A method and system for spatiotemporal analysis of lake shore zone coupled with lake shore distance and terrain
By using a spatiotemporal analysis method that couples shoreline distance with topography, the consistency and repeatability issues of shoreline environmental analysis in existing technologies have been resolved. This method enables unified comparison and quantitative regulation of multi-source ecological indicators, thereby improving the accuracy of lake ecological monitoring and drought risk identification.
Patent Information
- Application Number
- CN202611620817.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-20
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to unify shoreline distance gradients, topographic factors, and multi-temporal ecological indicators into a single computational framework, resulting in insufficient consistency of lakeside environmental analysis results and limited engineering application value. Furthermore, they lack a repeatable buffer-ring analysis structure and a quantified feature space inversion link.
By constructing a spatiotemporal analysis method for lakeshore zones that couples lakeshore distance with topography, including resampling, temporal synthesis, establishment of distance-topography joint analysis units, extraction of multi-indicator feature spaces, and trend analysis, we can achieve the same-caliber comparison of multi-source ecological indicators and quantitatively distinguish the relative contributions of lake regulation and topographic modulation effects.
It enables multi-scale, consistent ecological environment diagnosis and regulation zoning, improves the ability to identify lakeside thermal environment, vegetation status, and wet/dry conditions, and outputs verifiable standardized deliverables to support ecological monitoring and drought risk identification of plateau lakes.
Smart Images

Figure CN123115879A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of remote sensing ecological monitoring, geographic information processing, lakeside environmental diagnosis and ecological zoning, and in particular to a spatiotemporal analysis method and system for lakeside zones coupled with shoreline distance and topography. Background Technology
[0002] Large lakes and their surrounding shorelines are among the most sensitive areas for air exchange, heat regulation, water redistribution, and ecological responses. In the context of plateau lakes, the combined effects of lake water heat capacity, evaporation recharge, and complex topography result in significant differences in thermal, vegetation, and wet / dry conditions between nearshore and offshore areas.
[0003] Existing technologies typically conduct surface temperature analysis, vegetation index analysis, drought index analysis, or single spatial unit statistics separately, making it difficult to unify shoreline distance gradient, topographic factors, and multi-temporal ecological indicators into a single computational framework. Especially when identifying lake regulation effects at a quarterly scale, inconsistencies in distance zoning, topographic zoning, and temporal composite zoning make direct comparisons between different indicators difficult, hindering the formation of stable spatial zoning conclusions. Furthermore, existing technologies rely heavily on empirical judgment for identifying nearshore core regulation zones, intermediate transition zones, and offshore topographically dominant zones, lacking repeatable buffer-ring analysis structures, quantifiable feature spatial inversion links, and traceable coupling identification rules. In applications such as plateau lake ecological monitoring, wetting assessment, and drought risk identification, these deficiencies directly impact the consistency of results and their engineering application value.
[0004] Existing technologies for spatiotemporal analysis of the surface environment in lakeside zones have the following shortcomings: Firstly, in existing analyses of lakeside zones, surface temperature, vegetation index, and surface dryness and wetness are usually modeled separately, lacking a unified temporal synthesis and spatial alignment process, making it difficult to make comparisons between different indicators using the same caliber. Secondly, existing studies often only use statistics from a single buffer zone or the entire study area, which makes it difficult to express the continuous attenuation pattern of the lake's regulating effect along the distance from the shoreline, and also makes it difficult to identify the transition structure between the nearshore core area and the offshore dominant area. Third, the environmental processes in the lakeside zone are significantly affected by elevation, slope and aspect. Without the joint analysis of topographic factors and distance gradient, it is impossible to quantitatively distinguish the relative contributions of lake regulation and topographic modulation. Fourth, relying solely on LST or NDVI is insufficient to characterize changes in the surface ecological state of the lakeside zone, and its ability to identify the process of wet-dry transition is poor. Therefore, it is necessary to propose a spatiotemporal analysis method for the surface environment of the lakeside zone that simultaneously integrates the lake shore distance gradient, topographic factors of the digital elevation model, and multi-source ecological indicators such as LST, NDVI, and TVDI, so as to achieve multi-scale, consistent, and reusable ecological environment diagnosis and regulation zoning. Summary of the Invention
[0005] The purpose of this invention is to address the technical deficiencies in the existing technology by providing a spatiotemporal analysis method for lakeshore zones that couples lakeshore distance with topography.
[0006] Another objective of this invention is to provide a spatiotemporal analysis system for lakeshore zones that couples lakeshore distance with topography.
[0007] The technical solution adopted to achieve the purpose of this invention is: A spatiotemporal analysis method for lakeshore zones coupled with shoreline distance and topography includes the following steps: Step S1: Obtain LST data, NDVI data, DEM data, and shoreline boundary data of the area surrounding the target lake. Resample the LST data and NDVI data to a unified target grid to obtain the pixel values after resampling to the target grid. Step S2: Perform time-series synthesis on the cell values resampled to the target grid in step S1 according to the pre-divided quarters, calculate the quarterly synthesis results, and construct a spatiotemporal data cube. ; Step S3: Based on the shoreline boundary data from Step S1, construct the target buffer zone and divide it into multiple rings. Derive the elevation, slope, and aspect based on the DEM data, and compare them with... Perform registration and construct a distance-terrain joint analysis unit; Step S4: Extract effective pixel pairs from the time-series synthesized LST and NDVI data within the same quarter, establish the NDVI-LST feature space, extract dry-edge temperature samples and wet-edge temperature samples, perform linear fitting, and calculate the temperature-vegetation drought index. and will The indicator value is constrained to 0-1; Step S5, combine the time-series synthesized LST data, NDVI data, and As indicators, the mean, standard deviation, extreme values and quantiles of each indicator are calculated in various distance-topography joint analysis units. The Theil-Sen slope estimation and Mann-Kendall significance test are used to jointly identify the slope and significance of the indicator change trend. The Pearson correlation coefficient is calculated, and the output is a quarterly statistical table of LST, NDVI, and TVDI, a statistical table of lake shore distance ring, a trend analysis chart and an indicator relationship chart. Step S6: Based on the output of step S5, calculate the comprehensive adjustment score, classify the lakeside zone according to the comprehensive adjustment score, and output the standardized result package of the lakeside zone classification results.
