raster features such as vegetation index (NDVI) and land surface temperature, and perform radiometric and geometric corrections to keep the image coordinate system consistent with the vector data.

[0072] IoT real-time monitoring data: Three types of data are collected through a sensor network deployed within the area—soil moisture data (including volumetric water content and temperature, with a sampling frequency of 1 time / hour), air quality data (… , Indicators such as (1 time / hour) and negative oxygen ion concentration data (1 time / 10 minutes) were collected. During preprocessing, the moving average method was used to remove abnormal jump values, and the data were aggregated on an hourly basis to generate standardized time-series data.

[0073] Statistical report data: Statistical data for the base year are obtained from departments such as agriculture, environmental protection, and culture and tourism, including crop yield, forest stock volume, and tourist volume. During preprocessing, the statistical scope and accounting area of ​​the data are checked, and the municipal data is broken down into district, county or township units to ensure spatial scale matching.

[0074] Step 2: Based on the preprocessed data, solve the spatiotemporal heterogeneity problem of multi-source data through a dual algorithm:

[0075] Spatial interpolation-boundary correction fusion: This method addresses the spatial accuracy differences between raster and vector data by performing a two-step matching process.

[0076] Spatial interpolation: Kriging interpolation is used to encrypt low-precision raster data (such as air quality raster), improving the resolution of 3km data to 30m resolution, so that the raster cell size is adapted to the vector plot scale.

[0077] Boundary correction: Using the land change survey vector plots as boundary templates, control point pairs (such as road intersections and building corners) are created using ArcGIS spatial correction tools. The raster data and vector boundaries are aligned through an affine transformation model to eliminate raster outliers that cross plots and avoid spatial conflicts where "one raster unit corresponds to multiple plots".

[0078] Sliding window smoothing time-series fusion: For data with different temporal resolutions (e.g., monthly updates of remote sensing data, hourly updates of IoT data), a dynamic window is used for smoothing.

[0079] Window settings: Match window size to data type (set 30-day window for remote sensing data, set 24-hour window for IoT data);

[0080] Smoothing calculation: The weighted moving average method is used to fuse time series data. For example, daily IoT soil moisture data and monthly remote sensing vegetation index data are fused to generate continuous daily soil moisture raster data, eliminating drastic fluctuations in the time dimension.

[0081] Step 3: The merged data needs to undergo three unified processing steps to form the final dataset:

[0082] Format standardization: Vector data is converted to SHP format, raster data is converted to TIFF format, time series data is stored in CSV format, and all data is associated with a unique spatial code (such as plot ID, raster cell coordinates).

[0083] Unit normalization: This involves normalizing soil moisture (%) and negative oxygen ion concentration (cFU / cm³). Different dimensional indicators such as ) are converted into standardized values ​​in the range [0,1], and the dimensional differences are eliminated by the extreme value method;

[0084] Quality verification: Data integrity algorithms are used to screen for missing values. Missing areas in raster data are filled by interpolation based on neighboring cells, and missing attributes in vector data are filled by correlation with statistical report data, ensuring that the dataset integrity reaches more than 99%.

[0085] By achieving spatial alignment between raster and vector data through spatial interpolation and boundary correction, and fusing multi-temporal resolution data using the sliding window smoothing method, the problem of "spatiotemporal heterogeneity" of multi-source data is solved, addressing the "data silo" issue in traditional GEP accounting. Simultaneously, by constructing a standardized data foundation through preprocessing noise reduction, boundary alignment, and unit normalization, data errors are reduced from the source, providing accurate input for S300 index quantification and S400 spatialization calculations. Furthermore, leveraging a standardized dataset containing spatial, temporal, and statistical attributes, spatial units of the S100 semantic mapping library are accurately matched, achieving seamless integration with subsequent accounting stages and ensuring the overall process's continuity and consistency.

[0086] S300 constructs a two-tiered accounting indicator system for urban areas and typical districts. The urban area level includes three primary indicators: material products, regulatory services, and cultural services, as well as 14 secondary indicators. The typical district level consists of four types of distinctive areas with dedicated secondary indicators, which are enabled or disabled through dynamic switching. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model.

[0087] The specific execution process of S300 is as follows:

[0088] Step 1: Based on the characteristics of the accounting area and practical needs, design the indicator system in two layers to ensure coverage of the entire region and differences in specific areas:

[0089] Urban area hierarchical indicator design:

[0090] Based on the principle of "full coverage and universal adaptability", the specific connotations and calculation scope of the 3 primary indicators and 14 secondary indicators are clarified:

[0091] Material product indicators (4 secondary indicators): including agricultural products (output of grain, cash crops, etc. within the accounting area), forestry products (timber, bamboo, non-timber forest products, etc.), freshwater resources (total water use for domestic, production, and ecological purposes), and ecological energy (available biomass energy, hydropower, etc.).

[0092] Regulation service indicators (7 secondary indicators): including water conservation (rainfall interception, groundwater replenishment), soil conservation (reduction of soil erosion), carbon sequestration (carbon sequestration by vegetation and soil), oxygen release (oxygen release by vegetation photosynthesis), and air purification (absorption of oxygen). , Pollutant levels, water purification (degradation of pollutants such as COD, nitrogen, and phosphorus), and urban heat island mitigation (the regulation of surface temperature by vegetation and water bodies).

[0093] Cultural service indicators (3 secondary indicators): including landscape recreation (number of visitors to scenic spots, parks, etc.), nature education (number of participants in nature education bases), and health and wellness (number of people served by health and wellness venues);

[0094] All secondary indicators clearly define the data sources (such as agricultural products corresponding to agricultural department statistics, and air purification corresponding to environmental protection department monitoring data) and calculation dimensions (both physical quantity and value).

[0095] Typical area-level specific indicator design and dynamic switch configuration:

[0096] For four types of distinctive areas—science and technology parks, cultural and tourism towns, rural areas, and suburban parks—dedicated secondary indicators are configured, and their activation can be flexibly achieved through system switches.

[0097] Exclusive indicators for science and technology parks: carbon sequestration value of green buildings (carbon sequestration amount of photovoltaic power in green buildings within the park) and rainwater recycling value (utilization of rainwater recycling system in the park).

[0098] Exclusive indicators for cultural and tourism ancient towns: cultural heritage ecological premium (the added value of the ecological landscape in the core area of ​​the ancient town to the cultural heritage) and tourist carrying capacity regulation value (the benefits of tourist volume regulation based on ecological capacity).

[0099] Rural-specific indicators: farmland ecological cycle value (benefits of circular agriculture such as straw return to the field and agricultural film recycling), and courtyard economy ecological product value (quantity of products such as fruits and vegetables and livestock and poultry breeding in farmers' courtyards).

[0100] Country park-specific indicator: Ecological corridor connectivity value (the contribution of the park to the connectivity between the park and surrounding ecological patches);

[0101] The dynamic switch implementation logic is as follows: The system presets the mapping relationship between "region type - exclusive indicator". After the user selects the region type for calculation (such as "cultural and tourism ancient town"), the system automatically enables the corresponding exclusive indicator. If no region is selected, the indicator is blocked to avoid invalid calculation.

[0102] Step 2, with the goal of "accurately reflecting regional ecological differences," involves two steps to complete the physical quantity calculation:

[0103] Basic physical quantity calculation:

[0104] For each secondary indicator, the calculation basis is determined based on the standardized dataset generated by S200:

[0105] Material products: For example, the physical quantity of agricultural products = crop sown area in S200 statistical report data × yield per unit area; the physical quantity of forestry products = forest stock volume inverted from S200 remote sensing data × timber yield.

[0106] Regulation services: such as water conservation physical quantity = precipitation in S200 meteorological data - surface runoff - evapotranspiration (calculated based on digital elevation model); soil retention physical quantity = potential soil erosion - actual soil erosion (based on RUSLE model, parameters taken from S200 soil and vegetation data).

[0107] Cultural services category: such as the physical volume of landscape recreation = annual number of visitors to scenic spots in the S200 statistical report + daily reception volume of parks;

[0108] Introduction and application of regional correction coefficients:

[0109] To eliminate the impact of regional differences in natural conditions on physical quantities, correction coefficients are determined and applied in three steps:

[0110] Coefficient Classification: Based on the climate (north / south), topography (plains / mountains / hills), and ecosystem type (forests / farmland / wetlands) of the accounting area, the correction coefficient range is divided with reference to the "Technical Specifications for National Ecological Status Survey and Assessment". For example, the correction coefficient for soil conservation in the north is 1.2 (high risk of soil erosion) and 1.0 in the south; the correction coefficient for water conservation in mountains is 1.3 (strong precipitation interception capacity) and 0.9 in plains.

[0111] Coefficient matching: Associate the topographic and climate data of the accounting area in S200 with the correction coefficient range and automatically match the corresponding coefficient (e.g., the water quality purification correction coefficient for mountainous wetlands in the south is 1.1).

