Drought phase discrimination method combining GNSS water storage inversion and water budget constraint

Abstract

This invention discloses a drought stage discrimination method that integrates GNSS water storage inversion and water budget constraints. Addressing the problems of existing drought monitoring indicators being singular and unable to characterize the lag effects of meteorological deficits propagating to water storage deficits, as well as the impact of anthropogenic disturbances, this invention acquires precipitation, evapotranspiration, and GNSS vertical displacement sequences; inverts changes in terrestrial water storage and constructs anomaly sequences; constructs meteorological-side drought intensification / recovery and water storage deficit / recovery; sets candidate cumulative time scales and response lag sets, searches for the optimal match within a sliding window, calculates the drought initiation / recovery consistency factor and water storage mismatch residuals; and, based on dynamic closed-loop constraints, discriminates drought precursor, initiation, development, duration, mitigation, and recovery stages, and identifies natural meteorological-dominated, anthropogenically intensified, and water-buffered droughts based on mismatch residuals. This invention achieves refined stage discrimination and type identification of the drought propagation process, improving the accuracy and physical interpretability of drought monitoring.

Description

Technical Field

[0001] This invention belongs to the field of GNSS hydrogeometry, specifically involving a drought stage discrimination method that integrates GNSS water storage inversion and water budget constraints. Background Technology

[0002] Drought is a type of natural disaster caused by the combined effects of multiple hydrological and meteorological factors. It is characterized by slow formation, long duration, wide impact, complex propagation chains, and delayed recovery. Drought is usually not the result of a single moment or a single hydrological element anomaly, but rather a dynamic process in which meteorological water deficit gradually spreads to soil water, vegetation water, surface water, groundwater, and the overall regional terrestrial water storage deficit. Characterizing the process of drought propagation from meteorological water deficit to regional water storage deficit and its phased changes is one of the challenges facing modern drought monitoring. Therefore, accurately identifying the onset, development, duration, mitigation, and recovery stages of drought is of great significance for agricultural production management, water resource allocation, ecological environmental protection, groundwater extraction control, and drought relief and disaster reduction decision-making.

[0003] Currently, traditional drought monitoring and assessment methods mainly include standardized precipitation indices, standardized precipitation evapotranspiration indices, and Palmer drought indices constructed based on meteorological observation data, as well as comprehensive drought indices constructed based on remote sensing information such as vegetation indices, surface temperature, and soil moisture. These methods can characterize the occurrence and evolution of drought from the perspectives of precipitation deficit, atmospheric evapotranspiration, vegetation physiological response, and near-surface water status, and have been widely used in regional drought monitoring. However, existing methods still have certain limitations in characterizing the drought propagation process and the overall water storage response. For example, meteorological drought indices based on precipitation or precipitation-potential evapotranspiration relationships mainly reflect insufficient atmospheric water input and abnormal evapotranspiration, focusing on revealing the water surplus / deficit state under meteorological forcing conditions, and are difficult to directly characterize the actual degree of deficit and dynamic changes in regional terrestrial water storage. Remote sensing drought monitoring methods based on vegetation status, surface temperature, or soil moisture can reflect the response of the vegetation growing layer or near-surface moisture to drought. However, their monitoring results are easily affected by factors such as land cover type, vegetation phenology, crop growth stage, remote sensing inversion uncertainties, and surface heterogeneity. Their ability to comprehensively characterize changes in deep soil water, groundwater, surface water, and other water storage components is limited. Therefore, existing drought monitoring indicators often fail to fully reveal the process of meteorological water deficit spreading to regional comprehensive terrestrial water storage deficit, and also fail to adequately characterize the differences in response timescales, cumulative effects, and stage lag relationships among different hydrological elements. Consequently, the identification of key stages such as drought onset, development, mitigation, and recovery is not accurate enough.

[0004] With the development of space geodesy technology, the Global Navigation Satellite System (GNSS) can achieve all-weather, all-time, and millimeter-level precision surface displacement measurement. Based on the elastic load theory, using GNSS reference station vertical displacement to retrieve Terrestrial Water Storage Change (TWSC) has gradually become an important technical means for monitoring regional hydrological changes. Compared with traditional meteorological drought indices and some remote sensing drought indicators, GNSS-retrieved TWSC can comprehensively reflect the overall changes of multiple water storage components such as soil water, groundwater, and surface water, and has advantages such as continuous observation, high temporal resolution, and sensitivity to comprehensive water storage changes. However, existing drought monitoring methods based on GNSS water storage retrieval still have certain shortcomings. On the one hand, some methods mainly standardize the GNSS-retrieved TWSC, without fully integrating meteorological water balance information such as precipitation and actual evapotranspiration, making it difficult to explain the formation mechanism of water storage deficits. On the other hand, existing methods often compare or correlate TWSC with meteorological drought indices, assuming that meteorological water deficit and water storage response occur on the same time scale, lacking explicit characterization of cumulative time scale differences, response lags, and stage transitions during drought propagation. Furthermore, under the influence of anthropogenic disturbances such as irrigation water intake, groundwater extraction, reservoir regulation, or inter-basin water transfer, TWSC may not be entirely consistent with precipitation and evapotranspiration anomalies. Relying solely on GNSS water storage anomalies or a single drought index can easily lead to misjudgments of drought onset, relief, or drought type.