[0008] In the above technical solution, the formula for calculating the pixel value after resampling to the target grid in step S1 is: ; In the formula, To resample to the target grid in time phase Lower target mesh cell The pixel value at that location, To map to target mesh cells The source pixel set, For pixels, For the target mesh cell, For the original pixel set in time phase Lower target mesh cell The pixel value at that location, For variables In time phase lower pixel Effective mask; in, The calculation formula is: ; In the formula, For quality control marking, For variables The set of quality control codes that are allowed to pass.
[0009] In the above technical solution, the seasonal synthesis result in step S2 is expressed as follows: ; In the formula, For target mesh cells In the quarter The synthesized value below, Belonging to the quarter The set of time phases, To resample to the target grid in time phase Lower target mesh cell The pixel value at that location, For target mesh cells In the quarter The number of valid observations within; The Represented as: ; In the formula, For target mesh cells The LST composite value for each quarter For target mesh cells The NDVI composite value for each quarter For target mesh cells The corresponding digital elevation model data, For target mesh cells The corresponding distance to the lake shore, For target mesh cells The slope, For target mesh cells The slope direction; in, The calculation formula is: ; In the formula, and Elevation at direction and Gradient in the direction.
[0010] In the above technical solution, in step S2, the division of the year can be divided into four quarters: Q1, Q2, Q3 and Q4, or into two quarters: cold and warm, or into two quarters: dry and rainy. Preferably, the year is divided into four quarters: Q1, Q2, Q3 and Q4.
[0011] In the above technical solution, the calculation formula for the annular band in step S3 is: ; In the formula, For the first Level ring belt, For the lake shoreline, For the lake shoreline As the reference, with a radius of Cumulative buffer, For the lake shoreline As the reference, with a radius of Cumulative buffer, For the first The radius of the ring belt, for The radius of the ring belt; The distance-terrain joint analysis unit is represented as follows: ; In the formula, For the first First ring belt, the first Level Elevation, First Grade 1 and Grade 2 Distance-topography joint analysis unit formed by overlapping slope aspects, For the first Level elevation, For the first Grade slope, For the first Slope type.
[0012] In the above technical solution, in step S4, the... The calculation formula is: ; In the formula, For operators used to constrain results to the interval between 0 and 1, For the linear fit of the dry edge, This is a linear fit for the wet edge. For the target pixel's surface temperature; Represented as: ; In the formula, The slope of the linear fit along the dry edge. The intercept of the linear fit along the dry edge. Normalized Difference Vegetation Index; Represented as: ; In the formula, The slope of the linear fit for the wet edge. The intercept of the linear fit at the wet edge; The formula for calculating the dry edge temperature sample is: ; In the formula, For the first Dry edge temperature samples corresponding to each NDVI sub-bin. For high percentile operators, For the first NDVI boxes, For NDVI packaging; The formula for calculating the wet edge temperature sample is: ; In the formula, For the first Each NDVI sub-bin corresponds to a wet edge temperature sample. This is the low percentile operator.
[0013] In the above technical solution, the formula for calculating the mean value of each indicator within any distance-terrain joint analysis unit in step S5 is as follows: ; In the formula, As an indicator In the quarter and distance-terrain joint analysis unit The mean within, For quarters The number of effective pixels within, The serial number of the valid pixel. For the first The index corresponding to each valid cell ; The Theil-Sen slope estimation method identifies the Theil-Sen trend slope of each indicator, and its calculation formula is as follows: ; In the formula, As an indicator The slope of the Theil-Sen trend, For the first The index value at each observation time, For the first The index value corresponding to each observation time; The significance of the change in the indicator is identified by calculating the test statistic using the Mann-Kendall significance test. The formula for calculating the test statistic is as follows: ; In the formula, The Mann-Kendall test statistic is used. For symbolic functions, This represents the length of the time series.
[0014] In the above technical solution, the formula for calculating the comprehensive adjustment score in step S6 is: ; In the formula, For the first The overall adjustment score of the ring belt, For the first The relative cooling index of the ring belt. For the first The weighting coefficient of the relative cooling index of the ring belt. For the first The relative greening index of the first-level ring zone, For the first The weighting coefficient of the relative greening index of the ring zone. For the first The relative drought reduction index of the first-level ring zone, No. The weighting coefficients of the relative drought reduction index of the zonal rings. For the first Topographic support items for the ring zone, These are the weighting coefficients for the normalized terrain support term; in, The calculation formula is: ; In the formula, For reference area in quarterly LST mean under, For the first The tertiary ring belt in the quarter LST mean under, To prevent stable terms with a denominator of zero; The calculation formula is: ; In the formula, For reference area in quarterly The mean NDVI under the following conditions For the first The tertiary ring belt in the quarter The mean NDVI value under the following conditions; The calculation formula is: ; In the formula, For reference area in quarterly TVDI mean value For the first The tertiary ring belt in the quarter The mean TVDI value.
