CN115936996A - Method for improving spatial resolution of underground water reserves based on settlement feature weighted fusion - Google Patents

Abstract

The invention belongs to the technical field of hydrology, and particularly relates to a method for improving spatial resolution of underground water reserves based on settlement feature weighted fusion. The technical scheme of the invention is based on the extraction of the sedimentation characteristics and the derivation of a weight algorithm, and the novel sedimentation characteristic weighted fusion method fully considers the influence of two data characteristics of surface sedimentation and water level change on the underground water reserve change. And obtaining a brand new fusion image by weighting and fusing the underground water reserve abnormality and InSAR data inverted by the GRACE. The fusion result is more sensitive to the surface subsidence characteristics, can better show the water reserves change detail characteristics, can be more accurate for the utilization of water resource provide information service.

Description

Method for improving spatial resolution of underground water reserves based on settlement feature weighted fusion

Technical Field

The invention belongs to the technical field of hydrology, and particularly relates to a method for improving the spatial resolution of underground water reserves based on settlement feature weighted fusion.

Background

The exploitation of underground water in many arid and semiarid regions in the world exceeds natural supply, the excessive dependence of human beings on nonrenewable underground water resources can greatly reduce the underground water level, the pore water level of a weakly permeable layer and a water-bearing layer is reduced, and the stress transfer causes the compaction and deformation of cohesive soil, thereby causing the ground settlement. The problem of surface subsidence caused by the excess mining of underground water resources is increasingly prominent in China. Ground settlement may damage infrastructure such as buildings, highways, and bridges, and reduce flood discharge efficiency. In addition, inelastic settlement due to compaction of the aquifer system can cause the aquifer to permanently lose water holding capacity and affect groundwater sustainability. Rapid ground subsidence caused by groundwater super-mining is found in many cities in north china plains, such as zheng, beijing, tianjin, tezhou, cangzhou, etc. Implementing effective groundwater management measures is crucial to solving the existing water supply and demand problems.

At present, the traditional underground water monitoring scheme comprises an underground water level monitoring well and an underground water balance method. However, monitoring points are sparse, costly, time and labor intensive, and it is difficult to maintain good quality control. Therefore, it is difficult to capture the spatial details of aquifer and groundwater changes and to monitor the changes in real time without interruption. The groundwater balance method is also a common estimation method for regulatory agencies, but it ignores aquifer dynamics and local water circulation changes that occur as groundwater reserves change. In recent years, remote sensing monitoring shows huge application potential in the field of hydrology and water resources. The remote sensing monitoring is mainly to adopt optical or radar images to research underground water distribution, and a remote sensing model of underground water level distribution is established through soil layer water content. And obtaining a research area underground water burial depth distribution condition diagram by analyzing the correlation between the relative water content of the soil at different depths and the underground water burial depth through the Luo.

Synthetic Aperture interference Radar (InSAR) is increasingly used for monitoring surface subsidence and inferring hydrogeological properties due to its high spatial resolution to measure the characteristics of ground deformation. Researches show that the elastic water storage coefficient and the inelastic water storage coefficient of an aquifer can be estimated and underground water level changes can be monitored by adopting the InSAR technology to obtain time series deformation information. The InSAR has the technical advantages of high precision, high resolution, high coverage, all-time, all-weather, high automation of data processing and the like, but has certain limitation in practical application. Because the aquifer responds unevenly to pressure changes, the compacted thickness varies spatially. Even if pressure changes affect compressible aquifer structures, not all surface deformations are caused by aquifer depletion, especially in areas beyond typical InSAR detection thresholds. In recent years, a permanent scatterer (PS-InSAR) technology and a small baseline set (SBAS-InSAR) technology developed on the basis of a differential interferometry technology have been successfully applied to urban ground settlement monitoring, the phase and the amplitude of the technology are relatively stable in a longer time, the influence of factors such as space-time decoherence, atmospheric delay and noise can be effectively reduced, and the deformation monitoring precision is improved. The later added ScanSAR and TOPSAR are more helpful to provide large-area SAR images, and can draw surface deformation from cities to watershed/national scales.