[0005] Therefore, it is necessary to propose a drought stage discrimination method that can simultaneously reflect the water balance mechanism, response time scale differences, and disturbance mismatch information, to accurately distinguish the drought initiation, development, duration, mitigation, and recovery stages, and further identify drought types such as natural meteorological-dominated, anthropogenic water consumption-enhanced, and regulation-buffered drought types, thereby improving the reliability and physical interpretability of drought monitoring results. Summary of the Invention

[0006] This invention aims to provide a drought stage identification method that integrates meteorological water deficit, terrestrial water storage response, and the impact of anthropogenic disturbances. Addressing the shortcomings of existing drought monitoring indicators, such as their limited coverage and the difficulty in characterizing response lags, temporal scale differences, and anthropogenic disturbance identification in GNSS-based water storage monitoring, this invention constructs a drought process identification framework that includes GNSS terrestrial water storage inversion, multi-timescale water budget anomaly characterization, meteorological forcing and water storage response lag matching, dynamic closure constraints, and water storage mismatch residual identification. Specifically, this invention uses GNSS vertical displacement inversion to retrieve regional land water storage changes, combines precipitation and actual evapotranspiration data to construct multi-timescale water balance anomalies, and establishes directional consistency constraints, temporal consistency constraints, and amplitude consistency constraints by using the time scale and response lag between meteorological drought signals and meteorological recovery signals and water storage responses. This enables precise differentiation of drought initiation, development, duration, mitigation, and recovery stages, and further identifies natural meteorological-driven drought, anthropogenic water consumption-intensified drought, and water storage-buffered drought, thereby improving the accuracy and physical interpretability of drought stage differentiation.

[0007] According to one aspect of the present invention, a method for identifying drought stages by integrating GNSS water storage inversion and water budget constraints is provided, comprising: acquiring precipitation time series, actual evapotranspiration time series, and vertical displacement time series of GNSS observation stations in the study area; calculating and standardizing the cumulative water budget at each cumulative time scale based on the precipitation time series and actual evapotranspiration time series to obtain a standardized water budget index; constructing meteorological-side drought reduction and meteorological-side recovery based on the standardized water budget index; inverting the land water storage change sequence based on the vertical displacement time series, and constructing a land water storage anomaly sequence based on the land water storage change sequence; and so on. Based on the terrestrial water storage anomaly sequence, construct water storage deficit and water storage recovery; for drought-causing and recovery directions, respectively set candidate cumulative time scale sets and candidate response lag sets, and search for the optimal cumulative time scale and optimal response lag with the highest matching degree based on the terrestrial water storage anomaly sequence within a sliding window; based on the optimal cumulative time scale and optimal response lag, calculate drought-causing consistency factor and recovery consistency factor, as well as water storage mismatch residual; based on the meteorological drought-causing amount, meteorological recovery amount, water storage deficit, water storage recovery amount, drought-causing consistency factor, recovery consistency factor, and water storage mismatch residual, determine the drought stage and identify the drought type.

[0008] As a further technical solution, within a sliding window, the optimal cumulative time scale and optimal response lag with the highest matching degree are searched based on the anomaly sequence of terrestrial water storage. This includes: for each set of parameter combinations in the candidate set of cumulative time scales and the candidate set of response lags, calculating a matching score within the sliding window, wherein the matching score is obtained by weighted summation of relevant consistency terms and directional consistency terms; selecting the set of parameter combinations with the largest matching score as the optimal cumulative time scale and optimal response lag for the current sliding window; when multiple parameter combinations correspond to the same maximum value, selecting the combination with the smallest cumulative time scale; if the cumulative time scales are the same, selecting the combination with the smallest response lag.

[0009] As a further technical solution, the correlation consistency term is the larger of the correlation coefficient between the meteorological water balance cumulative anomaly sequence and the terrestrial water storage anomaly sequence within the sliding window and zero; the directional consistency term is the average of the products of the adjacent period differences of the meteorological water balance cumulative anomaly sequence and the adjacent period differences of the terrestrial water storage anomaly sequence within the sliding window.

[0010] As a further technical solution, the drought-causing consistency factor and the recovery consistency factor are respectively based on the drought-causing direction and the recovery direction, and are composed of the geometric average of the consistency factor of each free direction, the amplitude closure factor, the hysteresis stability factor and the persistence factor.

[0011] As a further technical solution, the water storage mismatch residual is calculated by the difference between the product of the meteorological-side drought amount or meteorological-side recovery amount and the dynamic conversion coefficient and the abnormal change in terrestrial water storage amount. The dynamic conversion coefficient is obtained by least squares estimation of the meteorological-side drought amount or meteorological-side recovery amount and the abnormal change in terrestrial water storage amount.