[0112] Corrected calculation: Final physical quantity = basic physical quantity × regional correction coefficient. For example, the actual grain yield of a certain mountain farmland in the south = (sown area × basic yield per unit area) × 1.1 (mountain correction coefficient).

[0113] Step 3, with the goal of "aligning with actual value fluctuations," calculates the value by considering differences in time and space scenarios:

[0114] Calculation of basic value:

[0115] Determine the underlying value based on physical quantity and market / substitution costs:

[0116] For material products: the market value method is used, such as the basic value of agricultural products = physical quantity of agricultural products × local average market price (taken from agricultural product price data in the S200 statistical report).

[0117] Adjustment services: The substitution cost method is adopted, such as the basic value of soil conservation = actual quantity of soil conservation × unit cost of dredging (taken from engineering quota data of the water conservancy department).

[0118] Cultural services: The travel cost method is adopted, such as the basic value of landscape recreation = number of tourists × average travel expenditure per person (taken from tourist consumption data statistics of S200 cultural and tourism departments).

[0119] Introduction and application of the spatiotemporal premium model:

[0120] To address the fluctuations in value over time and in hydrological contexts, a two-dimensional premium model is constructed and applied.

[0121] Time-based premium: The time premium coefficient is determined based on the season (peak season / off-season) and holidays (statutory holidays / ordinary days). For example, the tourism value of cultural and tourism ancient towns: the premium coefficient is 1.5 during holidays (a surge in tourist volume drives up consumption), and 0.8 during the off-season; the value of natural education: the premium coefficient is 1.3 during winter and summer vacations (due to concentrated study tours).

[0122] Hydrological premium: The hydrological premium coefficient is determined based on the hydrological period (flood season / non-flood season) and precipitation intensity (heavy rain / sparse rain). For example, the water conservation value: the flood season premium coefficient is 1.6 (high demand for flood storage and increased cost of alternative projects), and the non-flood season coefficient is 1.0; the water purification value: the heavy rain season premium coefficient is 1.2 (large amount of pollutants washed away, high demand for purification).

[0123] Premium calculation: Final value = Basic value × Spatiotemporal premium coefficient (time coefficient and hydrological coefficient are superimposed, such as the cultural service value during holidays + flood season = basic value × 1.5 × 1.2).

[0124] By employing a two-tiered design of "general indicators for urban areas + specific indicators for typical areas" with dynamic switching, the system addresses the poor adaptability of traditional GEP accounting indicator systems. This approach covers the ecological value of the entire region while accurately capturing the core ecological contributions of unique areas. The introduction of regional correction coefficients quantifies differences in natural conditions, improving the regional accuracy of physical quantity calculations and reducing indicator quantification errors. The use of a spatiotemporal premium model dynamically adjusts value quantities through time and hydrological dual-dimensional coefficients, enhancing scenario adaptability and ensuring that accounting results align with the true value of ecological products, providing a reference for ecological compensation, EOD projects, and other practical applications. Furthermore, the system clarifies the calculation logic of each indicator and links it to the S200 standardized dataset, building a bridge between "indicators, data, and calculations." This provides a clear "indicator-value" correspondence for S400 spatial accounting, ensuring the continuity and accuracy of subsequent processes.

[0125] S400, based on GIS, retrieves pre-trained digital elevation models and 3D building models, extracts parameters to correct soil erosivity factor values ​​and vegetation transpiration, quantifies the indicators to preset precision raster units, and obtains raster-level ecological value results through neighborhood analysis algorithms; it also accesses IoT monitoring data to calculate the deviation rate between measured and calculated values, identifies indicators exceeding thresholds, uses random forest algorithms to calibrate the calculation parameters, controls the deviation rate within a preset range, and combines GIS to generate a multi-year ecological value trend map.

[0126] The specific execution process of S400 is as follows:

[0127] Step 1: Load the pre-set model based on GIS technology, extract key parameters and correct core ecological factors to ensure that the factor values ​​match the actual geographical characteristics of the region.

[0128] Model retrieval and preprocessing:

[0129] Two types of core models that have been pre-trained are retrieved from the GIS spatial database: a digital elevation model (reflecting the topographic relief characteristics of the region) and a 3D building model (reflecting the distribution and height of buildings in urban areas). Coordinate matching (unified to the 2000 geodetic coordinate system) and accuracy verification (vertical accuracy error of the digital elevation model ≤ 2 meters, and planar position error of the 3D building model ≤ 1 meter) are completed to ensure that the model data and the spatial range of the calculation area are fully covered.

[0130] Key parameter extraction:

[0131] Topographic parameters—slope (reflecting the degree of surface inclination) and aspect (reflecting the orientation of the slope)—are extracted from the digital elevation model. These parameters are then calculated using the GIS "slope and aspect analysis tool" to generate slope raster maps (unit: degrees) and aspect raster maps (unit: degrees, range 0-360 degrees). Building parameters—building density (the ratio of building footprint area to the total area of ​​a given area)—are extracted from the 3D building model. The proportion of building footprints for each plot unit is statistically analyzed using "spatial overlay analysis" to generate a building density raster map (unit: %).

[0132] Ecological factor correction:

[0133] The core ecological factors are dynamically adjusted based on the extracted parameters:

[0134] Soil erosivity factor value correction: Combining slope parameters and following the rule that "the greater the slope, the stronger the soil erosivity", a correlation model between slope and erosivity factor is established (e.g., when the slope is <5 degrees, the erosivity factor takes a base value of 1.0; when the slope is 5-15 degrees, it takes 1.2-1.8; and when the slope is >15 degrees, it takes 1.8-2.5). At the same time, the slope aspect is taken into account (sunny slopes have strong sunlight and fast water evaporation, so the erosivity factor is 10%-15% higher than that of shady slopes) for fine-tuning, and finally a regionalized soil erosivity factor raster is generated.

[0135] Vegetation transpiration correction: Combining building density parameters and considering the impact of building shading on vegetation light and ventilation, a correction relationship between building density and transpiration is established (e.g., when building density is <20%, the transpiration is taken as the base value of 1.0; when it is 20% to 50%, it is taken as 0.7-1.0; and when it is >50%, it is taken as 0.4-0.7). At the same time, the vegetation cover data of S200 is superimposed (higher vegetation cover means higher transpiration) to generate a corrected vegetation transpiration raster.

[0136] Step 2: Associate the index values ​​calculated by S300 with spatial cells, and refine the results through raster quantization and neighborhood analysis:

[0137] Grid cell matching and index quantification:

[0138] Set a grid unit with a preset accuracy (e.g., 30m x 30m, adapted to the accuracy of satellite remote sensing data). Using the grid as the basic spatial unit, the physical quantity / value index value of S300 (e.g., water conservation capacity and carbon sequestration value of a plot) is allocated to the corresponding grid through "spatial correlation analysis". If the plot spans multiple grids, the index value is allocated according to the proportion of the overlapping area between the grid and the plot, ensuring that each grid unit has a unique index quantification result.

[0139] Neighborhood analysis algorithm processing:

[0140] Neighborhood analysis is performed using a "3×3 neighborhood window" to spatially smooth the index values ​​of each grid cell: the average index value of the current grid cell and its eight neighboring grid cells is calculated. If the deviation between the current grid cell value and the average value exceeds a preset threshold (e.g., 20%), the average value is used to replace the current value. This eliminates outliers at grid edges caused by plot boundary division and data sampling errors, ultimately generating continuous and smooth grid-level ecological value results (e.g., water conservation value grid map, carbon sequestration grid map).

[0141] Step 3: Compare the results with real-time monitoring data, and optimize the calculation parameters through algorithm iteration to control the error within a reasonable range.

[0142] Monitoring data access and deviation rate calculation:

[0143] The real-time IoT monitoring data in S200 (such as measured soil moisture and measured negative oxygen ion concentration) are associated with the corresponding grid calculation values ​​(soil moisture calculation values ​​and negative oxygen ion concentration calculation values ​​based on S300-S400). The deviation rate of each monitoring point corresponding to the indicator is calculated according to the formula "deviation rate = |measured value - calculated value| / measured value × 100%", and a deviation rate statistics table is generated.

[0144] Identification of out-of-threshold indicators and preparation of calibration data:

[0145] Set a preset threshold for the deviation rate (e.g., 5%, which can be adjusted according to the accuracy requirements of the calculation), filter out indicators whose deviation rate exceeds the threshold (e.g., the soil moisture deviation rate in a certain area reaches 8%), and extract the historical calculation data (measured values, calculated values, and deviation rates of the past 3 to 5 years) and influencing parameters (e.g., soil type, vegetation cover, and climate parameters) of these indicators to form the training dataset for the random forest algorithm.