In contrast to the InSAR method, GRACE gravity satellites can directly measure ground water reserve abnormalities (GWSA) in conjunction with hydrological models, assuming that other components (e.g., soil water, surface water, and snow water equivalents) have been subtracted from the total terrestrial water reserve. However, GRACE has a low spatial resolution (0.25 DEG to 3 DEG), and is suitable for an area of 100000km ² In the above areas, the typical Grace resolution is not sufficient to support small area-scale groundwater remediation. There are also many new ways to estimate the downscaling of water reserve anomalies for GRACE. GRACE inversion improvement algorithm shows that considering spatial variability of crust quality under GRACE resolution has important significance for better estimation of groundwater reservoir variation. The ground settlement rate and the settlement amount are closely related to the underground water level descending rate and the descending amount, so that the GRACE data can be subjected to scale reduction processing by utilizing ground monitoring data. The Castellazzi study finds that using InSAR data focused GWSA signals of GRACE inversion as a spatial prior map of groundwater mass loss distribution can provide quantitative, high-resolution groundwater depletion information. However, previous studies have only stayed on conceptual introduction and have not proposed specific and effective fusion algorithms and case studies.

Unlike existing research, GRACE is one of the means to effectively detect water reserve changes, the coarse resolution limits its application potential in small areas, and the phenomenon of inconsistency with InSAR and official ground water bit data occurs.

Disclosure of Invention

In order to overcome the defects in the prior art, the invention provides a method for improving the spatial resolution of underground water reserves based on settlement feature weighted fusion. By means of weighted fusion of underground water reserve abnormality and InSAR data inverted by GRACE, a brand new fusion image is obtained, and water reserve change detail characteristics are better represented.

The invention discloses a method for improving the spatial resolution of underground water reserves based on settlement characteristic weighted fusion, which performs spatial downscaling treatment on GRACE by a settlement characteristic weighted fusion method, and comprises the following steps:

s1, collecting data of regional underground water level monitoring wells, and calculating time sequence change trends of individual wells and change rate spatial distribution of the whole region;

s2, downloading Sentinel-1A images, high-precision DEM and POD precision orbit data, and detecting the spatial and temporal evolution of regional surface deformation by adopting a multi-temporal InSAR method;

s3, acquiring a JPL Mascon data set and a CSR Mascon data set, and extracting land water reserve abnormity from the two GRACE time sequence data;

and S4, detecting the surface deformation distribution of the underground water exhausted area and the underground water reserve abnormal value inverted by GRACE data by combining an InSAR technology.

Preferably, in step S2, the InSAR processing procedure includes determining that the primary and secondary images perform differential interference, terrain phase removal, phase unwrapping, phase filtering, and geocoding.

Preferably, the main and auxiliary images adopt 10 × 2 multi-view processing to suppress noise; after generating a phase interference image and an amplitude intensity image, removing a terrain phase from the interference phase by using DEM; and selecting a building with high coherence and stable scattering characteristics as a scattering point, and performing phase unwrapping by adopting a three-dimensional phase unwrapping algorithm.

Preferably, a Goldstein pruning method is adopted for phase filtering processing, and track errors and atmospheric phase interference are eliminated by subtracting a linear phase slope of each interferogram; and obtaining the sight line deformation time sequence of each pixel through least square inversion, and geocoding the result.

Preferably, in step S3, when the GRACE inverts the abnormal underground water reserve, the equivalent of soil water, surface water, vegetation water and snow water needs to be removed.

Preferably, at the same resolution, the groundwater reserves are subtracted from the land water reserves to obtain a groundwater reserves abnormality map expressed by the following formula:

GWSA＝TWSA-SMSA-SWSA

wherein GWSA is underground water reserve abnormality, TWSA is total land water reserve abnormality, SMSA is soil water reserve abnormality, and SWSA is surface water reserve abnormality.

Preferably, GRACE inverted land and ground water reservoir changes and NOAH inverted soil and surface water changes require interpolation and range leveling processing.

Preferably, in step S4, the GRACE and the weights are combined into new values based on a novel GRACE/InSAR sedimentation feature weighted fusion method, so as to improve the resolution of the GRACE signal.

Preferably, in step S1, more than 24 consecutive months are selected as a time reference interval, and a GPS time sequence provided by the CMONOC is acquired for seasonal item change analysis.

