CN-121998439-A - Shore line health evaluation method based on space-time coupling diagnosis

CN121998439ACN 121998439 ACN121998439 ACN 121998439ACN-121998439-A

Abstract

The invention relates to the field of river shoreline health evaluation, and discloses a shoreline health evaluation method based on space-time coupling diagnosis, which comprises the steps of obtaining original data of a river shoreline state and generating a river shoreline state multisource data set; extracting a shoreline sign time sequence index based on a river shoreline state multisource data set, constructing a space-time sign data cube, extracting features from the space-time sign data cube, performing time sequence causality diagnosis, vegetation shore fixing effect space heterogeneity analysis, predicting time sequence variation trend, demarcating a risk early warning area, realizing dynamic risk early warning, generating an evaluation conclusion based on the dynamic risk early warning, and covering scoring, diagnosis, early warning and countermeasures. According to the technical scheme, the method and the device can support efficient query and coupling analysis, identify the current good shore section but risk exists in the future, realize early intervention, break through the limitation of traditional post evaluation, and finally solve the problems of low efficiency, narrow coverage, shallow analysis and weak support of industry pain points when the natural condition of the river shoreline is evaluated.

Inventors

XIAO TINGTING
CHANG LI
LI YANMENG
ZHANG QING
QIU YI
CHEN YUCHONG
LUO CHAO
Zhan Jingya
HE JUAN

Assignees

中国电建集团贵阳勘测设计研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. A shoreline health evaluation method based on space-time coupling diagnosis is characterized by comprising the following steps: The method comprises the steps of obtaining river shoreline state original data, preprocessing and calibrating the river shoreline state original data to generate a river shoreline state multisource data set, wherein the river shoreline state original data is a contrastive multisource data set generated by preprocessing and calibrating an aerospace trinity monitoring system formed by satellite remote sensing data, unmanned plane LiDAR data and ground verification data; Extracting a shoreline sign time sequence index based on the river shoreline state multisource data set, and constructing a space-time sign data cube by combining annual precipitation distance level Pt, flood season flow level Ft and human activity variable Ht, wherein the shoreline sign time sequence index comprises a hysteresis stability index and a standard annual vegetation coverage scoring value Wherein the hysteresis stability index comprises a bank slope average gradient score value Vegetation coverage score value Elevation difference scoring value of top and foot of bank slope Matrix characteristic spectral score Grade of slope toe elevation difference Generating a timing feature, expressed as: 、、、、、、、 And ; Extracting features from the space-time physical sign data cube to perform time sequence causality diagnosis and vegetation shore fixing effect space heterogeneity analysis, generating a shoreline stability causal network diagram, predicting time sequence change trend, defining a risk early warning area and realizing dynamic risk early warning; and generating an evaluation conclusion based on the dynamic risk early warning, wherein the evaluation conclusion covers scoring, diagnosis, early warning and countermeasures.
2. The shore line health evaluation method based on space-time coupling diagnosis according to claim 1, wherein the preprocessing comprises performing radiation correction and geometric correction on satellite remote sensing data, performing point cloud denoising on unmanned aerial vehicle LiDAR data, and converting the unmanned aerial vehicle LiDAR data into a digital elevation model; the calibration processing comprises the steps of realizing three-level space-time alignment of satellite remote sensing data, unmanned aerial vehicle LiDAR data and ground verification data through affine transformation, unifying coordinate systems of the satellite remote sensing data, unmanned aerial vehicle LiDAR point clouds and the ground verification data, aligning high-frequency data of the ground verification data with low-frequency data of a satellite and an unmanned aerial vehicle to the same time node, identifying abnormal values, only eliminating equipment error difference normal points, and reserving risk extreme values.
3. The shoreline health assessment method based on spatiotemporal coupling diagnosis according to claim 1, wherein the constructing a spatiotemporal sign data cube comprises the steps of: dividing a target shoreline into evaluation units, wherein each evaluation unit is provided with a corresponding number i, i=1, 2, & gt, n; Extracting standardized scores of the shoreline sign time sequence indexes one by one based on the time sequence t; extracting a shoreline sign time sequence index and a river basin environment variable; generating a time sequence characteristic of a shoreline sign time sequence index and a watershed environment variable, and calculating a shoreline stability comprehensive value ; Construction of independent variable matrix ; A spatiotemporal sign data cube is constructed.
4. The shoreline health assessment method based on space-time coupling diagnosis according to claim 3, wherein the extracting of the shoreline sign time sequence index comprises: Average grade score value of bank slope Calculating the average slope of the bank slope based on the digital elevation model, and assigning values according to guideline grading standards; The vegetation coverage score The method comprises calculating vegetation coverage by using pixel bipartite model, and calculating variation coefficient based on 3 years coverage data According to the coefficient of variation Is a value of (2) for the vegetation coverage score Assigning a value; Elevation difference scoring value of top and foot of bank slope Extracting elevation difference H assignment of the top of the bank slope and the slope toe based on a digital elevation model; matrix characteristic spectral score Extracting a matrix characteristic spectrum through hyperspectral data of satellite remote sensing data, and classifying and identifying matrix types through a support vector machine, wherein the matrix types comprise bedrock, rock soil, clay and non-clay, and assigning values based on the matrix types; Grade of slope toe elevation difference Calculating the elevation difference delta Z of the toe of the two-stage digital elevation model, and scoring according to the range of the elevation difference delta Z of the toe; Reference year vegetation coverage score And assigning a value based on the calculated reference year vegetation coverage.