[0146] Random Forest Algorithm Parameter Calibration: Using "historical influence parameters" as feature variables and "deviation rate control target" (e.g., deviation rate ≤ 5%) as target variable, train the random forest model—learn the correlation between parameters and deviation rate through the model (e.g., the higher the soil clay content, the lower the soil moisture calculation parameters need to be by 5% to 10%), automatically adjust the calculation parameters of indicators exceeding the threshold (e.g., soil water holding capacity parameters in soil moisture calculation, vegetation release coefficient in negative oxygen ion concentration calculation), and recalculate the indicator values ​​after calibration until the deviation rate drops to within the preset threshold.

[0147] Step 4: Combining multi-period accounting data, present the spatiotemporal changes in ecological value using GIS visualization technology:

[0148] Multi-period data integration: Collect raster-level ecological value results of the accounting area for several years (e.g., 5 years) (1 period per year), and complete the alignment of time series data in GIS (ensuring consistent raster cell boundaries and accuracy each year) and attribute association (adding an "annual-value" attribute column to each raster).

[0149] Trend Analysis and Chart Creation:

[0150] Using GIS's "time series analysis tools," the annual trend of ecological value changes for each raster (such as annual growth rate and cumulative change) is calculated, and two types of maps are generated through visualization technology:

[0151] Spatial distribution trend map: A heat map is used to show the differences in the ecological value of different regions (e.g., green indicates growth and red indicates decline).

[0152] Time trend line chart: Generate annual value change line charts for each primary indicator (material products, regulatory services, and cultural services), mark key change nodes (such as a 15% increase in the value of regulatory services due to ecological restoration in a certain year), and finally form a complete multi-year ecological value trend map.

[0153] By employing raster-level quantification, terrain and building parameter correction, and neighborhood analysis, this approach overcomes the bottleneck of coarse spatialization in traditional GEP accounting, achieving precise spatial positioning of ecological value. This provides a spatial basis for precise ecological compensation down to the plot level and for optimizing EOD project site selection. Using IoT data as a real-world benchmark, a closed-loop logic of "real-time monitoring - deviation identification - random forest algorithm calibration" solves the problem of static parameters, dynamically adjusting accounting parameters to control the deviation rate within a preset range, thus improving the accuracy and reliability of the accounting results. Combined with GIS, a multi-year ecological value trend map is generated, intuitively displaying the spatiotemporal changes in ecological value, quantifying the temporal dimension patterns, and providing managers with "spatiotemporal dual-dimensional" decision support, assisting in judging the effectiveness of ecological restoration and long-term ecological management. Simultaneously, it builds upon the quantitative results of S300 indicators and transforms them into spatial results, while following up with the scenario-based output of S500 to provide basic data. Furthermore, it optimizes the S300 indicator calculations through calibration parameters, forming a complete link of "indicator quantification - spatialization - calibration - optimization," strengthening the coherence and logical closed loop of the overall accounting process and avoiding the accumulation of errors caused by disconnections in various links.

[0154] S500 outputs customized results based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios. It is verified through a three-level quality control mechanism: data level completes missing data, calculation level intercepts abnormal parameters, and result level outputs the final result after passing the sampling verification of third-party measured data.

[0155] The specific execution process of S500 is as follows:

[0156] Step 1: Based on the grid-level ecological value results and multi-year trend maps of S400, and combined with the practical needs of different scenarios, generate targeted outputs to ensure that the results are "ready to use immediately":

[0157] Output of Ecological Compensation Scenarios: With the goal of "precisely matching compensation amounts," the project integrates S400 plot-level ecological value data (such as the total regulatory service value and cultural service value of each plot) with local ecological compensation standards (such as the compensation value corresponding to 10,000 yuan of ecological value). Through a "value-value conversion model," the project calculates the compensation amount for each plot and generates a "plot-level ecological compensation amount map" (marking compensation amount ranges with different colors, such as red indicating compensation ≥ 500,000 yuan / plot and blue indicating 100,000-500,000 yuan / plot). At the same time, it outputs a compensation priority list (sorted by plot ecological value from high to low, marking high-value key compensation plots, such as forest land and wetlands in the top 10% of ecological value), supporting the natural resources department in formulating compensation implementation recommendations.

[0158] EOD Project Scenario Outputs: With the goal of "project revenue prediction and site selection optimization", based on the multi-year ecological value trend map of S400 (such as the annual growth rate of ecological value in the next 5 years), and superimposed on the EOD project planning scope (such as park boundaries and restoration project areas), an "EOD project ecological value increment prediction model" is constructed, which outputs the ecological value increment curve within the project cycle (such as 10 years) (marking the annual increment value and cumulative increment); at the same time, it is linked with project investment data (such as restoration costs and operating costs) to generate an "ecological value-cost-benefit analysis report", which clarifies the project's "ecological value break-even point" (such as the ecological value increment in the 3rd year covering the annual operating costs), providing investors with a basis for predicting returns.

[0159] Ecological audit scenario output: With the goal of "compliance verification and error traceability", the report summarizes the full-process data of S200-S400 (such as raw monitoring data, parameter calibration records, and spatialized calculation logs), sorts out the compliance points according to the "Technical Specification for Accounting of Gross Ecosystem Product", and outputs "Ecological Audit Compliance Report", which includes three core contents: compliance of data sources, compliance of parameter calculation (such as whether the regional correction coefficient and spatiotemporal premium coefficient have a basis), and result deviation rate (marking the final deviation rate after three levels of quality control), and attaches third-party verification documents to meet the audit department's compliance verification needs for the accounting process.

[0160] Step 2: Control errors throughout the entire process from "data source to calculation process to result output" to ensure the reliability of the calculation results.

[0161] Data-level missing data identification and completion:

[0162] 1. Missing data identification: The "data integrity scanning algorithm" traverses the S200 standardized dataset (such as raster data and statistical report data) and the intermediate data for S300 index calculation (such as draft values ​​of physical quantities) to mark missing items (such as the negative oxygen ion concentration data of a certain area being empty or the building density parameter of a certain plot being missing).

[0163] 2. Precise Completion: For missing spatial data (such as gaps in raster data), the "3×3 neighborhood interpolation method" is used (the mean of the surrounding 8 rasters is used for completion); for missing attribute data (such as missing statistical yield of a certain plot), the S100 semantic mapping library is associated to match the mean of similar plots (such as cultivated land with the same soil type and crop) for completion; after completion, data consistency needs to be verified (such as the completed yield value needs to be within ±10% of the yield of similar plots) to ensure data integrity ≥99%.

[0164] Computational level, abnormal parameter identification and interception:

[0165] 1. Threshold preset: Based on industry standards and historical accounting data, set reasonable ranges for core accounting parameters (such as soil erosivity factor value 1.0-2.5, vegetation transpiration correction coefficient 0.4-1.0, and spatiotemporal premium coefficient 0.8-1.6).

[0166] 2. Real-time interception: The system monitors the parameter values ​​during the calculation of S300 indicators (such as regional correction of physical quantities) and the calibration of S400 parameters (such as output parameters of random forest algorithm) in real time. If the values ​​exceed the reasonable range (such as soil erosivity factor value = 3.0), the interception mechanism is automatically triggered, and a pop-up window is displayed to indicate the abnormal parameters and the extent of the exceedance. The system also retrieves the parameter records of the last three similar areas and provides suggested correction values. The intercepted abnormal parameters need to be manually reviewed (such as to confirm whether they are caused by special terrain). After correction, the system re-enters the calculation process to prevent abnormal values ​​from being transmitted to the final result.

[0167] Result-level verification using third-party measured data sampling:

[0168] 1. Sample selection: The third-party organization selects verification samples from the grid-level ecological value results of S400 using the "stratified sampling method"—covering four types of ecosystems: forest, farmland, wetland, and urban areas. At least 8 sample points are selected for each type (total sample size ≥ 30). At the same time, field measurement data are collected from the sample points (such as measuring vegetation carbon sequestration through quadrat surveys and soil retention through runoff plots).

[0169] 2. Verification and Judgment: Calculate the "deviation rate of measured value - calculated value" for each sample point. If the proportion of samples with a deviation rate ≤ 5% is ≥ 90%, the judgment result is qualified. If it is unqualified, backtrack to the S400 parameter calibration step (such as readjusting the training dataset of the random forest algorithm), optimize, recalculate and verify, until the qualified standard is met.

[0170] Customized maps, models, and reports are developed for three scenarios: ecological compensation, EOD projects, and ecological audits. These are linked to S400 spatialization and trend data, transforming accounting results from "theoretical values" into actionable decision-making tools, thus addressing the disconnect between traditional GEP accounting results and practical applications. A three-tiered quality control system—data-level completion of missing data, computation-level interception of abnormal parameters, and result-level third-party verification—controls the accounting deviation rate to within 5%, eliminating the accumulation of errors across multiple stages and ensuring result accuracy. Furthermore, building upon S400 data as the foundation for output, quality control optimizes weak links between S200 and S400, forming a complete closed loop of "data input → indicator calculation → spatialization → scenario output → quality control." This ensures seamless process flow, prevents error accumulation, and enhances the stability and reliability of the overall accounting system.