The invention provides a novel settlement characteristic weighted fusion method, which is based on the extraction of settlement characteristics and the derivation of a weight algorithm, and fully considers the influence of surface settlement and water level change data characteristics on the change of underground water reserves.

And obtaining a brand new fusion image by weighting and fusing the underground water reserve abnormal data and InSAR data inverted by the GRACE. The fusion result is more sensitive to the surface subsidence characteristics, can better represent the water reserve change detail characteristics, enables the spatial resolution of the groundwater reserve abnormality to be improved by ten times (from 0.5 degrees multiplied by 0.5 degrees to 0.05 degrees multiplied by 0.05 degrees), and can provide information service for the utilization of water resources more accurately. The technical scheme of the invention is suitable for popularization and application in the related technical field.

Drawings

In order to more clearly explain the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that for those skilled in the art, other related drawings can be obtained from these drawings without inventive effort.

FIG. 1 is a geographic overview of the Beijing area;

FIG. 2 is a schematic flow diagram of the process of the present invention;

FIG. 3 is a diagram showing water level change rate in Beijing City, (a) water level change rate in 2005-2011, (b) water level change rate in 2012-2014, and (c) water level change rate in 2015-2018;

FIG. 4 is a time series diagram of the average groundwater level of Beijing plain in 2005-2018 and the water level change rate in different periods;

FIG. 5 is a schematic representation of the groundwater levels of 10 down trend monitoring wells in 2015-2018;

fig. 6 is a schematic diagram of deformation quantities of beijing plain in 5 months in 2017 to 7 months in 2020 of InSAR inversion, (a) a vector scattergram, A, B, C, D are four sedimentation funnels, and (b) a 0.05 ° × 0.05 ° grid graph;

fig. 7 is a graph showing the cumulative settling volume and average settling rate for four settling hoppers, (a) the cumulative settling volume, (b) the average settling rate;

FIG. 8 is a schematic diagram of a sinker grid extracted based on InSAR results;

fig. 9 is a time series comparison schematic of InSAR, GRACE, CLSM and Δ GWL between 2017-2020, (a) - (d) are raw time series comparison results at monitor wells P17, P66, P68 and P86, respectively, (e) - (h) are correlations of InSAR settlement trend terms for four monitor wells with CLSM and Δ GWL trend terms, respectively;

FIG. 10 is a graph showing the annual trend of each component when the underground water reserves are abnormal in the GRACE/GLDAS inversion in 2017-2020, (a) 0.5 ° JPL-TWSA, (b) 0.25 ℃ SR-TWSA, (c) 0.25 ° NOAH-SMSA, (d) 0.25 ° NOAH-SWSA, (e) - (h) are the results of interpolation of (a) - (d) to 0.05 ° respectively;

figure 11 is a GWSA schematic based on the novel sedimentation characteristic weighted fusion method,

wherein, (a) JPL-GWSA, (b) CSR-GWSA, (c) CLSM-GWSA, (d) JPL/InSAR fusion inversion GWSA, (e) CSR/InSAR fusion inversion GWSA, and (f) CLSM/InSAR fusion inversion GWSA;

figure 12 is a schematic diagram of error estimation of the GWSA trend for the fusion inversion GWSA and the groundwater level estimate;

figure 13 is a graph showing the relationship between GWSA, precipitation and precipitation, wherein (a) the average values of the precipitation and the area of the GWSA for beijing plain are compared, (b) the average value of the InSAR precipitation for beijing plain north is compared with the GPS precipitation at the site BJSH, and (c) the seasonal composition of the GWSA, precipitation and precipitation after decomposition;

FIG. 14 is a schematic representation of a time series settlement comparison of individual grids near a BJSH site and the average of northern plains;

FIG. 15 is a schematic diagram of Google Earth showing the topography of the terrain at the settling funnel.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.

To make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments obtained by a person of ordinary skill in the art without any inventive work based on the embodiments in the present application are within the scope of protection of the present application.