[0012] As a further technical solution, the identification of drought type includes: identifying it as anthropogenic water consumption-enhanced drought when the water storage mismatch residual is less than the negative mismatch threshold, identifying it as a storage and buffering type drought when the water storage mismatch residual is greater than the positive mismatch threshold, and identifying it as a natural meteorological-dominated drought when the water storage mismatch residual is between the negative mismatch threshold and the positive mismatch threshold.

[0013] As a further technical solution, constructing meteorological-side drought reduction and meteorological-side recovery includes: calculating the cumulative amount of the difference between precipitation anomalies and actual evapotranspiration anomalies over a preset cumulative time scale; standardizing the cumulative amount by dividing it by the standard deviation of the corresponding calendar month to obtain a standardized water balance index; taking the larger of the negative value and zero of the standardized water balance index as the meteorological-side drought reduction, and taking the larger of the standardized water balance index and zero as the meteorological-side recovery.

[0014] As a further technical solution, the drought stage determination includes: when the meteorological drought level exceeds the precursor threshold and the water storage deficit does not reach the drought initiation threshold, or the drought consistency factor is lower than the threshold, it is determined to be a drought precursor stage; when both the meteorological drought level and the water storage deficit exceed the drought initiation threshold and the drought consistency factor meets the closed threshold, it is determined to be a drought initiation stage; when the water storage deficit expands continuously for multiple periods and the drought consistency factor continuously meets the closed threshold, it is determined to be a drought development stage; when the meteorological recovery level is lower than the persistence threshold, the water storage deficit is higher than the persistence threshold, and the recovery consistency factor is lower than the recovery threshold, it is determined to be a drought persistence stage; when both the meteorological recovery level and the water storage recovery level exceed the mitigation threshold and the recovery consistency factor meets the mitigation threshold, it is determined to be a drought mitigation stage; when the water storage deficit falls back to the normal level and remains there for multiple periods, and the recovery consistency factor remains stable, it is determined to be a drought recovery stage.

[0015] As a further technical solution, the method also includes constructing a water storage decline rate and a water storage recovery rate based on the abnormal sequence of terrestrial water storage; when identifying a drought stage, the water storage decline rate and the water storage recovery rate serve as auxiliary references for trend confirmation and reverse suppression: when the water storage deficit is within a preset range and the water storage decline rate is continuously positive or exceeds the decline rate threshold, the confirmation of the drought initiation stage and development stage is enhanced; when the meteorological recovery amount increases and the water storage recovery rate is continuously positive or exceeds the recovery rate threshold, the confirmation of the drought mitigation stage and recovery stage is enhanced.

[0016] According to one aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the drought stage discrimination method that integrates GNSS water storage inversion and water budget constraints.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Improved the completeness of drought monitoring targets: By jointly analyzing precipitation, actual evapotranspiration and GNSS-inverted TWSC, it can characterize both meteorological water deficit or recovery and water storage deficit and recovery response, thus comprehensively depicting the regional integrated drought status.

[0018] 2. Enhanced physical constraints for drought stage identification: Based on the water balance relationship between precipitation-actual evapotranspiration anomalies and TWSC retrieved by GNSS, dynamic closed constraints are constructed, which provides clear water balance basis for drought stage identification and improves the physical interpretability of the results.

[0019] 3. Characterizing the time scale differences and lag response during drought propagation: By setting a set of candidate cumulative time scales and a set of candidate response lags, the optimal cumulative time scale and optimal response lag are adaptively searched within a sliding window, which can reflect the cumulative effect and lag response in the process of meteorological water deficit propagating to terrestrial water storage.

[0020] 4. Reduced risk of misjudgment during drought phases: By constructing drought-causing consistency factors, recovery consistency factors, and water storage mismatch residuals, the directional consistency, amplitude closure, hysteresis stability, and persistence between meteorological forcing and water storage response can be comprehensively judged, which can improve the accuracy and stability of identifying drought initiation, development, duration, mitigation, and recovery phases. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating the drought stage discrimination method integrating GNSS water storage inversion and water budget constraints provided in this embodiment of the invention; Figure 2 A schematic diagram of the dynamic physical closed-loop constraint between meteorological forcing and water storage side response provided in an embodiment of the present invention; Figure 3 A flowchart for drought stage discrimination based on dynamic closure constraints provided in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating drought type discrimination based on water storage mismatch, provided as an embodiment of the present invention. Detailed Implementation

[0023] This invention describes a novel method for drought stage discrimination and type identification based on GNSS vertical displacement inversion of terrestrial water storage changes and the construction of dynamic closed-loop constraints for water budget by combining precipitation and actual evapotranspiration. This method utilizes GNSS-inverted terrestrial water storage changes to constrain water deficit and recovery responses on the water storage side during drought processes. It proposes a method to construct multi-timescale budget anomalies using precipitation and actual evapotranspiration to characterize meteorological water forcing. This method not only alleviates the shortcomings of traditional drought monitoring methods in characterizing the propagation process from meteorological water deficit to terrestrial water storage deficit, but also improves the accuracy and physical interpretability of drought initiation, development, duration, mitigation, and recovery stages. It can be effectively applied to regional drought evolution monitoring, stage discrimination, recovery assessment, and drought type identification, belonging to the field of GNSS hydrogeometry.