[0015] In the above technical solution, in step S6, the lakeside zone is categorized into three types: lake regulation-dominant zone, transition zone, and topographically dominant zone. The categories of the lakeside zone are represented as follows: ; In the formula, For the first Category of ring zone or joint zone, For classification with a high threshold, For classification with a low threshold, where, , and It can be preset or determined based on sample statistics.
[0016] The above technical solution is characterized by including a data acquisition and standardization preprocessing module for implementing step S1; a quarterly synthesis and spatiotemporal data cube construction module for implementing step S2; a lake shore distance gradient and topographic zoning module for implementing step S3; an NDVI-LST feature space and TVDI inversion module for implementing step S4; a multi-scale statistics and trend analysis module for implementing step S5; and a lake effect-topographic coupling identification and result output module for implementing step S6.
[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention uses the lake shoreline as a unified spatial anchor point and the quarterly time series as a unified temporal anchor point to unify LST, NDVI, TVDI and topographic factors into the same analytical framework, so as to achieve the same-caliber comparison of multi-source ecological indicators. 2. This invention constructs a multi-level lake shore distance gradient through cumulative buffering and ring differential structure, which is beneficial for characterizing the attenuation law of lake regulation effect from near shore to far shore and reducing scale confusion caused by single buffer statistics; 3. This invention constructs multi-level ring zones through cumulative buffering and ring zone differential structure, and overlays the ring zones with elevation level, slope level, and aspect category to construct a distance-topography joint analysis unit. This allows the statistics, trend identification, and related calculations of each indicator to be completed within a unified joint analysis unit, thereby enabling the quantitative differentiation of the relative effects of lake regulation and topographic modulation. 4. This invention improves the ability to identify the co-evolution of lakeside thermal environment, vegetation status and dry / wet status by inverting TVDI in the NDVI-LST feature space and combining trend analysis and correlation analysis. 5. The invention further sets up a reference area and constructs relative cooling index, relative greening index, relative drought reduction index, and topographic support item. By comprehensively adjusting the score and classification threshold, the lakeside zone is quantitatively divided into lake regulation-dominant area, transition area, and topographic-dominant area. Threshold parameters, weight coefficients, buffer distance sequences, quarterly classification caliber, and version number information are recorded simultaneously to form a standardized result package that is verifiable, reproducible, and archiveable. It can be used for plateau lake ecological monitoring, wetting assessment, drought risk identification, protection zoning, and related result archiving. Attached Figure Description
[0018] Figure 1 The flowchart shown is a spatiotemporal analysis method for lakeshore zone coupling of lake shore distance and topography according to the present invention.
[0019] Figure 2 The diagram shown is a flowchart of the data acquisition and standardized preprocessing process of this invention.
[0020] Figure 3 The diagram shown is a flowchart of the quarterly synthesis and spatiotemporal data cube construction process of this invention.
[0021] Figure 4 The diagram shown is a flowchart of the lake shore distance gradient and terrain zoning of this invention.
[0022] Figure 5 The diagram shown is a flowchart of the NDVI-LST feature space and TVDI inversion process of this invention.
[0023] Figure 6 The diagram shown is a flowchart of the multi-scale statistical and trend analysis process of this invention.
[0024] Figure 7 The diagram shown is a flowchart of the lake effect-terrain coupling identification and result output of this invention.
[0025] Figure 8 This is a schematic diagram of the study area for the Qinghai Lake example, including the plateau background range and the analysis range of the target lakeshore zone.
[0026] Figure 9 This is a spatial distribution map of LST, NDVI, and TVDI at the quarterly scale from Q1 to Q4 in the Qinghai Lake example.
[0027] Figure 10 This is a schematic diagram of the NDVI-LST feature space for a certain quarter in the Qinghai Lake example, showing the fitting results of dry and wet edges.
[0028] Figure 11 This is a graph showing the elevation and slope variations along the shoreline in the Qinghai Lake example.
[0029] Figure 12 This is a diagram showing the quarterly variations of LST, NDVI, and TVDI under different shoreline distances in the Qinghai Lake example.
[0030] Figure 13 This is a quarterly time series trend analysis chart of LST, NDVI, and TVDI in the Qinghai Lake example.
[0031] Figure 14 This is a graph showing the relationship between the quarterly average values of LST, NDVI, and TVDI in the Qinghai Lake example. Detailed Implementation
[0032] The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0033] Example 1 A spatiotemporal analysis method for lakeshore zones coupled with shoreline distance and topography, referring to... Figure 1 This includes the following steps: Step S1, Data Acquisition and Standardization Preprocessing: Refer to Figure 2The study acquires land surface temperature (LST), vegetation index (NDVI), digital elevation model (DEM), shoreline boundary data, and optional hydrological and meteorological data for the area surrounding the target lake. It then analyzes the quality control bits, projection information, and temporal labels of the LST and NDVI data. Based on the quality control bits, an effective mask is constructed to remove cloud contamination, snow and ice interference, and obviously invalid pixels. Finally, the LST and NDVI data are standardized to obtain standardized LST and NDVI data. Standardization of the LST and NDVI data ensures that the buffer radius, ring width, and pixel statistical caliber have clear physical meaning.
[0034] Furthermore, the LST data is collected using MODIS MOD11A2 or an equivalent land surface temperature remote sensing product, the NDVI data is collected using MOD13A2 or a similar vegetation index product, and the DEM data is collected using SRTM or a similar digital elevation model product.