[0171] Taking the practical needs of ecological compensation and EOD projects in a mountainous city in southern China (covering forest land, wetlands, cultural and tourism towns, and science and technology parks) as an example, the solution ensures the accuracy and practicality of calculations through a full-process technical design. In the ecological compensation scenario, S100 first constructs a two-way semantic mapping library between land "forest land" and "wetland" plots and ecological "forest ecosystem" and "wetland ecosystem," and verifies the spatial matching of mountain plots by combining satellite imagery to avoid data semantic bias; S200 integrates land vector data, 30-meter resolution high-resolution remote sensing data, and IoT soil moisture monitoring data, improves the accuracy of air quality raster data to 30 meters through Kriging interpolation, and then aligns the boundaries of mountain plots through boundary correction to generate a standardized dataset; S300, tailored to the terrain of southern mountains, configures a regional correction coefficient of 1.3 for soil conservation and water conservation indicators to accurately calculate the soil of mountain forest land. Maintaining both physical quantity and value; the S400 retrieves the digital elevation model to extract the mountain slope (15-25 degrees), corrects the soil erosivity factor to 1.8-2.5, and then generates raster-level ecological value results through 3×3 neighborhood analysis. Finally, the S500 outputs a "plot-level ecological compensation amount map" (e.g., core mountain forest land compensation ≥ 500,000 yuan / plot, peripheral forest land 100,000-500,000 yuan / plot) and a priority list. At the same time, through three-level quality control, the negative oxygen ion concentration data of some remote mountain areas is supplemented, abnormal soil erosivity factor values ​​(e.g., miscalculated 3.0) are intercepted, and the deviation rate of third-party sampling verification is ≤5%, ensuring that the compensation amount is accurately allocated to high-value ecological plots.

[0172] In this urban mountain ecological restoration EOD project, the solution meets the needs of project benefit prediction and site selection through scenario-based output. Based on nearly 5 years of raster-level ecological value data, S400 generates an ecological value trend map of the mountain restoration area, showing that the annual growth rate of water conservation value after restoration reaches 8%. S300 introduces a hydrological premium model, setting the water conservation value premium coefficient to 1.6 for the southern flood season (May-September), accurately calculating the ecological value increment during the flood season. Based on the above data, S500 constructs an EOD project ecological value increment prediction model, outputting the cumulative ecological value increment within a 10-year project cycle, while also linking restoration costs to generate a cost-benefit analysis report, clarifying that the ecological value increment in the third year of the project can cover the annual operating costs, providing investors with a clear return point. In addition, the solution also corrects the vegetation transpiration around the science and technology park through the building density parameter of S400, avoiding the impact of building shading on the ecological value calculation, helping the project site selection avoid high building density areas, and prioritizing mountain restoration plots with great potential for ecological value increment.

[0173] In this case, the solution ensures both "accuracy" and "practicality" through multi-stage technical design: In terms of accuracy, the S400 uses a random forest algorithm to calibrate the water quality purification calculation parameters of the mountain wetland, reducing the deviation rate between the measured and calculated values ​​from the initial 8% to 4%. The S500 further verifies reliability through third-party verification at the result level (32 sample points were extracted, and 29 of them had a deviation rate ≤ 5%). In terms of practicality, in response to the unique needs of the cultural and tourism ancient town, the S300 enables the "cultural heritage ecological premium" indicator through a dynamic switch to calculate the added value contribution of the surrounding mountain landscape to the cultural heritage. The S500 simultaneously outputs an ecological audit compliance report, clarifying the data source (such as the number of tourists in the ancient town being taken from the statistical report of the cultural and tourism department) and the basis for parameter calculation, meeting the project audit requirements, and ultimately achieving a deep fit between the technical solution and the practical needs of ecological compensation and EOD projects.

[0174] In this embodiment of the application, in order to eliminate the impact of annual ecological changes on the accounting results and achieve comparability of cross-year GEP data, it is achieved through three steps: "historical ecological data collection and preprocessing - construction of ecological value decay coefficient model - time-series correction of current accounting results". The specific process is as follows:

[0175] Step 1: First, acquire three types of core historical data and complete standardized preprocessing to ensure data continuity and availability.

[0176] Data collection scope and sources:

[0177] Historical ecological data of the accounting area over the past 10 to 20 years (the time span must cover at least two ecological change cycles, such as vegetation growth cycles and land use planning cycles), with clearly defined data sources:

[0178] Historical climate change data: obtained from meteorological departments, including indicators such as annual average temperature, annual precipitation, and number of extreme precipitation days (e.g., daily average temperature and monthly average precipitation datasets from 2010 to 2023); Vegetation cover change data: obtained from natural resources departments or remote sensing data platforms, including historical vegetation cover (NDVI) and vegetation type transfer data (e.g., area of ​​forest land converted to cultivated land and area of ​​wetland vegetation degradation from 2015 to 2020), in the form of annual remote sensing raster imagery (30-meter resolution); Land use transformation data: obtained from historical land use change survey data from natural resources departments, extracting land use type transfer records (e.g., the area and coordinates of the transformation from "cultivated land to construction land" and "unused land to forest land" each year from 2010 to 2023).

[0179] Data preprocessing:

[0180] Time alignment: Unify all historical data to an "annual" time granularity (e.g., aggregate monthly climate data into annual averages, and take the annual maximum value for quarterly vegetation cover data) to ensure consistency in time series dimensions; Spatial matching: Unify the coordinate system of historical vegetation raster data and land use data to the 2000 geodetic coordinate system to align with the spatial range of the current accounting area (e.g., remove historical data outside the accounting area using pruning tools); Outlier handling: Use the "3σ principle" to screen outliers in historical data (e.g., annual precipitation is more than 3 times the average of the past 5 years), and replace outliers with the moving average of the indicator over the past 3 years to ensure data continuity.

[0181] Step 2: Based on the preprocessed historical ecological data and current accounting data, a model is constructed through "time series comparison - factor correlation - coefficient calculation" to quantify the degree of value decay of different ecosystem types.

[0182] Model input and core alignment dimensions:

[0183] The model uses "historical ecological data" and "current accounting data" as dual inputs, and selects three key comparison dimensions for four core ecosystems: forests, wetlands, farmland, and urban areas.

[0184] Functional decline dimension: such as "annual vegetation cover change" for forest ecosystems (NDVI=0.8 in 2010, NDVI=0.6 in 2023, indicating vegetation cover decline), and "annual water area change" for wetland ecosystems (wetland area 100km² in 2015, 85km² in 2023, indicating wetland shrinkage); Environmental driving dimension: such as "annual precipitation change" for farmland ecosystems (average annual precipitation decreased by 15% from 2018 to 2023, indicating insufficient water supply leading to productivity decline), and "annual building density change" for urban ecosystems (building density 20% in 2015, 45% in 2023, indicating green space encroachment leading to regulatory function decline); Land use transformation dimension: such as "impact of land use transformation on ecological functions over the years" (a certain area was converted from forest land to construction land in 2020, indicating complete decline in carbon sequestration and oxygen release functions).

[0185] Attenuation coefficient calculation logic:

[0186] Single-factor decay calculation: For each comparison dimension, the decay is calculated as follows: "Single-factor decay = (Base year data value - Current year data value) / Base year data value" (the base year is selected as the year with the best ecological function in historical data, such as 2010 when forest NDVI was the highest). For example, for forest ecosystems, "Vegetation cover decay = (2010 NDVI - 2023 NDVI) / 2010 NDVI = (0.8 - 0.6) / 0.8 = 0.25";

[0187] Weighted calculation of the overall degradation coefficient: Weights are assigned to the three comparison dimensions (functional decline dimension: 0.5; environmental driving dimension: 0.3; utilization transformation dimension: 0.2). For example, for a forest ecosystem in a certain region: Overall degradation coefficient = (0.25 × 0.5) + (0.1 × 0.3, degradation due to decreased precipitation) + (0 × 0.2, no land transformation) = 0.125 + 0.03 = 0.155, meaning the value of the forest ecosystem in this region has decreased by 15.5%.

[0188] Coefficient range calibration: Referring to the "Ecosystem Service Value Assessment Specification", the comprehensive attenuation coefficient is limited to the range of [0,1] (0 represents no attenuation, 1 represents complete attenuation). If the calculation result is negative (such as the vegetation coverage of a certain area increases), then 0 is taken, which means that no attenuation correction is required.

[0189] 3. Time-series correction of current accounting results based on attenuation coefficient:

[0190] The calculated ecological value decay coefficient is applied to the current GEP accounting results to eliminate the impact of historical decay and generate comparable data across years:

[0191] Correction logic and formula: For the value of different ecosystem types in the current accounting results (e.g., the current accounting carbon sequestration value of forest ecosystem = 50 million yuan), the correction is performed according to "time-series corrected value = current accounting value / (1 - comprehensive attenuation coefficient)" (the denominator "1 - comprehensive attenuation coefficient" represents the "value retention ratio" of the current ecosystem).