[0024] The technical solution of the present invention includes the following steps: Step 1: Data acquisition and inversion of terrestrial water storage.

[0025] After obtaining the precipitation time series, actual evapotranspiration time series, and vertical displacement time series of GNSS observation stations in the study area, the data were first spatially matched and the time scale was unified to ensure data consistency. Then, outlier removal, missing data completion, trend term separation, and seasonal component standardization were performed to reduce the influence of various non-hydrological load signals. Finally, the TWSC sequence was obtained by inversion based on the processed vertical displacement time series.

[0026] Step 2: Meteorological anomaly construction.

[0027] For a pre-defined set of cumulative time scales, the cumulative water balance between precipitation and actual evapotranspiration at different time scales is calculated and standardized to obtain a multi-time-scale water balance anomaly sequence that reflects the water surplus and deficit status. Based on this, meteorological anomaly indicators are constructed: meteorological drought amount, used to characterize the driving intensity of meteorological conditions on the occurrence and development of drought; and meteorological recovery amount, used to characterize the meteorological recovery process caused by enhanced precipitation replenishment or weakened evapotranspiration.

[0028] Step 3: Constructing anomalies on the water storage side.

[0029] Based on the aforementioned TWSC sequence, the degree of deviation from the average state for the same period over many years was calculated, and anomaly indicators for terrestrial water storage were constructed: water storage deficit, used to characterize the overall deficit degree of regional terrestrial water storage; water storage recovery, used to characterize the degree of recovery of regional comprehensive water storage under the influence of precipitation replenishment and hydrological processes; and water storage decline rate and water storage recovery rate, used to characterize the dynamic changes in the aggravation or recovery of water storage deficit.

[0030] Step 4: Constructing dynamic closed-loop constraints for water volume revenue and expenditure.

[0031] Candidate cumulative timescale sets and candidate response lag sets are set up for drought-causing and recovery directions respectively to describe the cumulative and lag effects in the transmission process from meteorological water volume anomalies to terrestrial water storage anomalies. The optimal cumulative timescale and optimal response lag that best match the TWSC are searched within a sliding window. Drought-causing consistency factors and recovery consistency factors are constructed based on directional consistency, amplitude closure, lag stability, and persistence to characterize the degree of directional consistency between meteorological anomalies and water storage anomalies. Water storage mismatch residuals are calculated to identify deviations caused by insufficient water budget closure, groundwater regulation anomalies, or human activity disturbances.

[0032] Step 5: Determining the drought stage and type based on closed constraints.

[0033] Based on meteorological drought amount, water storage deficit, water storage recovery amount, water storage decline rate, water storage recovery rate, drought induction consistency factor, and recovery consistency factor, a state transition discrimination rule for drought initiation, development, duration, mitigation, and recovery is established, and the drought stage and severity are output. Based on the characteristic combination of meteorological anomalies, water storage anomalies, and mismatch residuals during the drought process, the drought type is further distinguished as a storage and buffer type, a natural meteorological-dominated type of drought, and an anthropogenic water consumption-intensified type of drought.

[0034] The terms “comprising” and “having”, and any variations thereof, in the specification, claims, and accompanying drawings of this invention are intended to cover a non-exclusive inclusion, such as a process, method, system, product, or apparatus that includes a series of steps or units, not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. In addition, the technical features of the various embodiments or individual embodiments provided by the present invention can be arbitrarily combined to form new technical solutions. Such combinations are not bound by the order of steps and / or structural composition patterns, but must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0036] See Figure 1 The drought stage discrimination method that integrates GNSS water storage inversion and water budget constraints provided in this embodiment of the invention has the following specific steps.

[0037] Step 1: First, collect raw observation data from GNSS stations within the study area, and download PPP (Precise Point Positioning) error correction products such as precise ephemeris and precise clock bias for the corresponding time period. Perform precise point positioning calculations on the raw GNSS observation data to obtain the vertical displacement time series of each GNSS station. Subsequently, the vertical displacement time series was subjected to gross error removal, step correction, missing measurement interpolation, non-tidal ocean load correction, non-tidal atmospheric load correction, thermal expansion correction, and trend term separation to obtain the GNSS vertical displacement series that mainly reflects hydrological load changes. Then, the regional TWSC series was retrieved based on the Green's function constrained by the Laplace matrix. : (1), in, The coefficient matrix of the Green's function. Let be the vector of terrestrial water storage to be estimated; This is the GNSS vertical hydrological load displacement observation vector; It is a smoothing factor; The matrix is ​​Laplace smoothed. Then, contemporaneous precipitation time series for the study area are collected. and actual evapotranspiration time series Precipitation and actual evapotranspiration data were unified to a monthly scale and the spatial range of the study area; high temporal resolution data were aggregated at the monthly scale, and data with different spatial resolutions were resampled or regionally averaged. Subsequently, quality control measures such as outlier identification and correction, and missing data completion were implemented to obtain monthly precipitation and actual evapotranspiration sequences.