The present invention is based on Beijing area as an example, and as shown in FIG. 1, geographical location 115 ° 30 'E-117 ° 30' E,39 ° 00 'N-41 ° 00' N is located at North China plain West North China in Beijing City. Wherein, the Beijing plain has 13 areas including urban eight areas and dense cloud and gentle Tongzhou Pinggu partial areas, and the Beijing belongs to a typical warm-zone semi-humid continental monsoon climate, is cold and dry in winter, and is high in temperature and rainy in summer. The rainfall is extremely unevenly distributed in time and space, the rainfall in mountainous areas is larger than that in plain, the annual rainfall is concentrated in 6-9 months and accounts for 60-85% of the annual rainfall. In addition, the annual change of precipitation is large, and continuous drought or water-rich years often occur. Five continuous drought years have been encountered since 1999, the contradiction between water resource supply and demand is prominent, groundwater is continuously overflowed for guaranteeing water supply, and the average burial depth of groundwater in the plain area is reduced from 11.9m at the end of 1998 to 24.3m at the end of 2012. About two thirds of water supply in Beijing comes from underground water, and the ground settlement of the longterm over-mining-induced area of sixty years is in a rapid increasing trend.

The Beijing has complex terrain and is high in northwest and low in southeast. The Beijing urban area is located on the impulsive accumulation plains formed by rivers such as the Yongding river and the Wen elm river and inclined towards the southeast. As a alluvial plain, the compressible mass of the aquifer forms an overburden whose thickness increases from west (tens of meters) to east (hundreds of meters). The regions with greater sedimentation gradient are located at the quaternary depressed boundary region. According to the deposition rule of the fourth system, the deposition era, the structure of the underground aquifer, the current situation of exploitation and utilization of underground water and other factors, the aquifer in the Beijing area can be divided into four strata. The first water-containing layer group is mainly diving and micro confined water, and the buried depth is less than 50m; the second aquifer is a pressure bearing area buried at 80-120m deep; the third aquifer is a pressure bearing area buried at 150-180m deep; the fourth aquifer is a pressure bearing area buried deep at 300 m. Wherein, the lithology of the diving area is mainly silt, and the lithology of the pressure bearing area is silt, clay, fine sand and coarse sand. The primary pumping layer of industrial and domestic water is the third confined aquifer, which is widely distributed throughout plains. There is concern that groundwater overstrike in combination with compressible aquifers causes extremely severe settlement. The second and third aquitards present elastoplasticity along with the change of the water level of the confined water of the middle and deep layers and have creep characteristics, and are main contribution layers of ground settlement in Beijing area.

From the beginning of the 50-70 th age of the 20 th century, the groundwater in Beijing city belongs to the initial development stage, the mining and supplementing are basically balanced, the super mining only occurs in the local area of suburbs in the city, and the ground settlement only occurs in the super mining area. In 1961-1970, the increase of underground water exploitation amount in suburbs of cities is particularly obvious. The exploitation degree of underground water in Beijing city is obviously improved from the middle of 70 s to the beginning of 80 s in the 20 th century, and the exploitation amount reaches the 20 th centuryTwice as early as the first 60 s. By 1978, the groundwater in suburbs of cities was severely overstrained, and the groundwater storage capacity decreased by 12.42X 10 in ten years ⁸ m ³ . The underground water exploitation amount in the suburb counties is greatly increased, but the serious excess exploitation degree is not reached. In the 20 th century, the early 80 s to the end 90 s, the underground water mining amount in Beijing is relatively stable, the underground water level is in a slow descending state, the underground water super-mining area is gradually expanded from the early stage mainly concentrated in the suburbs to the suburbs, and the area of the super-mining area accounts for 70% of the area of the plain area.

However, compression of groundwater mining layers is the primary cause of surface subsidence, with the range of the subsidence funnels gradually expanding from suburbs to cis-and Changping areas. In 10 years before the 21 st century, the underground water funnels are mainly distributed in regions from Huanggang and Changsheng in sunny areas to Mizhuang in Shuizhijie areas, and the distribution condition of the underground water funnels is basically consistent with the region with serious ground settlement. By 2010, the area with the ground settlement accumulation amount of more than 50mm in Beijing plain area reaches 4288km ² And expands at a rate of 30-60 mm/yr, seriously threatening urban safety. In recent years, although the government strictly limits the exploitation amount of underground water, and the south-to-north water central line regulation project gradually recovers the underground water level in Beijing city by means of ecological water supplement, irrigation water infiltration and the like, the phenomenon of over-mining still exists in part of regions. When facing problems, the system monitors groundwater in a large range in detail, and can help people to know water resource conditions and reasonably exploit groundwater.