[0035] The standardization process involves resampling the LST and NDVI data to a unified target grid to obtain the resampled pixel values. The calculation formula is as follows: ; In the formula, To resample to the target grid in time phase Lower target mesh cell The pixel value at that location, To map to target mesh cells The source pixel set, For pixels, For the target mesh cell, For the original pixel set in time phase Lower target mesh cell The pixel value at that location, For variables In time phase lower pixel An effective mask.
[0036] The calculation formula is: ; In the formula, For quality control marking, For variables The set of quality control codes that are allowed to pass.
[0037] Step S2, Quarterly Synthesis and Spatiotemporal Data Cube Construction: Refer to Figure 3The pixels in the standardized LST and NDVI data from step S1 are synthesized according to a pre-defined quarterly execution time sequence to generate a quarterly composite result. The quarterly composite result is then cropped to the study area and stacked in chronological order to form a spatiotemporal data cube. The construction of the spatiotemporal data cube ensures strict consistency between LST, NDVI, and subsequent TVDI results in both the temporal and spatial dimensions, and also provides a basis for comparing different indicators with the same caliber.
[0038] Furthermore, the division of the year can be divided into four quarters: Q1, Q2, Q3, and Q4; or into two quarters: a cold season and a warm season; or into two quarters: a dry season and a rainy season.
[0039] Set the minimum number of valid observations for each quarter. When the number of effective observations for a certain pixel in a certain quarter is less than If the pixel is not detected, it will be marked as missing to prevent low-quality composite values from entering the subsequent calculation process.
[0040] The quarterly composite results are expressed as follows: ; In the formula, For target mesh cells In the quarter The synthesized value below, Belonging to the quarter The set of time phases; For target mesh cells In the quarter The number of valid observations within the range.
[0041] The spatiotemporal data cube is represented as follows: ; In the formula, For spatiotemporal data cubes, For target mesh cells The LST composite value for each quarter For target mesh cells The NDVI composite value for each quarter For target mesh cells The corresponding digital elevation model data, For target mesh cells The corresponding distance to the lake shore, For target mesh cells The slope, It is a slope.
[0042] The calculation formula is: ; In the formula, and Elevation at direction and Gradient in the direction.
[0043] Step S3, Lakeshore Distance Gradient and Topographic Zoning: Refer to Figure 4 Based on the shoreline boundary data obtained in step S1, a target buffer zone is constructed and divided into multiple ring zones. The elevation, slope, and aspect are derived from the DEM and compared with the spatiotemporal data cube. Registration was performed, and distance-topography joint analysis units were constructed based on ring zones, elevations, slopes, and aspects. These units were then divided according to year, quarter, distance ring zones, and topography.
[0044] Furthermore, the elevation and slope are discretized according to a preset interval, and the slope direction is divided into 8 or 16 directional categories.
[0045] The formula for calculating the annular band is: ; In the formula, For the first Level ring belt, For the lake shoreline, For the lake shoreline As the reference, with a radius of Cumulative buffer, For the first The radius of the ring belt, For the lake shoreline As the reference, with a radius of Cumulative buffer, for The radius of the ring belt.
[0046] The distance-terrain joint analysis unit is represented as follows: ; In the formula, For the first First ring belt, the first Level Elevation, First Grade 1 and Grade 2 Distance-topography joint analysis unit formed by overlapping slope aspects, For the first Level elevation, For the first Grade slope, For the first Slope type.
[0047] Step S4, NDVI-LST Feature Space and TVDI Inversion: Refer to Figure 5 Within the same quarter, effective pixel pairs of the standardized LST and NDVI data from step S1 are extracted to establish an NDVI-LST feature space. Binning is performed according to the NDVI step size, and wet edge percentiles and dry edge percentiles are extracted to obtain dry edge temperature samples and wet edge temperature samples. Linear fitting is then performed to calculate the temperature-vegetation drought index. ,in accordance with The formula inverts each pixel pair and obtains the result. The indicator value is constrained to 0-1.
[0048] Furthermore, according to the preset NDVI step size The LST data samples are binned, and a minimum sample size threshold is set for each bin to avoid unstable fitting of dry and wet edges due to insufficient bin samples. In each NDVI bin, the high percentile value of LST is extracted as the dry edge sample and the low percentile value is extracted as the wet edge sample.
[0049] The formula for calculating the dry edge temperature sample is as follows: ; In the formula, For the first Dry edge temperature samples corresponding to each NDVI sub-bin. For high percentile operators, Normalized Difference Vegetation Index (NDVI) For the first NDVI boxes, For the target pixel's surface temperature, For NDVI packaging.
[0050] The formula for calculating the wet edge temperature sample is as follows: ; In the formula, For the first Each NDVI sub-bin corresponds to a wet edge temperature sample. This is the low percentile operator.
[0051] The linear fit of the dry edge is expressed as: ; In the formula, The slope of the linear fit along the dry edge. The intercept is the linear fit of the dry edge.
[0052] The linear fit of the wet edge is expressed as: ; In the formula, The slope of the linear fit for the wet edge. The intercept is the linear fit of the wet edge.
[0053] The temperature and vegetation drought index The calculation formula is: ; In the formula, This is an operator used to constrain the result to the interval between 0 and 1.