[0038] Finally, the multi-year contemporaneous climate averages of TWSC, precipitation, and actual evaporation were calculated, and the corresponding multi-year averages were subtracted from each series to obtain the monthly-scale anomalous time series: water storage anomaly. Abnormal precipitation and abnormal evaporation .in: (2), (3), (4), In the above formula Indicates a time sequence number; Indicates time period The corresponding calendar month number; , , These represent the sequences of terrestrial water storage, rainfall, and actual evapotranspiration, respectively. , , These represent the multi-year average terrestrial water storage, precipitation, and actual evapotranspiration, respectively. The resulting anomalous sequences, after the above processing, reflect the degree of deviation from normal climate conditions and serve as input for subsequent drought discriminant analysis.

[0039] Step 2: Obtain precipitation anomalies in the study area and actual evaporation anomalies Subsequently, a meteorological-side water forcing is constructed based on the principle of water balance. Regional TWSCs are typically influenced by precipitation input, evapotranspiration, runoff output, and lateral water exchange. However, data on runoff, artificial water intake, and water storage often suffer from gaps or spatial discontinuities at the regional scale. Since this invention does not directly replace GNSS-based TWSC retrieval with meteorological water balance, but rather aims to establish a dynamic closed-loop constraint between meteorological forcing and GNSS water storage response, the difference between precipitation and actual evapotranspiration is used to characterize the primary meteorological-side forcing driving the TWSC under conditions of incomplete data.

[0040] For the preset cumulative time scale Calculate the cumulative amount of water revenue and expenditure The water surplus / deficit status of a region at different cumulative time scales is quantified. The calculation formula is as follows: (5).

[0041] When precipitation is below average for a prolonged period and actual evapotranspiration is above average... A trend towards lower or even negative values ​​indicates that meteorological conditions are conducive to drought; when precipitation is above average or actual evapotranspiration decreases, An increase indicates that the meteorological side is trending towards recovery. To eliminate dimensional differences across different calendar months and cumulative time scales, measurements were performed at each time scale. Standardization is performed to obtain the standardized water quantity budget index. : (6), in, and These represent the mean and standard deviation of the cumulative water volume and expenditure for the corresponding calendar month and the corresponding cumulative time scale within the base period, respectively. Indicates time period The corresponding calendar month number is used for calculation of all samples from the same calendar month within the base period. and .

[0042] To unify the directionality of subsequent stages of judgment, the amount of aridification on the meteorological side is further defined. and meteorological recovery amount The calculation formula is as follows: (7), (8).

[0043] After the above treatment , and These values ​​are used to characterize meteorological water balance anomalies, drought-causing directional forcing, and recovery directional forcing at different cumulative time scales. The larger the value, the stronger the driving force in the current direction, providing a basis for subsequently determining the time scale and response lag that best match the water storage-side response.

[0044] Step 3: Based on the terrestrial water storage anomaly sequence obtained in Step 1, calculate the terrestrial water storage anomaly index: (9), in, and These represent the multi-year average terrestrial water storage and standard deviation for the corresponding calendar month within the baseline period, respectively. Based on this, the water storage deficit is further calculated. Water storage recovery rate of decline in water storage and the rate of recovery of water storage : (10) (11), (12), (13) in, Used to reflect the degree of loss in the region's overall water storage. Used to characterize the degree of water storage recovery and These are used to characterize the dynamic intensity of water storage decline and recovery, respectively. Since terrestrial water storage comprehensively reflects changes in soil water, shallow groundwater, and other water storage elements, it is more effective in characterizing the overall latent drought state of a region compared to indicators that only focus on single soil moisture or single reservoir storage; it also incorporates... and It helps to identify the key turning points in the spread and recovery process of drought.

[0045] Step 4: After obtaining multi-timescale water budget anomaly indicators and water storage-side indicators, a dynamic physical closed-loop constraint is constructed to measure the degree of matching between meteorological forcing and water storage-side response. The working principle of the dynamic physical closed-loop constraint between meteorological forcing and water storage-side response is detailed in the appendix. Figure 2 .

[0046] For both drought-causing and recovery directions, candidate cumulative timescale sets and candidate response lag sets are established respectively: (14) in, Each scale in Indicates the current time As the endpoint, accumulate backwards. Accumulated window of abnormal meteorological water volume and expenditure in a time unit; Each lag in This indicates that the meteorological anomaly lags behind the water storage response sequence in the comparison time. One time unit Indicates a synchronous response. This indicates a lag in the response of water storage to meteorological forcing. To improve search stability, the sliding window length is set to... Set the lower limit of valid samples to Only if the number of valid samples in the current window is not less than Parameter search is only performed at certain times.

[0047] Under ideal closed conditions, TWSC is primarily driven by the difference between precipitation input and evapotranspiration output. Considering that factors such as surface runoff leakage, groundwater recharge and discharge, human water intake and storage projects can cause a mismatch between meteorological forcing and water storage changes, this invention no longer simply assumes strict synchronization between the two. Instead, it allows different cumulative time scales and response lags for different watersheds and different stages. (In candidate parameter combinations...) A sliding window search is performed. For any sliding window, a matching score is calculated. It is a weighted score of correlation coefficient and directional consistency.