5. The shoreline health assessment method based on spatiotemporal coupling diagnosis of claim 3, wherein the constructing a spatiotemporal sign data cube comprises the steps of: constructing a time sequence sign data matrix of each bank section i Expressed as: ; To all the shore sections The space-time sign data cube of the target river is formed by the space-time sign data cube and the corresponding basic information, time sequence information and quality control information; Storing the associated data of the space-time physical sign data cube into PostGIS space-time database, storing the DEM grid and satellite image data into MinIO cloud, and associating with PostGIS space-time database through file path.
6. The shoreline health assessment method based on spatiotemporal coupling diagnosis of claim 3, wherein the time-series causal relationship diagnosis comprises the steps of: generating univariate timing sequences based on hysteresis stability index for each land segment i Wherein j is a hysteresis stability index number and j=1..5; By ADF inspection model pairs ADF inspection is realized; determining a hysteresis order p; Combining the F test and Bonferroni correction, ensuring that the level of significance reaches a = 0.005; Generating a land section level causal network, which comprises defining nodes of the land section level causal network as hysteresis stability indexes, drawing directed edges among the nodes by using obvious Granges; nodes and directed edges are shown in a shoreline stability causal network graph.
7. The shoreline health assessment method based on space-time coupling diagnosis according to claim 6, wherein the vegetation-fixed shore effect spatial heterogeneity analysis is implemented by a geographic and time weighted regression model, comprising the steps of: Definition of shoreline stability Complex value Defining independent variable matrix for dependent variable All variables in the system are independent variables, and the longitude and latitude of the central point of each shore section i are calculated As a space coordinate, normalizing a time variable t as a time coordinate; Preprocessing an independent variable set; constructing a space-time weight matrix, quantifying the composite adjacency of the geographic distance and the time distance, and determining the sample weight distribution of local regression; Based on cross verification, bandwidth parameter optimization of a geographic and time weighted regression model is realized; performing iterative calculation and t-test of local regression coefficients, estimating the relation strength of space-time non-stationarity, and testing the statistical significance of local estimation, including that for each bank segment i, based on space-time weight matrix A weighted least squares regression model is constructed, expressed as: , wherein, As a function of the vector of variables, A matrix is designed for the independent variables and, As a vector of the local regression coefficients, Is a residual vector, wherein, Is the comprehensive value of the dynamic stability of the shoreline of the ith shoreside segment at the moment t, Is the local intercept coefficient of the shore section i, Is that Local coefficients of (2), Is that Local coefficients of (c.) Is that The coefficient iterative solution is expressed as: , wherein, As a local regression coefficient estimation value for shore segment i, Calculating t statistic of kth independent variable in the shore section i According to Determining significance; Outputting a vegetation fixed shore effect space distribution thermodynamic diagram based on the geographic and time weighted regression model, and predicting a time sequence change trend.
8. The method for shoreline health assessment based on spatiotemporal coupling diagnosis of claim 7, wherein the output vegetation fixed shore effect spatial distribution thermodynamic diagram comprises the steps of: for all shore sections Carrying out standardized treatment, and classifying a target river bank into a high-effect area, a medium-effect area and a low-effect area; And coloring the grading result of each bank section by taking a space vector diagram of the target river bank line as a base diagram, superposing auxiliary information, marking a key area, and realizing the space distribution thermodynamic diagram drawing of the vegetation fixed-bank effect.
9. The method for shoreline health evaluation based on space-time coupling diagnosis according to claim 7, wherein when the cross-validation is realized, a test set is constructed by sequentially constructing the dependent variable and the independent variable of each shoreside segment, the dependent variable and the independent variable of the remaining shoreside segment are used as training sets, and geography and time weighted regression models are trained based on the training sets to predict the test set Calculating a prediction error CV; wherein, the error formula defining the cross validation is as follows, , wherein, For cross-validation total error at bandwidth h, Is the measured shoreline stability integrated value of the shoreside section i, For the band width h of the lower shore section i The predicted value, n is the total number of shore segments; setting candidate ranges of the bandwidths h, calculating CV values corresponding to different h one by one, and selecting h corresponding to the smallest CV value as the optimal bandwidth.
10. The shoreline health evaluation method based on space-time coupling diagnosis according to claim 1, wherein the implementation of dynamic risk early warning comprises the steps of: Identifying dominant factors based on a shoreline stability causal network graph; Analyzing the dominant factor change trend through Mann-Kendall trend test and Sen's Slope change rate estimation; Determining a multi-level dynamic risk early warning triggering condition, and performing space plotting to realize dynamic risk early warning; And outputting an early warning report, wherein the early warning report comprises a shoreline dynamic risk early warning list, a risk early warning spatial distribution diagram and specific intervention measures.