[0192] Example: The comprehensive decay coefficient of a forest ecosystem in a certain area is 0.155, and the current calculated carbon sequestration value is 50 million yuan. Then the corrected value is 50 million / (1 - 0.155) ≈ 59.17 million yuan. This value is equivalent to "the carbon sequestration value comparable to the ecological function level in 2010 after eliminating the decay impact from 2010 to 2023".

[0193] Generation of comparable results across years: After correcting all primary indicators (material products, regulatory services, and cultural services) in the current accounting according to the above logic, a "Comparable GEP Accounting Report Across Years" is generated. The report should simultaneously present the "Current Original Accounting Value", "Corrected Comparable Value", and "Comprehensive Degradation Coefficient", and indicate the base year (e.g., "Comparable GEP with 2010 as the base year") to facilitate users' intuitive comparison of the ecological value change trends in different years (e.g., 2010, 2020, and 2023), avoiding misjudgments such as "decreased accounting value but effective actual ecological protection" due to annual ecological degradation.

[0194] In this embodiment of the application, the method further includes:

[0195] Step 1: Through multi-source data fusion and spatial analysis, accurately locate the scope of sensitive areas and determine the protection level, laying the foundation for subsequent weight configuration.

[0196] Three types of authoritative data are integrated to ensure comprehensive identification: boundary delineation data, obtained from the natural resources department, including "Ecological Protection Red Line Delineation Results," "Lists of Nature Reserves at All Levels" (including approval documents and vector boundaries of national, provincial, and municipal nature reserves), and "Wetland Protection List" (clearly defining the core and buffer zones of wetlands); biodiversity data, obtained from the forestry department, including "Distribution Maps of Endangered Species Habitats," containing species names, protection levels, and critical habitat ranges (such as foraging areas and breeding areas); and remote sensing-assisted data, using high-resolution satellite imagery (sub-meter level) for interpretation and verification, and using areas with abnormal vegetation cover (such as concentrated areas of primary forests) and the integrity of surface water bodies (such as undisturbed natural wetlands) to help confirm the actual boundaries of sensitive areas and exclude areas encroached upon by human development.

[0197] Spatial calibration: Unify the above data to the 2000 geodetic coordinate system, compare the administrative boundaries with the remote sensing interpretation boundaries using the GIS "spatial overlay analysis" tool, and correct any deviation areas (such as boundary offsets caused by human development) based on the latest remote sensing imagery to ensure that the boundary accuracy error is ≤3 meters;

[0198] Protection level classification: Classified according to the principle of "national level > provincial level > municipal level" and "core area > buffer zone". For example, the core area of ​​a national nature reserve is level I, the core area of ​​a provincial wetland is level II, and the habitat of endangered species in a municipality is level III. Each level is assigned a unique code (such as I-001, II-002) to facilitate subsequent weight matching.

[0199] Step 2: Based on the protection level of sensitive areas and the scarcity of ecosystem services, construct a two-dimensional weighting system to achieve differentiated configuration where "the higher the protection level and the scarcer the function, the greater the weight":

[0200] Weighting factors: Determination of dimensions and quantification standards:

[0201] Protection level weight (60%):

[0202] Level I sensitive areas (national core areas) have a base weight of 1.6 (reflecting the highest protection priority); Level II sensitive areas (provincial core areas / national buffer zones) have a base weight of 1.4; Level III sensitive areas (municipal sensitive areas / provincial buffer zones) have a base weight of 1.2.

[0203] Ecosystem service scarcity weight (40%):

[0204] According to the "Guidelines for Assessing Ecosystem Service Scarcity," regions that provide irreplaceable services receive bonus points:

[0205] The only habitat for endangered species, add 0.4 (irreplaceable function); the core wetland area (the only water conservation core in the region), add 0.3; the nature reserve (undisturbed native ecosystem), add 0.2;

[0206] Comprehensive weight calculation and constraints: The comprehensive weight is calculated according to the formula "Comprehensive weight = Protection level weight + Scarcity weight". For example, Level I sensitive area + unique habitat of endangered species: 1.6 + 0.4 = 2.0; Level II sensitive area + wetland core area: 1.4 + 0.3 = 1.7. At the same time, an upper limit of 2.0 is set (to avoid excessive amplification) and a lower limit of 1.2 (to ensure that the lowest sensitive area still has a premium). Finally, a "Sensitive area level - comprehensive weight" correspondence table is formed and stored in the system database.

[0207] Step 3: Link the weighting factors with the existing accounting system, making targeted adjustments only to indicators within the sensitive area, thus highlighting their value without affecting the results in non-sensitive areas:

[0208] Embedded logic and scope of application: A "sensitive area weight switch" is added to the "urban area - typical area dual-level indicator system" of S300. When the system identifies that a grid cell or plot belongs to a sensitive area (through spatial coding matching), the corresponding comprehensive weight is automatically enabled; the default weight of non-sensitive areas is 1.0, and the original calculation logic remains unchanged.

[0209] The weights apply to all secondary indicators within the sensitive area. Among them, regulatory services (such as water conservation and carbon sequestration) and cultural services (such as nature education) are given priority. Material product indicators (such as forestry products) are subject to development restrictions, and their weights are only 1.1 times the normal value (to avoid encouraging development).

[0210] Weighted Calculation and Result Output: The final calculated value of a sensitive area = the conventional calculated value (S300-S400 calculation result) × the comprehensive weight. For example, the conventionally calculated water conservation value of a certain Class I sensitive area is 80 million yuan, and after weighting, it is 80 million × 2.0 = 160 million yuan, which directly reflects its ecological importance.

[0211] The final output is a "Special Report on the Accounting of Ecologically Sensitive Areas", which includes: a spatial distribution map of sensitive areas (labeled with levels and weight values); a comparison table of accounting results before and after weighting (highlighting the increase in regulatory service indicators); and an analysis of the proportion of ecological value of sensitive areas (e.g., the value of Level I sensitive areas accounts for 30% of the total regional value), providing data support for the management of ecological protection red lines and priority compensation for sensitive areas.

[0212] By assigning exclusive and special accounting weighting factors to ecologically sensitive areas, the problem of their value being underestimated is solved. This strengthens the guidance of accounting results on protection practices such as the formulation of ecological compensation standards for sensitive areas and the assessment of ecological access for development projects. Furthermore, through a "switch mechanism," it seamlessly integrates with the existing system, achieving refined management of "differentiated accounting" while maintaining overall stability, thus meeting the ecological management needs of "protection first."

[0213] In this embodiment of the application, the method further includes:

[0214] Step 1: Acquire three types of core human activity data through multiple channels, complete standardized preprocessing, and ensure that the data is quantifiable and spatially matchable.

[0215] Population density distribution data: The latest annual population statistics of townships (streets) in the accounting area are obtained from the statistics department. Combined with the administrative division vector map, the township-level population density (unit: people / square kilometer) is calculated by "population density = number of people in the area / area of ​​the area". At the same time, nighttime light remote sensing data (such as NPP-VIIRS data) is used to assist in verification and correct the deviations in the population statistics (such as density correction in areas with concentrated migrant population).

[0216] Industrial layout data: Vector boundaries of industrial parks (such as industrial parks, chemical parks, and agricultural parks) in the accounting area are obtained from natural resources and industry and information technology departments. Pollution / interference intensity is marked by industry type (e.g., chemical parks are high interference, agricultural parks are medium interference, and cultural and creative parks are low interference). At the same time, the industrial land area ratio is extracted (industrial land area in a certain area / total area of ​​the area, unit: %).

[0217] Traffic network density data: Obtain road vector data (including expressways, national highways, provincial highways, county roads, and rural roads) from the transportation department, assign weights according to road level (expressway 1.0, national highway 0.8, provincial highway 0.6, county road 0.4, rural road 0.2), and calculate the regional traffic network density (unit: km / square kilometer) using the formula "Traffic network density = Σ (road length × road weight) / area".

[0218] The three types of data are uniformly converted into the 2000 geodetic coordinate system and rasterized according to the preset raster precision of S400 (such as 30m×30m). Population density and transportation network density are generated into continuous rasters through Kriging interpolation, and industrial layout data are assigned the industrial interference intensity to the corresponding raster cells through "raster overlay". Finally, a human activity dataset that is completely matched with the ecological value accounting raster is formed.

[0219] Step 2: Based on the preprocessed human activity data, a model is constructed through "single-factor standardization - weight allocation - comprehensive index calculation" to quantify multi-dimensional human activities into a unified intensity index.

[0220] Single-factor standardization:

[0221] The "extreme value method" is used to standardize the three types of data to the [0,1] interval, eliminating dimensional differences:

[0222] Standardized population density value = (population density of a grid - minimum population density of the region) / (maximum population density of the region - minimum population density of the region); Standardized industrial disturbance value = industrial land area ratio × industry type weight (e.g., chemical industrial park weight 1.0, agricultural industrial park weight 0.6); Standardized transportation network density value = (transportation network density of a grid - minimum transportation density of the region) / (maximum transportation density of the region - minimum transportation density of the region).