[0048] (15) in For relevant consistency items, This is the direction consistency function.

[0049] (16) (17) in This corresponds to the cumulative anomalies in the meteorological water balance within the sliding window. To and The matching water storage-side response sequence is composed of terrestrial water storage anomaly sequences; and Representing the sliding window respectively and The average value. The optimal scale and optimal lag are defined for the drought-causing direction, respectively. And the optimal scale and optimal hysteresis in the recovery direction It is obtained by maximizing the consistency score within a sliding window, i.e.: (18).

[0050] If multiple combinations correspond to the same optimal value, the one with the smaller cumulative time scale is preferred; if they are still the same, the one with the smaller lag is preferred. For the optimal meteorological forced sequence selected in the current stage... Introducing dynamic conversion coefficient Used to characterize the local transformation intensity of the water storage-side response to the meteorological forcing term. Click the sliding window The samples within are estimated using least squares estimation as follows: (19) in, Indicates the first The changes in terrestrial water storage anomalies retrieved by GNSS over a given period. Based on the relevant dynamic conversion coefficient. Further construct water storage mismatch residual : (20).

[0051] Water storage mismatch residual This characterizes the portion of water storage changes that cannot be explained by meteorological anomalies under the current optimal cumulative timescale and optimal response lag conditions. A significantly negative value indicates that the decline in water storage is greater than the decline that can be explained by meteorological forcing, and may indicate groundwater over-extraction, increased irrigation water withdrawal, or other additional water consumption; when When the value is significantly positive, it indicates that the decline in actual water storage has been significantly buffered, possibly due to processes such as reservoir regulation and external water inflow compensation.

[0052] To further characterize the degree of closure consistency in different directions, drought-causing consistency factors were constructed respectively. and recovery consistency factor Both are composed of a directional consistency factor, an amplitude closure factor, a hysteresis stability factor, and a persistence factor, and are calculated using a geometric mean. (twenty one), (twenty two), in, Used to determine whether the direction of meteorological forcing is consistent with the direction of change in water storage. Used to determine the degree of similarity between the measured change in water storage and the expected change in water storage. Used to evaluate the stability of the optimal lag within a continuous window. This is used to evaluate whether the consistent relationship between meteorological forcing and water storage response persists across multiple periods. Therefore, and It is no longer used as a weight for uniformly scaling a single composite index, but rather as a gating factor for whether migration is permitted during drought phases: It is mainly used to confirm whether the development of drought to the level of drought has been established. It is mainly used to confirm whether relief and recovery have been achieved. This is used to identify drought types such as anthropogenic water consumption enhancement or water storage buffering. When the meteorological forcing and the water storage response are in the same direction, have similar amplitudes, and a stable lag relationship, the directional consistency factor in the corresponding direction... The value is relatively large; when the two are in opposite directions, have significant amplitude deviations, or have unstable lag relationships, the corresponding consistency factor is... The value decreases.

[0053] Step 5: After obtaining , , , , , , , and Subsequently, a drought stage state transition rule based on dynamic closed constraints was established to comprehensively judge whether a region is experiencing drought, what stage of evolution it is in, and whether it has entered a mitigation or recovery process.

[0054] The determination is made using a combination of staged conditions, closed-loop gating, and mismatch criteria. First, closed-loop gating conditions for drought-causing and recovery directions are constructed: (twenty three), in, For indicator functions, and These are the closure discrimination thresholds for the drought-causing direction and the recovery direction, respectively.

[0055] When the meteorological drought level has increased, but the water storage deficit has not yet reached the drought initiation confirmation criteria, or the drought consistency factor has not yet met the closed-gate requirements, it is judged to be in the drought precursor stage: (twenty four), (25) in, The threshold for meteorological drought reduction during the early stages of drought is used to determine whether the meteorological water deficit has intensified. This indicates the upper limit threshold of water storage deficit during the early stages of drought, used to determine whether the water storage deficit has not yet reached the level required to confirm the onset of drought. The drought consistency factor threshold represents the drought precursor stage and is used to determine whether the closed consistency between the meteorological drought signal and the water storage response has not yet reached the requirements for confirming drought.

[0056] When both meteorological drought level and water storage deficit exceed the drought initiation threshold, and the drought consistency factor meets the closed-loop gate condition, the drought is considered to have entered the initial stage. (26) The threshold for meteorological drought initiation is used to confirm that meteorological drought forcing has reached the drought initiation level. This indicates the threshold for water storage deficit at the onset of a drought, used to confirm that the water storage deficit has reached the level required to initiate a drought. The drought consistency factor gating threshold represents the initial stage of drought and is used to confirm that there is sufficient directional consistency, amplitude closure and hysteresis stability between meteorological drought forcing and water storage deficit response.