[0054] Step S5, Multiscale Statistical and Trend Analysis: Refer to Figure 6 Using the time-series synthesized LST data and NDVI data obtained in step S2 and step S4 as indicators, calculate the time-series synthesized LST data and NDVI data obtained in step S2 and step S4. The distribution characteristics of the means, standard deviations, extreme values, and quantiles of each indicator within different distance-topography joint analysis units were analyzed. The Theil-Sen slope estimation and Mann-Kendall significance test were used to jointly identify the slope and significance of the indicator change trends to reduce the interference of outliers on long-term trend estimation. By sampling effective pixels, Pearson correlation coefficient analysis was performed to construct the correlation and regression relationships between topographic factors (elevation, slope) and vegetation factors (LST, NDVI, TVDI). The indicators include relative cooling indicators, relative greening indicators, relative drought reduction indicators, and topographic support items. The output includes quarterly statistical tables of LST, NDVI, and TVDI, statistical tables of lake shore distance rings, trend analysis charts, and indicator relationship charts.
[0055] Furthermore, the distance-terrain joint analysis unit includes quarterly-ring zones, quarterly-terrain categories, and quarterly-distance-terrain joint units.
[0056] The formula for calculating the mean of each indicator within any distance-topography joint analysis unit is as follows: ; In the formula, As an indicator In the quarter and joint analysis unit The mean within, For quarters The number of effective pixels within, For the first The pixel value of each valid pixel.
[0057] The Theil-Sen slope estimation method identifies the Theil-Sen trend slope of each indicator, and its calculation formula is as follows: ; In the formula, As an indicator The slope of the Theil-Sen trend, For the first The index corresponding to each observation time The pixel value, For the first The pixel value at each observation time.
[0058] The significance of the change in the indicator is identified by calculating the test statistic using the Mann-Kendall significance test. The formula for calculating the test statistic is as follows: ; In the formula, The Mann-Kendall test statistic is used. For symbolic functions, The length of the time series. For time series later than the first The sequence number of the observation value at each observation time.
[0059] Step S6, Lake Effect-Topography Coupling Identification and Results Output: Refer to Figure 7 Based on the output of step S5, a reference area is set, the mean values of LST, NDVI and TVDI are calculated, the comprehensive adjustment score is calculated, the lakeside zone is classified according to the comprehensive adjustment score, and a standardized result package of the lakeside zone classification results is output.
[0060] Furthermore, the reference area is pre-set as the outermost ring zone, the entire background zone, or a verified stable background zone according to the analysis task. The reference area is used as a background benchmark zone to provide the background mean values of indicators such as LST data, NDVI data, and TVDI at the quarterly scale, so as to calculate the relative cooling, relative greening, and relative drought reduction of each ring zone relative to the background state, thereby characterizing the ecological environment regulation effect of ring zones at different distances from the lake shore.
[0061] The formula for calculating the comprehensive adjustment score is as follows: ; In the formula, For the first The overall adjustment score of the ring belt, For the first The relative cooling index of the ring belt. For the first The weighting coefficient of the relative cooling index of the ring belt. For the first The relative greening index of the first-level ring zone, For the first The weighting coefficient of the relative greening index of the ring zone. For the first The relative drought reduction index of the first-level ring zone, No. The weighting coefficients of the relative drought reduction index of the zonal rings. For the first Topographic support items for the ring zone, The weighting coefficients for the normalized terrain support term.
[0062] in, , , and It can be preset or determined based on sample statistical results.
[0063] Furthermore, The calculation formula is: ; In the formula, For reference area in quarterly LST mean under, For the first The tertiary ring belt in the quarter LST mean under, To prevent stable terms with a denominator of zero.
[0064] The calculation formula is: ; In the formula, For reference area in quarterly The mean NDVI under the following conditions For the first The tertiary ring belt in the quarter The mean NDVI value.
[0065] The calculation formula is: ; In the formula, For reference area in quarterly TVDI mean value For the first The tertiary ring belt in the quarter The mean TVDI value.
[0066] The It is obtained by encoding and normalizing the elevation level, slope level and aspect category within the target zone according to preset rules.
[0067] Furthermore, the lakeside zone categories include lake-regulated areas, transitional areas, and topographically dominated areas, and the lakeside zone categories are represented as follows: ; In the formula, For the first Category of ring zone or joint zone, For classification with a high threshold, For classification with a low threshold, where, , and It can be preset or determined based on sample statistics.
[0068] Furthermore, threshold parameters, weight coefficients, buffer distance sequences, quarterly classification criteria, and version number information are recorded synchronously to form a standardized deliverable package that is verifiable, reproducible, and archiveable.
[0069] Example 2 A system for implementing the lakeshore distance and topography coupling spatiotemporal analysis method for lakeshore zones as described in Embodiment 1 includes a data acquisition and standardization preprocessing module, used to implement step S1, acquiring surface temperature data, vegetation index data, digital elevation model data, and shoreline boundary data of the area surrounding the target lake, and performing quality control, spatial registration, resampling, and study area clipping on the surface temperature data and vegetation index data; a quarterly synthesis and spatiotemporal data cube construction module, used to implement step S2, performing temporal synthesis on the preprocessed surface temperature data and vegetation index data, and constructing a spatiotemporal dataset under a unified spatial grid; and a lakeshore distance gradient and topographic zoning module, used to implement step S3. A multi-level lakeshore distance ring is constructed, and its elevation, slope, and aspect are calculated to form a distance-topography joint analysis unit. The NDVI-LST feature space and TVDI inversion module is used to implement step S4, constructing the NDVI-LST feature space and inverting the Temperature-Vegetation-Drought Index (TVDI). A multi-scale statistical and trend analysis module is used to implement step S5, performing hierarchical statistical and trend analysis according to time scale, lakeshore distance scale, and topographic scale. A lake effect-topography coupling identification and results output module is used to implement step S6, comprehensively identifying relative indicators and topographic support items based on pre-set reference areas, weight coefficients, and classification thresholds, and outputting ecological regulation zoning results and statistical results.