[0223] Weighting and calculation of the comprehensive intensity index: Based on the weight ratio of "industrial interference (40%) > population density (30%) > transportation network density (30%)", the comprehensive human activity intensity index is calculated by "comprehensive human activity intensity index = (standardized value of industrial interference × 0.4) + (standardized value of population density × 0.3) + (standardized value of transportation network density × 0.3)", and the final index range is [0,1].

[0224] Intensity Level Classification: Human activity intensity is divided into four levels based on index values: low intensity (0-0.25), medium intensity (0.25-0.5), high intensity (0.5-0.75), and extremely high intensity (0.75-1.0), which facilitates subsequent targeted adjustments.

[0225] Step 3: Spatial correction of ecological value and generation of heat map based on intensity index.

[0226] Based on the intensity level of human activities, correction coefficients are determined to spatially adjust the raster-level ecological value results generated by S400, visually presenting the distribution of the impact of human activities:

[0227] Correction coefficient setting and correction logic: Following the principle that "the higher the intensity of human activities, the stronger the inhibition / disruption of ecological value, and the smaller the correction coefficient," four levels of correction coefficients are set:

[0228] Low intensity: Correction coefficient 1.0 (human activities have minimal impact on ecological value, no need to adjust downwards); Medium intensity: Correction coefficient 0.8 (ecological value is slightly disturbed, adjust downwards by 20%); High intensity: Correction coefficient 0.6 (ecological value is significantly disturbed, adjust downwards by 40%); Very high intensity: Correction coefficient 0.4 (ecological value is severely disturbed, adjust downwards by 60%); The correction coefficient for sensitive areas (S700 identified areas) can be increased by 10% (e.g., the correction coefficient for medium intensity activities in sensitive areas is 0.88) to avoid excessively underestimating the ecological value of sensitive areas;

[0229] Corrected calculation and heat map generation: Outside the sensitive area, the corrected ecological value = original ecological value of S400 × corresponding intensity correction coefficient; Inside the sensitive area, the corrected ecological value = original ecological value of S400 × corresponding intensity correction coefficient × 1.1;

[0230] For example: The original calculation of the water conservation value of an industrial park (extremely high intensity, non-sensitive area) was 5 million yuan, which was corrected to 5 million × 0.4 = 2 million yuan; the original calculation of the carbon sequestration value of a township center (medium intensity) in a sensitive area was 8 million yuan, which was corrected to 8 million × 0.88 = 7.04 million yuan.

[0231] Finally, based on the correction results, a "Heat Map of the Impact of Human Activities" was generated: different areas were marked with red (extremely high intensity, with the largest decrease in value), orange (high intensity), yellow (medium intensity), and green (low intensity, with value basically unchanged), clearly presenting the spatial interference pattern of human activities on ecological value.

[0232] By constructing a human activity intensity index model, the degree of disturbance caused by population density, industrial layout, and transportation network density is quantified, solving the problem that conventional accounting does not distinguish between densely populated areas and natural areas and cannot reflect the true impact of "human-ecological interaction." Through grid-level ecological value correction and activity impact heatmaps, areas with severe human disturbance are accurately located, providing a spatial basis for targeted ecological restoration and activity control. At the same time, its correction process is based on the grid-level results of S400, and the model parameters can be dynamically adjusted in combination with S200 statistical data without changing the original accounting logic, only adding the dimension of human activity impact, further improving the comprehensiveness of ecological value accounting.

[0233] In this embodiment of the application, the method further includes:

[0234] Step 1 involves acquiring three core disaster data categories. Standardization processes are used to ensure data quantifiability and risk assessment, laying the foundation for subsequent matrix construction: Flood frequency data for the past 30 years in the accounting area is obtained from the water resources department, including annual occurrence frequency, return period, and the affected ecosystem area caused by each flood; Standardized Precipitation Index (SPI) and Standardized Precipitation Evapotranspiration Index (SPEI) for the same period are obtained from the meteorological department, and the average annual number of drought days and the types of ecosystems affected by drought are statistically analyzed according to "slight, moderate, and severe" levels; Detailed regional geological disaster survey reports are obtained from the natural resources department, extracting geological erosion modulus and erosion types, and classifying erosion intensity levels according to "micro, slight, moderate, and severe" levels. In the preprocessing stage, all data are integrated into an "annual" time granularity to unify the risk assessment benchmark. Disaster data are associated with 30-meter precision raster units in the accounting area through GIS spatial overlay, generating a "raster-disaster attribute" association table. Simultaneously, the "3σ principle" is used to remove extreme anomalies, and the data is replaced with a 5-year moving average to ensure data representativeness.

[0235] Step 2: Construct a "frequency-loss" dual-dimensional assessment matrix based on preprocessed data, quantifying qualitative disaster information into calculable ecological value loss risk coefficients. The matrix is ​​divided horizontally into three disaster occurrence frequency levels: "low frequency (≤1 time per year / mild), medium frequency (1-3 times per year / moderate), and high frequency (≥3 times per year / severe)". Vertically, it is divided into three ecological loss levels: "low loss (ecological value affected <5%), medium loss (5%-15%), and high loss (>15%)". These correspond to matrix cell relationships of "low risk, medium risk, and high risk" (e.g., low frequency-low loss corresponds to low risk, medium frequency-medium loss corresponds to medium risk, and high frequency-high loss corresponds to high risk). When quantifying risk coefficients, initial coefficients are assigned according to risk level (0.1 for low risk, 0.3 for medium risk, and 0.5 for high risk), with the coefficient range limited to [0, 0.5]. If a grid is affected by multiple disasters simultaneously, the maximum value of the risk coefficients of each disaster is taken to avoid double counting. Referring to the "Technical Guidelines for Ecosystem Risk Assessment", the risk coefficients of special ecosystems such as wetland core areas and endangered species habitats are increased by 20% to highlight the special risk characteristics of sensitive areas.

[0236] Step 3: Embed the risk coefficient into the S300 spatiotemporal premium parameter correction mechanism to generate an ecological value accounting result that includes "risk premium" to reflect the impact of disaster risk on value: On the basis of S300 "time premium coefficient × hydrological premium coefficient", a "risk premium adjustment item" is added to construct the calculation logic of "comprehensive spatiotemporal risk premium coefficient = original spatiotemporal premium coefficient × (1 + risk coefficient)", where "1 + risk coefficient" is used to reflect "the higher the risk, the stronger the 'risk compensation demand' of ecological value" (e.g., the risk coefficient of a high-risk flood area is 0.5, the original spatiotemporal premium coefficient during the flood season is 1.6, and the comprehensive premium coefficient is 1.6 × (1 + 0.5) = 2.4). When adjusting the value, the ecological value adjustment value outside the sensitive area = S300 basic value × comprehensive spatiotemporal risk premium coefficient, and the adjustment value of the sensitive area (S700 identified area) = S300 basic value × comprehensive spatiotemporal risk premium coefficient × S700 sensitive area weight, ensuring dual consideration of risk and sensitivity attributes. For example, if the basic value of farmland water conservation in a non-sensitive area is 10 million yuan, the hydrological premium coefficient during the flood season is 1.6, and the high-risk coefficient for flooding is 0.5, the adjusted value is 10 million × 1.6 × (1 + 0.5) = 24 million yuan. Finally, the output is a "Natural Disaster Risk-Ecological Value Premium Thematic Map" (marking the risk level and corresponding premium coefficient with different colors) and a "Multi-hazard Risk Value Distribution Table" (statistically calculating the risk coefficient and the value before and after adjustment by grid unit), clearly presenting the pattern of the impact of risk on ecological value.

[0237] By quantifying the potential loss risk to ecological value from natural disasters such as floods and droughts using risk coefficients, this addresses the problem that conventional accounting only focuses on static current value and does not include this risk. It upgrades the accounting results to dynamic value that takes into account risk, thus meeting the needs of "risk prevention and control." At the same time, the results that include risk premiums can directly guide the formulation of ecological restoration priorities and ecological insurance pricing in high-risk disaster areas, enhancing practical relevance. Furthermore, the risk coefficient is only integrated into the S300 spatiotemporal premium system as an "adjustment item," and the data preprocessing is consistent with S200 and S400, which both supplements the risk dimension and ensures the stability and consistency of the overall accounting system.