As shown in fig. 2, the method for improving spatial resolution of underground water reserves based on sedimentation characteristic weighted fusion according to the present invention performs spatial downscaling processing on a GRACE by a sedimentation characteristic weighted fusion method, including:

s1, collecting data of regional underground water level monitoring wells, and calculating time sequence change trends of individual wells and change rate spatial distribution of the whole region;

s2, downloading Sentinel-1A images, high-precision DEM and POD precision orbit data, and detecting the spatial and temporal evolution of regional surface deformation by adopting a multi-temporal InSAR method;

s3, acquiring a JPL Mascon data set and a CSR Mascon data set, and extracting land water reserve abnormity from the two GRACE time sequence data;

and S4, detecting the surface deformation distribution of the underground water exhausted area and the underground water reserve abnormal value inverted by GRACE data by combining an InSAR technology.

And deducing a statistical normalization algorithm according to Max-Min normalization, using the statistical normalization algorithm to take the deformation result as the weight of spatial priori knowledge, and combining the GRACE and the weight into a new value based on a novel GRACE/InSAR settlement characteristic weighting fusion method, thereby improving the resolution of the GRACE signal. 2005-2018 was chosen as the study period in view of the quality and availability of the data, as well as the actual situation of the study area. Finally, the simulation precision is evaluated by comparing the downscaling result with the absolute error index of the measured value, and the data processing flow is shown in fig. 3.

Fig. 3 and 4 show a space distribution diagram of the change rate of the water level of the beijing plain and a time series curve of the average buried depth of the region respectively. FIGS. 3 (a) - (c) show the groundwater level change rates for three periods of time, 2005-2011, 2012-2014 and 2015-2018, respectively. As can be seen from FIG. 3, the groundwater level is mostly in a descending trend in 2005-2011 due to continuous super mining of groundwater; in 2012-2014, due to the increase of rainfall, the underground water level gradually tends to be stable; after 2015, water starts to flow through the central line engineering for south-to-north water diversion, the phenomenon of shortage of water resources of Beijing is relieved, the underground water level is obviously increased, and the change rate of most of the water levels realizes the conversion from negative to positive. This transition can also be found from the area mean burial depth of fig. 4, and the mean burial depth trends for the three periods are-0.91 m/yr, 0.16m/yr, and 0.03m/yr, respectively, as obtained by linear fitting. However, the trend of the change of partial monitoring wells is continuously reduced, and a total of 22 monitoring wells is found. The 10 monitoring wells with descending trends were selected to show the time series change of the groundwater level for the third period, as shown in fig. 5.

FIG. 6 (a) shows the cumulative deformation of Beijing plain between 2017 and 2020. For subsequent fusion inversion, the dot-plot is converted to a 0.05 ° resolution grid-plot, as shown in fig. 6 (b). It can be seen that most of the area is relatively stable with no significant sedimentation. In terms of spatial distribution, the Yucrouchan pool is taken as a reference point, and the settlement areas are mainly distributed in a junction area between a sunny area and a Tongzhou area and a junction area between a Changping area and a Huishi area to form four settlement funnels. From the time distribution, the accumulated deformation amount is calculated by taking the main image as a reference value (5 months and 20 days in 2017), the accumulated deformation range at the four funnels is-183.70 mm-92.63 mm (fig. 7 (a)), and the maximum annual average deformation rate reaches-53.09 mm/yr (fig. 7 (b)). This is consistent with the distribution of ground subsidence in the prior studies in Beijing. Most of previous researches are carried out by utilizing ENVISAT data to extract Beijing deformation in 2007-2010 time period, wherein the deformation speed reaches 150mm/yr. The ground subsidence in Beijing area is mainly caused by groundwater overstraining, and through groundwater replenishment and strict groundwater mining policies in recent years, the ground subsidence in most urban areas has been gradually relieved in 2017-2020.