[0070] Example 3 This embodiment is an application of the lakeshore distance and topographic coupling spatiotemporal analysis method described in Embodiment 1 to the lakeshore area of Qinghai Lake. Specifically: Study Area, Data Sources, and Parameter Settings: The Qinghai Lake shoreline was used as the target study area, with a time range from January 2000 to December 2024. Input data included MODIS MOD11A2 land surface temperature products, MOD13A2 vegetation index products, SRTM DEM data, shoreline boundary data, and optional hydrological and meteorological data of Qinghai Lake from 1956 to 2020. The study area was defined as the target analysis range extending 10 km outward from the shoreline, and further divided into five concentric zones: 0-2 km, 2-4 km, 4-6 km, 6-8 km, and 8-10 km. The study area is shown below. Figure 8 As shown, where Figure 8 (a) is a schematic diagram of the plateau background, showing the location of Qinghai Lake in the Qinghai-Tibet Plateau region and the background of plateau permafrost distribution; Figure 8 (b) is a schematic diagram of the analysis range of the target lake shoreline, showing the range of the study area formed by extending 10 km outward from the shoreline and the distribution of elevation, highways and railways within the area.
[0071] In terms of parameter settings, the quarters are divided into Q1 (January to March), Q2 (April to June), Q3 (July to September), and Q4 (October to December); the uniform spatial resolution is set to 1 km; the NDVI bin step size is preferably 0.01; the wet edge percentile is preferably 0.5%, and the dry edge percentile is preferably 99.5%; the minimum number of effective observations per quarter is also specified. Set to 2 to 3 times to ensure time coverage while maintaining sample stability.
[0072] For distance-topography joint analysis, it is preferable to divide the slope aspect into eight directional categories, classify the slope at intervals of 2° or 5°, and classify the elevation at intervals of 50m or 100m. In spatial correlation analysis, approximately 20,000 samples can be randomly selected from the effective pixels to balance statistical stability and computational efficiency.
[0073] Data standardization preprocessing and spatiotemporal data cube construction: Read the LST and NDVI raw files, parse their quality control bits, projection information and temporal labels, construct an effective mask, remove cloud pollution, snow and ice interference and obvious invalid pixels, unify the projection coordinate system, spatial resolution and target grid, and clip according to the study area boundary.
[0074] For quarterly-scale analysis, effective observations of LST and NDVI are synthesized according to quarterly time sets to generate quarterly raster sequences from Q1 to Q4. Simultaneously, annual mean raster data can be calculated for long-term trend analysis. If the number of effective observations for a pixel within a quarter is less than... If the pixel is not detected, it will be marked as missing to prevent low-quality composite values from entering the subsequent calculation process.
[0075] After processing, all quarterly rasters are aligned in a unified grid, forming a spatiotemporal data cube that can be directly accessed. This step ensures strict consistency of land surface temperature, vegetation index, and subsequent TVDI results in both temporal and spatial dimensions, and also provides a basis for comparing different indicators using the same caliber.
[0076] The lake shore distance gradient and terrain zoning are constructed by generating five levels of equal-width ring zones based on a unified spatial grid, starting from the lake shoreline. Simultaneously, elevation, slope, and aspect are derived from the DEM, slope values are calculated, and then elevation, slope, and aspect category layers are formed according to preset intervals.
[0077] Under the data conditions of this embodiment, along the lake shore, from nearshore to farshore, the elevation generally increases from approximately 3250m to approximately 3468m, while the slope gradually decreases from approximately 7.2° to approximately 5.3°, thus forming a terrain structure of low-altitude steep slopes nearshore and high-altitude gentle slopes farshore. The corresponding relationship between elevation and slope changes under the distance gradient is as follows: Figure 11 As shown.
[0078] Based on the aforementioned topographic structure, a distance-topography joint analysis unit was further constructed. This unit was further divided according to year, quarter, distance ring, and topography. This joint unit provides a spatial basis for subsequent explanations of the nearshore moderating effect, the offshore topographic cold and wet effect, and the transitional relationship between the two.
[0079] NDVI-LST Feature Space Fitting and TVDI Inversion: Within each quarter, effective LST and NDVI pixels are paired and an NDVI-LST scatter plot is constructed. Preferably, bins are divided according to an NDVI step size of 0.01, and each bin is required to contain at least 30 effective samples; when the number of samples in a bin is insufficient, it can be merged with adjacent bins before percentile extraction. Dry edge samples and wet edge samples are extracted for each bin, and linear fitting of dry and wet edges is performed. After obtaining the fitting parameters, the quarterly TVDI inversion is completed, and the results are restricted to the interval between 0 and 1. The NDVI-LST feature space fitting results for a certain quarter are shown below. Figure 10 As shown.
[0080] Taking Q2 of 2000 as an example, the feature space can exhibit a dry-at-the-top, wet-at-the-bottom structure, and can fit the dry and wet edges as they vary with NDVI. The resulting TVDI results can be used to characterize the differences in dryness and wetness in the Qinghai Lake shoreline zone during different seasons and in different annular zones. The spatial distribution results of LST, NDVI, and TVDI in different seasons are shown below. Figure 9 As shown, where Figure 9 (a) to (d) represent the spatial distribution of LST for Q1 to Q4, respectively. Figure 9 (e) to (h) represent the spatial distribution of NDVI from Q1 to Q4, respectively. Figure 9 The spatial distribution of TVDI is shown in (i) to (l) for the four quarters from Q1 to Q4.