[0238] In this embodiment of the application, "solving the problem of 'spatial fragmentation' accounting of ecosystem services, quantifying cross-regional ecological value transfer relationships, and supporting regional collaborative ecological management" is achieved through three steps: "construction of functional flow network model - calculation of spatial flow volume and transfer efficiency - cross-regional value listing and association map generation." The specific method also includes:

[0239] Step 1: Based on multi-source spatial data, clarify the spatial chain of "supply-transmission-benefit" of ecosystem services and build a visual network model. First, collect four types of core data: Obtain ecosystem type distribution maps of the accounting area and its surroundings from the natural resources department (to identify supply areas, such as mountains with high vegetation cover and water conservation wetlands), and ecological corridor vector data (such as rivers and greenways, as transmission paths); obtain river system network data from the water resources department (to mark watershed boundaries and water flow direction to assist in identifying water regulation service transmission paths); and obtain population and industrial agglomeration data from the statistics department (to identify beneficiary areas, such as urban built-up areas and major agricultural production areas, and areas that rely on external ecosystem services). Subsequently, a model was constructed using GIS spatial analysis tools: adopting a "node-corridor" network structure, the supply area was set as a "supply node" (weighted according to the ecological service supply capacity, such as 1.0 for 10,000 mu of forest land and 0.8 for 1,000 mu of wetland), the transmission path was set as a "connection corridor" (labeled according to the transmission capacity, such as "high transmission efficiency" for main rivers and "medium transmission efficiency" for secondary greenways), and the beneficiary area was set as a "beneficiary node" (weighted according to the demand intensity, such as 1.2 for urban areas with a population of one million and 0.9 for 10,000 mu of farmland); at the same time, the minimum transmission cost from the supply node to the beneficiary node was calculated using the "cost distance analysis" tool (such as the water transport distance from wetlands to urban areas and the diffusion cost of vegetation carbon sequestration to industrial areas through atmospheric circulation), and invalid connections with transmission costs exceeding the threshold (such as exceeding 50 kilometers) were eliminated, finally forming a complete "supply-transmission-benefit" functional flow network model.

[0240] Step 2: Employ specialized algorithms to quantify the scale and efficiency of ecosystem service transmission in spatial chains, avoiding "vague value transmission": First, identify the core ecosystem service types for accounting (focusing on three types of services that can be transmitted across regions: water regulation, carbon sequestration and oxygen release, and soil conservation). Then, select appropriate spatial flow analysis algorithms for different services: For water regulation services, use the "water conservation" module of the InVEST model, combining precipitation and evapotranspiration data to calculate the actual water yield of the supply area (e.g., mountainous areas), and then calculate using the "water decay model" of the river network (considering evaporation and seepage losses). The amount of water reaching the beneficiary area (such as downstream urban areas) is the water transfer volume. For carbon sequestration and oxygen release services, the "atmospheric diffusion model" is used to simulate the carbon sink diffusion range after carbon sequestration in the supply area (such as forests) based on wind speed and direction data. Combined with the carbon emissions of the beneficiary area (such as industrial areas), the amount of carbon sink absorbed by the beneficiary area is calculated, which is the carbon sink transfer volume. For soil conservation services, the RUSLE model is used to calculate the amount of soil erosion reduced in the supply area (such as protective forests around sloping farmland). Then, the "sediment transport model" is used to calculate the amount of sediment intercepted by downstream farmland (beneficial area), which is the soil conservation transfer volume. Based on this, the value transfer efficiency is calculated as follows: "Transfer efficiency = (Actual amount of ecosystem services received by the beneficiary area / Amount of ecosystem services output by the supply area) × 100%". For example, if a wetland supply area outputs 10 million cubic meters of water regulation services annually, and the downstream urban area actually receives 8 million cubic meters (after deducting 2 million cubic meters of evaporation and seepage), then the water regulation service transfer efficiency = 800 / 1000 × 100% = 80%. At the same time, the efficiency value is corrected for loss factors in the transmission path (such as river pollution leading to a decline in water service quality, and greenway interruption leading to obstruction of vegetation service transmission). For example, the efficiency is reduced by 10% due to polluted river sections and by 20% due to greenway interruption, to ensure that the results are consistent with the actual transmission scenario.

[0241] Step 3: Separately calculate and visualize the ecological value transferred across regions to clearly present the ecological value correlation between regions: First, according to the value calculation logic of S300, convert the spatial flow volume into cross-regional value: For example, the cross-regional value of water regulation services = water flow volume × unit water resource value (taken from local water price and ecological compensation standard, such as 5 yuan / cubic meter), and the cross-regional value of carbon sinks = carbon sink flow volume × carbon trading price (such as 60 yuan / ton); Then, add a separate "cross-regional ecological service value" module to the overall calculation results, which is listed in three parts: the value output by the supply area, the value obtained by the beneficiary area from the outside, and the value of transmission path loss, and mark the corresponding related areas (such as "the water value of the urban area comes from XX wetland"). Finally, an ecological service spatial correlation map is generated: using GIS visualization technology, different colors are used to mark the supply area (green), transmission path (blue, the direction of the arrow indicates the transmission direction, and the thickness of the arrow indicates the amount of flow), and beneficiary area (yellow); numerical labels are superimposed on the map to mark the flow of each node (e.g., "wetland → urban area: 8 million cubic meters of water") and transmission efficiency (e.g., "80%)", and a heat map is used to show the dense areas of value transmission (e.g., the dense areas of value transmission along the river are marked in dark red), intuitively presenting the cross-regional ecological value correlation pattern.

[0242] By constructing an ecosystem service function flow network model, the supply-transmission-benefit relationship of cross-regional ecosystem services is quantified, which solves the problem that conventional accounting fragments regions, leading to underestimation of the value of supply areas and overestimation of the value of beneficiary areas, and makes the results consistent with the spatial mobility characteristics of ecosystem services. At the same time, cross-regional value listing can directly support horizontal transfer payments for ecological compensation and the formulation of cross-regional ecological protection agreements, and enhance regional collaborative ecological management.

[0243] In this embodiment of the application, the method further includes:

[0244] Step 1: First, collect three types of core data: Obtain vegetation coverage and biodiversity index from the ecological and environmental departments, combined with S800 human activity intensity and S900 disaster frequency, as the basis for assessing resilience (ability to withstand disturbance); obtain historical vegetation recovery rate and ecological restoration input data after disturbance from the natural resources departments to support the calculation of resilience (ability to recover after disturbance); obtain future climate prediction data from the meteorological departments, combined with species adaptation assessment reports, to determine the adaptability (ability to adapt to long-term changes) index. Then, standardize the data to the [0,1] interval, assign weights according to "resistance 40%, resilience 35%, adaptability 25%", and calculate using "three-dimensional resilience index = resistance × 0.4 + resilience × 0.35 + adaptability × 0.25", with an index range of [0,1], where higher values ​​indicate stronger resilience.

[0245] Step 2: Using the total traditional ecological value (V) calculated by S300 and the three-dimensional resilience index (E) as coupling objects, first calculate the degree of correlation between the two using the "coupling degree formula" (C∈[0,1], ≥0.7 indicates high coupling), and assess the overall development level using the "coordination degree formula" (D∈[0,1], ≥0.7 indicates excellent coordination). Then, construct a comprehensive value model according to the coordination degree classification: when the coordination degree is excellent (D≥0.7), the comprehensive value = V×0.6 + (E×1000)×0.4 (E is magnified 1000 times to match the magnitude of V); when the coordination degree is moderate (0.4≤D<0.7), the comprehensive value = V×0.7 + (E×1000)×0.3; when the coordination degree is unbalanced (D<0.4), the comprehensive value = V×0.5 + (E×1000)×0.5, while setting a baseline constraint (when E≤0.3, the comprehensive value is reduced by 10%~20%).

[0246] Step 3: Add a "Resilience Feature Module" to the traditional GEP report, listing the three-dimensional resilience sub-scores, coupling coordination degree, and comprehensive value (including traditional value and resilience contribution value). Then, classify the health level according to the dual indicators of "comprehensive value + resilience index": Excellent (comprehensive value ≥ 30 million yuan / km² and E ≥ 0.7), Good (2000 ≤ comprehensive value < 30 million yuan / km² and 0.5 ≤ E < 0.7), Medium (1000 ≤ comprehensive value < 20 million yuan / km² and 0.3 ≤ E < 0.5), Poor (comprehensive value < 10 million yuan / km² or E < 0.3). Generate a health level map through GIS, using dark green, light green, yellow, and orange-red to mark each level, intuitively presenting the "value-resilience" matching status.

[0247] By filling the gap in "ecosystem resilience" assessment, it addresses the problem of traditional accounting focusing on the short term while neglecting the long term, enabling accounting results to balance short-term value with the long-term stability of ecosystems in resisting, recovering, and adapting to disturbances. At the same time, the health map it generates can directly guide differentiated ecological restoration (such as prioritizing the enhancement of resilience in areas with "high value but low resilience" and carrying out systematic restoration in areas with "low value and weak resilience"), thereby improving the guidance of ecological management.

[0248] This application discloses an ecosystem gross product accounting device based on GIS spatial analysis, referring to... Figure 2 ,include:

[0249] The semantic mapping library construction module 001 calculates regional basic data and satellite imagery data to construct a two-way dynamic semantic mapping library corresponding to the land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes a primary mapping and a secondary mapping. The primary mapping associates the land and ecosystem primary categories, and the secondary mapping achieves the alignment of secondary category terms through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with the annual land data.