[0057] When water storage deficit is continuous The period is expanding, and the drought-causing consistency factors are increasing. When the closed-gate condition is continuously met, it is determined to be in the drought development stage, among which... The continuous discrimination period indicates the length of continuous time used to confirm the continued expansion of water storage deficit. When the meteorological drought level begins to decrease, but the water storage deficit remains high, and the recovery consistency factor has not yet met the recovery threshold, it is determined to be a drought persistence phase. (27) The upper limit threshold for meteorological recovery during the drought duration is used to determine whether the meteorological recovery signal is still weak. The threshold for water storage deficit during a prolonged drought is used to determine whether the water storage deficit remains at a high level. The recovery consistency factor threshold, representing the drought duration phase, is used to determine whether a stable closure relationship has not yet been formed between the meteorological recovery signal and the water storage recovery response. When the meteorological recovery increases, the water storage recovery shows an upward trend, and the recovery consistency factor meets the mitigation threshold condition, the drought is considered to be in a mitigation phase. (28) The meteorological recovery threshold represents the drought mitigation phase and is used to confirm that the recovery signal formed by increased precipitation supply or reduced actual evapotranspiration has met the mitigation criteria. The threshold for the recovery of water storage during the drought relief phase is used to confirm that terrestrial water storage has substantially recovered. This represents the recovery consistency factor gating threshold during drought mitigation, used to confirm sufficient directional consistency, amplitude closure, and hysteresis stability between the meteorological recovery signal and the water storage recovery response. When the water storage deficit returns to normal levels and remains thereafter... When the drought recovery period has lasted for more than a certain period and the recovery consistency factor remains stable, it is determined to be in the drought recovery phase. See the attached flowchart for the drought phase identification process. Figure 3 .

[0058] During the aforementioned state transition, the rate of decrease in water storage... and the rate of recovery of water storage As an auxiliary reference for trend confirmation and reverse suppression: when the water storage deficit approaches or reaches a threshold, if the rate of decline in water storage is continuously positive or exceeds a preset decline rate threshold, the confirmation of the drought's onset and development stages is strengthened; if the decline rate weakens or turns to rebound, the development stage is not directly confirmed, and the process moves to the candidate discrimination of persistence or mitigation. When meteorological recovery increases, if the rate of recovery in water storage is continuously positive or exceeds a preset recovery rate threshold, the confirmation of mitigation and recovery stages is strengthened; if the recovery rate is insufficient or turns to decline again, it is only considered a precursor to mitigation, and recovery is not confirmed. Therefore, the rate of change in water storage is used to assist in judging whether the stage transition has a sustained water storage response, avoiding misjudgment of the stage due to abnormal fluctuations in a single month.

[0059] After completing the stage assessment, based on the water storage deficit... Historical quantile intervals, duration of phases, and mismatched residuals Determine the severity and type of drought.

[0060] (29) in, This represents the historical percentile threshold for water storage deficit during the baseline period. Further, it is combined with the duration of the phase. Adjust the level: (30).

[0061] Finally, combining the mismatch residuals Identify drought types: (31), in, and These represent the negative and positive mismatch thresholds, respectively. See the attached diagram for a detailed illustration of drought type discrimination. Figure 4 .

[0062] Based on the above steps, this invention can comprehensively obtain meteorological water deficit and recovery information, GNSS-inverted terrestrial water storage anomaly response information, the dynamic closed-loop relationship between meteorological forcing and water storage dependence, and the discrimination results of drought stage and type. This method can achieve continuous identification of the entire process of drought initiation, development, duration, mitigation, and recovery, reduce the risk of misjudging drought stages, and further output drought type, thereby improving the accuracy, stability, and physical interpretability of regional drought monitoring and discrimination results.

[0063] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it implements all or part of the steps of the drought stage discrimination method integrating GNSS water storage inversion and water budget constraints provided by this invention.

[0064] The computer-readable storage medium may be any medium capable of storing data, including but not limited to: read-only memory (ROM), random access memory (RAM), disk (such as hard disk, floppy disk), optical disk (such as CD-ROM, DVD), solid-state drive (SSD), USB flash drive, flash memory card, or other optical storage, magnetic storage, semiconductor storage devices, or any combination thereof.

[0065] In one specific implementation, the computer program code corresponding to steps 1 to 5 described in the above embodiments is pre-stored in a solid-state drive or optical disc. When drought stage discrimination is required, the storage medium is connected to a computer system or server, and the processor reads and executes the computer program, which can automatically complete the entire process of data acquisition, meteorological anomaly construction, water storage anomaly construction, dynamic closure constraint construction, and drought stage discrimination and type identification.

[0066] In another embodiment, the computer-readable storage medium is built into a separate computing device that receives external GNSS vertical displacement time series, precipitation time series and actual evapotranspiration time series through a data interface. After the processor executes the program in the storage medium, it directly outputs the drought stage discrimination result and drought type.