[0081] Multi-scale statistics, trend analysis, and significance testing: Statistical analysis was conducted on LST, NDVI, and TVDI by quarter, year, distance zone, and terrain category, respectively. The mean of each analysis unit was calculated. In the time trend analysis, the Theil-Sen trend slope was calculated, and the Mann-Kendall significance test was carried out to identify the long-term change direction and significance level of each indicator from 2000 to 2024.
[0082] In the statistical results of this embodiment, LST reaches its annual high in Q2 and Q3, with the average temperature in the nearshore 0-2 km zone around Q2 being approximately 23.8°C, while the temperature in the mid-to-farshore zone is approximately 25.4°C to 26.8°C. NDVI reaches its annual high in Q3 and generally increases with distance from the lake, with the average value increasing from approximately 0.48 nearshore to approximately 0.51 offshore. The quarterly variations of LST, NDVI, and TVDI under different zones are shown below. Figure 12 As shown, where Figure 12 (a) shows the quarterly variation of the mean LST with distance from the lake shore. Figure 12 (b) shows the quarterly variation of the mean NDVI with distance from the lake shore. Figure 12 (c) represents the quarterly variation of the mean TVDI with distance from the lake shore.
[0083] Regarding topographic coupling, after calculating the Pearson correlation coefficients, the correlation coefficient between elevation and LST was -0.38, the correlation coefficient between elevation and TVDI was -0.45, and the correlation coefficient between elevation and NDVI was 0.30. The quarterly time series trends and indicator relationships are as follows: Figure 13 and Figure 14 As shown, where Figure 13 (a) to (d) show the time series trend and significance test results of the mean LST for the quarters Q1 to Q4, respectively. Figure 13 (e) to (h) show the time series trend and significance test results of the mean NDVI values from Q1 to Q4, respectively. Figure 13 (i) to (l) represent the time series trend and significance test results of the mean TVDI from Q1 to Q4, respectively. Figure 14 In the middle (a), the relationship between the mean LST and the mean NDVI is shown. Figure 14 (b) shows the relationship between the mean TVDI and the mean NDVI. Figure 14 In the middle (c), the relationship between the mean TVDI and the mean LST is shown.
[0084] Lake effect-topography coupling identification and results output: After completing multi-scale statistics, the outermost ring zone or the verified stable background area is used as the reference area. The relative cooling, relative greening and relative drought reduction indices of each ring zone are calculated, and the topographic support items and the three types of ecological indicators are used to calculate the comprehensive regulation score.
[0085] Weighting coefficient , , and and classification threshold and The weighting is determined either by pre-setting or based on sample statistical results; in this embodiment, to maintain the comparability of the contributions of different indicators, a pre-set equal weighting or a standardized near-equal weighting setting is preferred. Subsequently, each annular zone or joint zone is classified to obtain the lake-dominant regulation zone, transition zone, and topographic-dominant zone.
[0086] The final output of this embodiment includes quarterly statistics on LST, NDVI, and TVDI, as well as statistics on lakeshore distance zones, topographic factors, trend analysis maps, indicator relationship maps, feature spatial maps, and lakeshore ecological regulation zoning maps. Based on these results, the nearshore core regulation zone, intermediate transition zone, and offshore topographically dominant zone can be identified, further serving lakeshore ecological monitoring, drought risk diagnosis, wetting assessment, and differentiated protection management.
[0087] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A spatiotemporal analysis method for lakeshore zones coupled with shoreline distance and topography, characterized in that, Includes the following steps: Step S1: Obtain LST data, NDVI data, DEM data, and shoreline boundary data of the area surrounding the target lake. Resample the LST data and NDVI data to a unified target grid to obtain the pixel values after resampling to the target grid. Step S2: Perform time-series synthesis on the cell values resampled to the target grid in step S1 according to the pre-divided quarters, calculate the quarterly synthesis results, and construct a spatiotemporal data cube. ; Step S3: Based on the shoreline boundary data from Step S1, construct the target buffer zone and divide it into multiple rings. Derive the elevation, slope, and aspect based on the DEM data, and compare them with... Perform registration and construct a distance-terrain joint analysis unit; Step S4: Extract effective pixel pairs from the time-series synthesized LST and NDVI data within the same quarter, establish the NDVI-LST feature space, extract dry-edge temperature samples and wet-edge temperature samples, perform linear fitting, and calculate the temperature-vegetation drought index. and will The indicator value is constrained to 0-1; Step S5, combine the time-series synthesized LST data, NDVI data, and As indicators, the mean, standard deviation, extreme values and quantiles of each indicator are calculated in various distance-topography joint analysis units. The Theil-Sen slope estimation and Mann-Kendall significance test are used to jointly identify the slope and significance of the indicator change trend. The Pearson correlation coefficient is calculated, and the output is a quarterly statistical table of LST, NDVI, and TVDI, a statistical table of lake shore distance ring, a trend analysis chart and an indicator relationship chart. Step S6: Based on the output of step S5, calculate the comprehensive adjustment score, classify the lakeside zone according to the comprehensive adjustment score, and output the standardized result package of the lakeside zone classification results.
2. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S1, the formula for calculating the pixel value after resampling to the target grid is: ; In the formula, To resample to the target grid in time phase Lower target mesh cell The pixel value at that location, To map to target mesh cells The source pixel set, For pixels, For the target mesh cell, For the original pixel set in time phase Lower target mesh cell The pixel value at that location, For variables In time phase lower pixel Effective mask; in, The calculation formula is: ; In the formula, For quality control marking, For variables The set of quality control codes that are allowed to pass.
3. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S2, the seasonal synthesis result is expressed as follows: ; In the formula, For target mesh cells In the quarter The synthesized value below, Belonging to the quarter The set of time phases, To resample to the target grid in time phase Lower target mesh cell The pixel value at that location, For target mesh cells In the quarter The number of valid observations within; The Represented as: ; In the formula, For target mesh cells The LST composite value for each quarter For target mesh cells The NDVI composite value for each quarter For target mesh cells The corresponding digital elevation model data, For target mesh cells The corresponding distance to the lake shore, For target mesh cells The slope, For target mesh cells The slope direction; in, The calculation formula is: ; In the formula, and Elevation at direction and Gradient in the direction.
4. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S2, the division of quarters can be divided into four quarters: Q1, Q2, Q3, and Q4; or into two quarters: cold and warm; or into two quarters: dry and rainy.
5. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S3, the calculation formula for the annular band is: ; In the formula, For the first Level ring belt, For the lake shoreline, For the lake shoreline As the reference, with a radius of Cumulative buffer, For the lake shoreline As the reference, with a radius of Cumulative buffer, For the first The radius of the ring belt, for The radius of the ring belt; The distance-terrain joint analysis unit is represented as follows: ; In the formula, For the first First ring belt, the first Level Elevation, First Grade 1 and Grade 2 Distance-topography joint analysis unit formed by overlapping slope aspects, For the first Level elevation, For the first Grade slope, For the first Slope type.
6. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S4, the The calculation formula is: ; In the formula, For operators used to constrain results to the interval between 0 and 1, For the linear fit of the dry edge, This is a linear fit for the wet edge. For the target pixel's surface temperature; Represented as: ; In the formula, The slope of the linear fit along the dry edge. The intercept of the linear fit along the dry edge. Normalized Difference Vegetation Index; Represented as: ; In the formula, The slope of the linear fit for the wet edge. The intercept of the linear fit at the wet edge; The formula for calculating the dry edge temperature sample is: ; In the formula, For the first Dry edge temperature samples corresponding to each NDVI sub-bin. For high percentile operators, For the first NDVI boxes, For NDVI packaging; The formula for calculating the wet edge temperature sample is: ; In the formula, For the first Each NDVI sub-bin corresponds to a wet edge temperature sample. This is the low percentile operator.
7. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S5, the formula for calculating the mean value of each indicator within any distance-terrain joint analysis unit is as follows: ; In the formula, As an indicator In the quarter and distance-terrain joint analysis unit The mean within, For quarters The number of effective pixels within, The serial number of the valid pixel. For the first The index corresponding to each valid cell ; The Theil-Sen slope estimation method identifies the Theil-Sen trend slope of each indicator, and its calculation formula is as follows: ; In the formula, As an indicator The slope of the Theil-Sen trend, For the first The index value at each observation time, For the first The index value corresponding to each observation time; The significance of the change in the indicator is identified by calculating the test statistic using the Mann-Kendall significance test. The formula for calculating the test statistic is as follows: ; In the formula, The Mann-Kendall test statistic is used. For symbolic functions, This represents the length of the time series.
8. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S6, the formula for calculating the comprehensive adjustment score is: ; In the formula, For the first The overall adjustment score of the ring belt, For the first The relative cooling index of the ring belt. For the first The weighting coefficient of the relative cooling index of the ring belt. For the first The relative greening index of the first-level ring zone, For the first The weighting coefficient of the relative greening index of the ring zone. For the first The relative drought reduction index of the first-level ring zone, No. The weighting coefficients of the relative drought reduction index of the zonal rings. For the first Topographic support items for the ring zone, These are the weighting coefficients for the normalized terrain support term; in, The calculation formula is: ; In the formula, For reference area in quarterly LST mean under, For the first The tertiary ring belt in the quarter LST mean under, To prevent stable terms with a denominator of zero; The calculation formula is: ; In the formula, For reference area in quarterly The mean NDVI under the following conditions For the first The tertiary ring belt in the quarter The mean NDVI value under the following conditions; The calculation formula is: ; In the formula, For reference area in quarterly TVDI mean value For the first The tertiary ring belt in the quarter The mean TVDI value.
9. The spatiotemporal analysis method for lakeshore zone coupled with shoreline distance and topography as described in claim 1, characterized in that, In step S6, the lakeside zone is categorized into three types: lake regulation-dominant zone, transition zone, and topographically dominant zone. The lakeside zone categories are represented as follows: ; In the formula, For the first Category of ring zone or joint zone, For classification with a high threshold, For classification with a low threshold, where, , and It can be preset or determined based on sample statistics.
10. A system for implementing the spatiotemporal analysis method of lakeshore distance coupled with topography as described in claim 1, characterized in that, The module includes: a data acquisition and standardization preprocessing module for implementing step S1; a quarterly synthesis and spatiotemporal data cube construction module for implementing step S2; a lake shore distance gradient and topographic zoning module for implementing step S3; an NDVI-LST feature space and TVDI inversion module for implementing step S4; a multi-scale statistics and trend analysis module for implementing step S5; and a lake effect-topographic coupling identification and results output module for implementing step S6.