Fig. 8 shows 258 settlement grids for InSAR estimation with empty areas as elevated areas. And in the sedimentation area, four descending monitoring wells with good data quality, namely P17, P66, P68 and P86 are used for analyzing the underground water level and the surface sedimentation, and the four monitoring wells are all positioned in the sedimentation funnel area. FIGS. 9 (a) - (d) are the comparison of groundwater level change, JPL-GWSA, CSR-GWSA, CLSM-GWSA and InSAR settlement at the four selected sites, with more consistent trends. Where the GWSA of the JPL and CSR inversion is the average change per year and the time series range is 2018.6-2020.12. From the results of comparing Δ GWL and InSAR settlement, it is seen that the surface settlement lags behind the change in groundwater level by about 1 to 6 months. In order to reasonably analyze the correlation between the two, the method needs to preprocess the Δ GWL, and takes 5 months in 2017 as a reference, subtracts a reference value on the basis of the original data, decomposes a trend item by using an STL method, and then compares the trend item. Fig. 9 (e) - (h) show the correlation of InSAR settlement at four monitoring wells with CLSM-GWSA and Δ GWL, respectively, with correlation coefficients greater than 0.85. In the period of 2017-2020, the settlement of the four monitoring wells has better consistency with the change trend of the underground water level. The underground water level is reduced, and the change rate of the accumulated settlement amount is increased; when the groundwater level rises back, the rate of change of the corresponding settlement amount is also slowed down.

FIG. 10 shows the annual trend plots of the variables in the 2017-2020 GRACE inversion of GWSA, including JPL-TWSA, CSR-TWSA, NOAH-SMSA, and NOAH-SWSA. Fig. 10 (a) - (d) are spatial trend plots at the original resolution for four variables, 0.25 ° × 0.25 °, respectively, except that JPL is 0.5 ° × 0.5 °. Fig. 10 (e) - (h) are graphs showing the trend of four variables interpolated to 0.05 ° × 0.05 ° by cubic spline interpolation, respectively. The whole space distribution before and after interpolation is basically consistent, so the invention does not consider the error brought by interpolation. It was found from fig. 10 (a) and 10 (b) that the land water reserves estimated by the two GRACE data are not uniformly spatially distributed, which may be due to the fact that the two solutions use different striping strategies, and the striping residuals have different effects on them. The GRACE spherical harmonic coefficient truncation and the de-banding and Gaussian smoothing filtering process cause leakage errors of the GRACE inversion result. Currently the most appropriate GRACE solution for the study area is not yet determined, so JPL, CSR and CLSM data will be used for GWSA inversion for comparative analysis.

The ground settlement rate and the settlement amount are closely related to the groundwater level descending rate and the groundwater level descending amount. According to the study of scholars in cities such as Cangzhou, texas and Gallery, the empirical relationship between ground settlement and groundwater level change generally presents an exponential, quadratic or linear correlation relationship. According to the method, the InSAR settlement result is used as priori knowledge, the underground water reserve change of Beijing plain is reconstructed based on a novel settlement characteristic weighted fusion method, and the spatial resolution of the underground water reserve inverted by GRACE is improved. Fig. 11 (a) and (b) show trends in GWSA of 0.5 ° × 0.5 ° in 2018-2020 obtained by inversion of two GRACE data based on the equation for ground water equilibrium. Fig. 11 (c) shows the CLSM-GWSA 0.5 ° × 0.5 ° trend, time span and SAR image matching between 2017-2020. Fig. 11 (d) - (f) are the results for three GWSAs down-scales of 0.05 ° x 0.05 ° based on the novel weighted fusion method for sedimentation characteristics, respectively. Compared with the original GWSA, the spatial resolution of the merged GWSA is greatly improved, and the requirement of a water resource management mechanism can be met.

The surface subsidence weighted mass distribution provides a better groundwater reserve change map, and shows that InSAR measurement not only provides the spatial position of groundwater reserve loss occurrence, but also provides important quantitative information. Furthermore, despite differences in lithology (aquifer restriction, thickness and compressibility) within the study area, there is an important correlation between groundwater depletion rate and settling rate. The different data sources all presented similar inversion results (fig. 11 (d) - (f)), demonstrating the potential for combining the two. The novel GRACE/InSAR sedimentation characteristic weighted fusion inversion method provided by the invention is applied to future generations of GRACE Mascon tasks, and is expected to improve the spatial resolution.