Description

Shore line health evaluation method based on space-time coupling diagnosis Technical Field The invention relates to the field of river shoreline health evaluation, in particular to a shoreline health evaluation method based on space-time coupling diagnosis. Background The establishment of national river and lake health files and the periodic development of river and lake health evaluation are one of the core tasks of water conservancy work in new era. The natural state of the shoreline is used as a key component of the ecological system of the river and the lake to directly determine the basic state of the health of the river and the lake, and the evaluation work is a core link of the health evaluation system of the river and the lake. The core target of the river and lake health evaluation is not simply to output a grading conclusion, but accurately identifies the sign problem of the shoreline ecological system, prejudges the integral change trend of the shoreline and the river and lake, provides scientific basis for shoreline protection and repair, comprehensive river and lake treatment and normalized supervision, and promotes the continuous development of the river and lake ecological system. The current technical guidelines for river and lake health evaluation (trial run) already defines the evaluation index and calculation method of the natural condition of the shoreline, and lays a standardized foundation for river and lake health monitoring work. However, the current actual evaluation work still has obvious shortboards that each index measurement mainly depends on manual field investigation or static analysis of single remote sensing data, and is difficult to adapt to the modern requirements of dynamic supervision of river and lake. On one hand, manual surveying has the problems of high cost, large terrain limitation, limited coverage range and the like, and cannot realize the global coverage of long-distance and complex river sections, and on the other hand, the traditional evaluation period is as long as 1-2 months, so that the timeliness is seriously insufficient, and the working requirements of normalized dynamic monitoring and on-schedule completion of general survey of the name, the river and the lake are not met. In addition, the evaluation mode based on single remote sensing data improves efficiency to a certain extent, but has intrinsic limitations that firstly, slope, vegetation coverage and other shoreline indexes are treated in isolation, space-time correlation and synergy among factors are ignored, the evaluation result lacks systematicness and scientificity, secondly, analysis is carried out by relying on single time point data, dynamic evolution rules of shoreline ecology cannot be reflected, hidden problems such as gradual degradation of the shoreline are difficult to identify, thirdly, the evaluation result only reflects a grading conclusion, diagnosis and worsening risk early warning of the 'etiology' of the shoreline degradation are lacking, accurate butt joint with actual treatment requirements cannot be achieved, and long-term assessment and forward-looking accurate management decisions of the river and the lake are difficult to effectively support. Based on the above, it is urgent to construct a set of automatic, intelligent and deep shoreline natural condition evaluation systems. The system effectively solves the problems of low efficiency, narrow coverage, shallow analysis dimension, weak decision support and other pain points of the traditional shoreline evaluation, achieves the aims of multi-source data fusion, space-time dynamic analysis, accurate risk early warning and targeted treatment, promotes the river and lake shoreline evaluation to be changed from passive result output to active risk early warning, provides solid scientific support for systematic protection and repair of river and lake ecological shorelines, and comprehensively builds a foundation defense line for river and lake health. Disclosure of Invention In order to achieve the above purpose, the application provides a shoreline health evaluation method based on space-time coupling diagnosis, which comprises the following steps: The method comprises the steps of obtaining river shoreline state original data, preprocessing and calibrating the river shoreline state original data to generate a river shoreline state multisource data set, wherein the river shoreline state original data is a contrastive multisource data set generated by preprocessing and calibrating a space-sky three-in-one monitoring system formed by satellite remote sensing data, unmanned plane LiDAR data and ground verification data; Extracting a shoreline sign time sequence index based on a river shoreline state multisource data set, and constructing a space-time sign data cube by combining annual precipitation distance level Pt, flood season flow level Ft and human activity variable Ht, wherein the shoreline sign time sequence index comprises