[0250] The multi-source data fusion module 002 acquires the corresponding multi-source data, which includes land change survey vector data, satellite remote sensing raster data that meets the required accuracy, IoT real-time monitoring data, and statistical report data. The monitoring data includes soil moisture, air quality, and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed using spatial interpolation-boundary correction and sliding window smoothing methods to obtain the corresponding standardized dataset.

[0251] The accounting indicator system configuration and calculation module 003 constructs a two-tier accounting indicator system for urban areas and typical districts. The urban area level includes three primary indicators: material products, regulatory services, and cultural services, as well as 14 secondary indicators. The typical district level consists of four types of characteristic areas with exclusive secondary indicators, which are enabled or disabled through dynamic switches. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model.

[0252] The GIS spatial calculation and parameter calibration module 004 retrieves pre-trained digital elevation models and 3D building models based on GIS, extracts parameters to correct soil erosivity factor values ​​and vegetation transpiration, quantifies the indicators to preset precision raster cells, and obtains raster-level ecological value results through neighborhood analysis algorithms; it also connects IoT monitoring data to calculate the deviation rate between measured and calculated values, identifies indicators exceeding thresholds, uses random forest algorithms to calibrate the calculation parameters, controls the deviation rate within a preset range, and combines GIS to generate a multi-year ecological value trend map.

[0253] The scenario-based output and quality control module 005 outputs customized results based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios. It is verified through a three-level quality control mechanism: data level completes missing data, calculation level intercepts abnormal parameters, and result level outputs the final result after sampling and verification by third-party measured data.

[0254] This application also discloses an ecosystem gross product accounting device based on GIS spatial analysis, including a processor, wherein the processor runs a program for the ecosystem gross product accounting method based on GIS spatial analysis described in any one of the above.

[0255] This application also discloses a storage medium storing a program for the ecosystem gross product accounting method based on GIS spatial analysis as described in any one of the above embodiments.

[0256] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for calculating the Gross Ecosystem Product based on GIS spatial analysis, characterized in that, include: The accounting area's basic data and satellite imagery data are used to construct a two-way dynamic semantic mapping library corresponding to the land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes a primary mapping and a secondary mapping. The primary mapping associates the land and ecosystem primary categories, and the secondary mapping achieves the alignment of secondary category terms through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with the annual land data. The corresponding multi-source data is acquired, including land change survey vector data, satellite remote sensing raster data of the required accuracy, IoT real-time monitoring data and statistical report data. The monitoring data includes soil moisture, air quality and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed through spatial interpolation-boundary correction and sliding window smoothing method to obtain the corresponding standardized dataset. A two-tiered accounting indicator system is constructed, consisting of urban areas and typical districts. The urban area level includes three primary indicators: material products, regulatory services, and cultural services, as well as 14 secondary indicators. The typical district level is configured with exclusive secondary indicators for four types of characteristic areas, and these exclusive secondary indicators are enabled or disabled through dynamic switching. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model. Based on GIS, pre-trained digital elevation models and 3D building models are retrieved, parameters are extracted to correct soil erosivity factor values ​​and vegetation transpiration, and the indicators are quantified to a preset precision raster unit. The raster-level ecological value results are obtained by processing the neighborhood analysis algorithm. The deviation rate between measured and calculated values ​​is calculated by accessing IoT monitoring data, and indicators exceeding the threshold are identified. The random forest algorithm is used to calibrate the calculation parameters and control the deviation rate within a preset range. Combined with GIS, a multi-year ecological value trend map is generated. Customized results are output based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios, and are verified through a three-level quality control mechanism: data level to complete missing data, calculation level to intercept abnormal parameters, and result level to output the final result after passing the sampling verification of third-party measured data.

2. The method for calculating the Gross Ecosystem Product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: Obtain historical ecological data for the accounting area, including annual climate change data, vegetation cover change data, and land use transformation data. An ecological value decay coefficient model is constructed, and historical ecological data is compared with current accounting data over time to calculate the value decay coefficient for different ecosystem types. Based on the value decay coefficient, the current accounting results are corrected over time to generate comparable ecosystem GDP accounting results across years.

3. The method for calculating the Gross Ecosystem Product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: Identify ecologically sensitive areas within the accounting area, including nature reserves, wetland core areas, and habitats of endangered species; A unique accounting weighting factor is assigned to ecologically sensitive areas, and the weighting factor is determined based on the protection level and the scarcity of ecosystem services in the ecologically sensitive areas. By embedding special accounting weighting factors into a two-level accounting indicator system, the accounting results for ecologically sensitive areas are adjusted with weights.

4. The method for calculating the gross ecosystem product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: Collect human activity data for the accounting area, including population density distribution, industrial layout data, and transportation network density; Construct a human activity intensity index model to quantify human activity data into a standardized intensity index; Spatial corrections are made to the ecological value quantification results based on the human activity intensity index to generate a heat map of activity impacts.

5. The method for calculating the gross ecosystem product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: Obtain historical data on natural disasters in the accounting area, including flood frequency, drought intensity, and geological erosion degree; Establish a natural disaster risk assessment matrix and convert historical natural disaster data into ecological value loss risk coefficients; By incorporating the risk coefficient into the spatiotemporal premium parameter correction mechanism, an ecosystem GDP accounting result containing a risk premium is generated.

6. The method for calculating the gross ecosystem product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: Construct an ecosystem service function flow network model to identify the supply areas, transmission paths and beneficiary areas of ecosystem services within the accounting region; Calculate the spatial flow volume and value transfer efficiency of ecosystem services based on spatial flow analysis algorithms; The cross-regional ecosystem service value is listed separately in the accounting results, and a spatial correlation map of ecosystem services is generated.

7. The method for calculating the gross ecosystem product based on GIS spatial analysis according to claim 1, characterized in that, The method also includes: An ecosystem resilience assessment dimension is introduced to construct a three-dimensional resilience index that includes resistance, resilience, and adaptability; By coupling the resilience index with traditional ecological value indicators, a comprehensive ecosystem value assessment model is established. Based on the integrated ecosystem value assessment model, the output of the ecosystem gross product accounting results including resilience characteristics is used to generate an ecosystem health grading map.

8. An ecosystem gross product accounting device based on GIS spatial analysis, characterized in that, include: The semantic mapping library construction module calculates regional basic data and satellite imagery data to construct a two-way dynamic semantic mapping library corresponding to land parcel classification and ecosystem type in the land change survey. The two-way dynamic semantic mapping library includes primary mapping and secondary mapping. The primary mapping associates land and ecosystem primary categories, and the secondary mapping achieves secondary category term alignment through ontological semantic matching. The two-way dynamic semantic mapping library is automatically updated with annual land data. The multi-source data fusion module acquires corresponding multi-source data, including land change survey vector data, satellite remote sensing raster data of the required accuracy, IoT real-time monitoring data, and statistical report data. The monitoring data includes soil moisture, air quality, and negative oxygen ion concentration data. Based on the multi-source data, spatiotemporal fusion is performed using spatial interpolation-boundary correction and sliding window smoothing methods to obtain the corresponding standardized dataset. The accounting indicator system configuration and calculation module constructs a two-tier accounting indicator system for urban areas and typical districts. The urban area level includes three primary indicators (material products, regulatory services, and cultural services) and 14 secondary indicators. The typical district level is configured with exclusive secondary indicators for four types of characteristic areas, and these exclusive secondary indicators are enabled or disabled through dynamic switches. The physical quantity and value of all accounting indicators are calculated simultaneously. The calculation of physical quantity introduces a regional correction coefficient, and the calculation of value introduces a spatiotemporal premium model. The GIS spatial calculation and parameter calibration module retrieves pre-trained digital elevation models and 3D building models based on GIS, extracts parameters to correct soil erosivity factor values ​​and vegetation transpiration, quantifies the indicators to preset precision raster cells, and obtains raster-level ecological value results through neighborhood analysis algorithms. It also connects IoT monitoring data to calculate the deviation rate between measured and calculated values, identifies indicators exceeding thresholds, uses random forest algorithms to calibrate the calculation parameters, controls the deviation rate within a preset range, and combines GIS to generate a multi-year ecological value trend map. The scenario-based output and quality control module outputs customized results based on ecological compensation scenarios, EOD project scenarios, and ecological audit scenarios. It is verified through a three-level quality control mechanism: data level completes missing data, calculation level intercepts abnormal parameters, and result level outputs the final result after passing the sampling verification of third-party measured data.

9. An ecosystem gross product accounting device based on GIS spatial analysis, characterized in that, Includes a processor running a program for an ecosystem gross product accounting method based on GIS spatial analysis as described in any one of claims 1-7.

10. A storage medium, characterized in that, The program stores the method for calculating the gross ecosystem product based on GIS spatial analysis as described in any one of claims 1-7.

Images