As shown in fig. 12, it can be observed from the error distribution rule that most regions have small errors, and below 10mm/yr, the most error places are located at the northeast corner and the west. The GWSA signals for these regions cannot be interpreted by InSAR derived mass distribution maps, indicating that the northeast and western subsidence signals cannot be completely attributed to groundwater reserves changes, possibly where the deformation is greatly influenced by natural hydrological processes. And after removing the northeast corner and the west part, carrying out spatial correlation analysis on the remaining areas, and finding that the estimation result of the novel settlement characteristic weighted fusion method is more consistent with that of the underground water level monitoring well.

The error between the fusion result and the measured data can be explained in four aspects. First, GRACE-GWSA does not fully reflect regional groundwater depletion due to leakage error of GRACE. Second, in the original GWSA trend plot of 0.25 °, 36% of the pixels are not calculated ((406-258)/406 ≈ 0.36), indicating that the aquifer of some grids does not sag, or compaction is performed at a rate below the InSAR detection threshold. Third, GRACE and official groundwater balance differ in the mechanism of measuring groundwater depletion. GRACE provides information about the mass loss associated with groundwater, including all dynamic effects within the aquifer, but it is contaminated by spatial leakage inherent to gravitational field resolution. Fourth, the dynamic effects of aquifer depletion, wastewater recharge, and water supply distribution system infiltration are neglected by official groundwater budgets. It is worth noting that deep confined water head changes have a greater impact on ground settlement than diving and shallow confined water. Particularly, the water level changes of the aquifers which are adjacent up and down at the deepest part of the monitoring well and have the largest contribution to settlement are main factors for forming ground settlement in the area, and the underground water level belongs to unpressurized diving data.

Figure 13 shows a comparison of GWSA, settlement, rainfall data in terms of long term trends and seasonal changes. The area average and plain north average from the entire beijing plain were analyzed (plain divided into three parts, north, middle and south). FIG. 13 (a) compares the Beijing plain average of the three types of data, and finds that the long-term trends of GRACE-GWSA and InSAR settlement are more consistent and in a descending trend; and the CLSM-GWSA is more consistent with the GPS settlement, and has no obvious trend and is more stable. The average settlement rate in 2015-2020 was higher than CLSM-GWSA, probably because the CLSM hydrological model structure lacks human active components and does not reflect the positive effect of north-south water switch on surface water. The difference between InSAR and GPS is that the settlement of the whole Beijing area cannot be accurately represented because the number of Beijing GPS sites is small. Fig. 13 (b) compares the average InSAR settlement for north of beijing plain with the GPS settlement at site BJSH. It can be seen that the long-term trend and the settlement magnitude of the InSAR time series and the GPS vertical displacement are in agreement on the area average, which also indicates that InSAR can capture high-precision deformation. Figure 13 (c) shows the seasonal signature after the trend removal term was used and the seasonal composition showed (1) significant seasonal oscillations in GWSA, settling and rainfall, but the amplitude of the oscillations was very different, with the rainfall range-62.22-87.27mm, the GWSA range-9.69-13.14 mm, and the settling floated within ± 3 mm. (2) The settling generally lags behind GWSA, a time difference that is likely due to the low vertical conductivity of the aquifer system and slow consolidation of the soil in wet weather. (3) Sedimentation is inversely correlated with rainfall, with a correlation coefficient of about 0.45 absolute, and less correlated with GWSA (0.06), indicating that shaking is most likely due to rainfall changes.

Because the water reserves change and the sedimentation time sequence are damaged by noise, certain errors exist in the estimation of the fusion result and the decomposition of the seasonal term. It can also be seen from the correlation coefficient that the seasonal term is much less correlated than the long-term trend term. Therefore, the invention averages the time series of the deformed pixels near the BJSH site to reduce random noise. Fig. 14 shows a time-series settling comparison of a single grid near the BJSH site and the plain north InSAR mean, finding that the mean varies similarly with the time series of adjacent grids, except that the range of fluctuation of the mean is small, and thus, estimation of the plain scale is feasible.

Figure 15 google earth shows that there are many vegetable greenhouses, factories and residential buildings on the settling funnel site, indicating that human activities such as agricultural production may also be one of the causes of settling. In summary, groundwater mining, human activities, rainfall and hydrogeological conditions work together on surface deformation.

The technical scheme of the invention is based on the extraction of sedimentation characteristics and the derivation of a weight algorithm, and the novel sedimentation characteristic weighted fusion method fully considers the influence of two data characteristics of surface sedimentation and water level change on the underground water reserve change. And obtaining a brand new fusion image by weighting and fusing the underground water reserve abnormality and InSAR data inverted by the GRACE. The fusion result is more sensitive to the surface subsidence characteristics, can better show the water reserves change detail characteristics, can be more accurate provide information service for the utilization of water resource.

Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the embodiments. Thus, the present embodiments are not intended to be limited to the embodiments shown herein but are to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention shall be included in the scope of the present invention.

Claims (9)

1. The method for improving the spatial resolution of underground water reserves based on settlement feature weighted fusion is characterized in that the method for carrying out spatial downscaling treatment on GRACE by using the settlement feature weighted fusion method comprises the following steps:

s1, collecting data of regional underground water level monitoring wells, and calculating time series change trends of individual wells and change rate spatial distribution of the whole region;

s2, downloading Sentinel-1A images, high-precision DEM and POD precision orbit data, and detecting the spatial and temporal evolution of regional surface deformation by adopting a multi-temporal InSAR method;

s3, acquiring JPL Mascon and CSR Mascon data sets, and extracting land water reserve abnormality from the two GRACE time series data;

and S4, detecting the surface deformation distribution of the underground water exhausted area and the underground water reserve abnormal value inverted by GRACE data by combining an InSAR technology.

2. The method for improving the spatial resolution of the underground water reserves based on the weighted fusion of the subsidence features as claimed in claim 1, wherein in the step S2, the InSAR processing procedure comprises determining the main and auxiliary images to perform differential interference, terrain phase removal, phase unwrapping, phase filtering and geocoding.

3. The method for improving spatial resolution of groundwater reserves based on weighted fusion of sedimentation features as claimed in claim 2, wherein the primary and secondary images are processed with 10 x 2 multiview to suppress noise; after generating a phase interference image and an amplitude intensity image, removing a terrain phase from the interference phase by using DEM; and selecting a building with high coherence and stable scattering characteristics as a scattering point, and performing phase unwrapping by adopting a three-dimensional phase unwrapping algorithm.

4. The method for improving spatial resolution of groundwater reserves based on weighted fusion of sedimentation features as claimed in claim 2, wherein Goldstein pruning is used for phase filtering to eliminate orbit error and atmospheric phase interference by subtracting linear phase slopes of each interferogram; and obtaining the sight line deformation time sequence of each pixel through least square inversion, and geocoding the result.

5. The method for improving the spatial resolution of the underground water reserves based on the weighted fusion of the settlement features as claimed in claim 1, wherein in the step S3, when the GRACE inverts the abnormal underground water reserves, the equivalent of soil water, surface water, vegetation water and snow water needs to be removed.

6. A method for improving spatial resolution of groundwater reserves based on weighted fusion of sedimentation features according to claim 5, wherein the groundwater reserve abnormality map is obtained by subtracting the soil water and the surface water from the land water reserve at the same resolution and is expressed by the following formula:

GWSA＝TWSA-SMSA-SWSA

GWSA is the underground water reserve abnormality, TWSA is the total land water reserve abnormality, SMSA is the soil water reserve abnormality, and SWSA is the surface water reserve abnormality.

7. The method of improving spatial resolution of groundwater reserves based on subsidence feature weighted fusion of claim 5, wherein GRACE inverted land and groundwater reserve changes and NOAH inverted soil and surface water changes require interpolation and leveling processing.

8. The method for improving spatial resolution of groundwater reserves based on sedimentation feature weighted fusion as claimed in claim 1, wherein in step S4, the resolution of the GRACE signal is improved by combining the GRACE and the weights into new values based on a novel GRACE/InSAR sedimentation feature weighted fusion method.

9. The method for improving the spatial resolution of underground water reserves based on the weighted fusion of the settlement features as claimed in claim 1, wherein in the step S1, more than 24 consecutive months are selected as time reference intervals to obtain a GPS time series provided by CMONOC for seasonal item variation analysis.

Images