[0067] The computer-readable storage medium embodiments of the present invention are based on the same technical concept as the foregoing method embodiments, and have the same technical features and effects, which will not be repeated here.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method of distinguishing drought stages by fusing GNSS water storage inversion and water budget constraints, characterized in that, include: Obtain the time series of precipitation, actual evapotranspiration, and vertical displacement of GNSS observation stations in the study area; Based on the precipitation time series and the actual evapotranspiration time series, the cumulative water balance at each cumulative time scale is calculated and standardized to obtain the standardized water balance index. Based on the standardized water balance index, the meteorological drought amount and the meteorological recovery amount are constructed. The land water storage change sequence is obtained by inverting the vertical displacement time series, and the land water storage anomaly sequence is constructed based on the land water storage change sequence. The water storage deficit and water storage recovery are constructed based on the land water storage anomaly sequence. For drought-causing and recovery directions, candidate cumulative time scale sets and candidate response lag sets are set respectively. Within a sliding window, the optimal cumulative time scale and optimal response lag with the highest matching degree are searched based on the land water storage anomaly sequence. Based on the optimal cumulative time scale and optimal response lag, the drought-causing consistency factor and the recovery consistency factor, as well as the water storage mismatch residual, are calculated. Based on the meteorological drought level, meteorological recovery level, water storage deficit, water storage recovery level, drought consistency factor, recovery consistency factor, and water storage mismatch residual, the drought stage is determined and the drought type is identified.

2. The method according to claim 1, wherein, Within a sliding window, the search for the optimal cumulative time scale and optimal response lag that best match the anomaly sequence of terrestrial water storage includes: For each set of parameters in the candidate cumulative time scale set and the candidate response lag set, a matching score is calculated within the sliding window. The matching score is obtained by a weighted sum of relevant consistency terms and directional consistency terms. The parameter combination with the highest matching score is selected as the optimal cumulative time scale and optimal response lag for the current sliding window; When multiple parameter combinations correspond to the same maximum value, the combination with the smallest cumulative time scale is selected; if the cumulative time scales are the same, the combination with the smallest response lag is selected.

3. The method according to claim 2, wherein, The correlation consistency term is the larger of the correlation coefficient between the cumulative anomaly sequence of meteorological water balance and the anomaly sequence of terrestrial water storage within the sliding window and zero; the directional consistency term is the average of the products of the differences between adjacent periods of the cumulative anomaly sequence of meteorological water balance and the anomaly sequence of terrestrial water storage within the sliding window.

4. The method of claim 1, wherein the method is characterized by: The drought-causing consistency factor and recovery consistency factor are respectively for the drought-causing direction and the recovery direction, and are composed of the geometric average of the consistency factor of each free direction, the amplitude closure factor, the hysteresis stability factor and the persistence factor.

5. The method of claim 1, wherein the method is characterized by: The water storage mismatch residual is calculated by the difference between the product of the meteorological-side drought amount or meteorological-side recovery amount and the dynamic conversion coefficient and the abnormal change in terrestrial water storage. The dynamic conversion coefficient is obtained by least squares estimation of the meteorological-side drought amount or meteorological-side recovery amount and the abnormal change in terrestrial water storage.

6. The method of claim 1, wherein the method is characterized by: The drought type identification includes: identifying anthropogenic water consumption-enhanced drought when the water storage mismatch residual is less than the negative mismatch threshold, identifying storage and buffering drought when the water storage mismatch residual is greater than the positive mismatch threshold, and identifying natural meteorological-dominated drought when the water storage mismatch residual is between the negative and positive mismatch thresholds.

7. The method of claim 1, wherein the method is characterized by: The construction of meteorological-side drought reduction and meteorological-side recovery includes: Calculate the cumulative amount of the difference between precipitation anomalies and actual evapotranspiration anomalies over a preset cumulative time scale; The cumulative amount is standardized by dividing it by the standard deviation of the corresponding calendar month to obtain the standardized water quantity expenditure index. The larger of the negative value and zero of the standardized water balance index is taken as the meteorological drought amount, and the larger of the standardized water balance index and zero is taken as the meteorological recovery amount.

8. The method of claim 1, wherein the method is characterized by: Determining the stage of drought includes: When the amount of drought on the meteorological side exceeds the precursor threshold and the water storage deficit does not reach the drought initiation threshold, or when the drought consistency factor is lower than the threshold, it is determined to be a drought precursor stage. When both the amount of drought on the meteorological side and the amount of water storage deficit exceed the drought initiation threshold and the drought consistency factor meets the closed gate control condition, it is determined to be the drought initiation stage. When the water storage deficit expands continuously over multiple periods and the drought consistency factor continuously meets the closed gate condition, it is determined to be in the drought development stage. When the meteorological recovery amount is lower than the duration threshold, the water storage deficit is higher than the duration threshold, and the recovery consistency factor is lower than the recovery threshold, it is determined to be a drought duration stage. When both the meteorological recovery amount and the water storage recovery amount exceed the mitigation threshold and the recovery consistency factor meets the mitigation threshold condition, it is determined to be in the drought mitigation stage; When the water storage deficit returns to normal levels and remains there for several consecutive periods, and the recovery consistency factor remains stable, it is determined to be in the drought recovery phase.

9. The drought stage discrimination method integrating GNSS water storage inversion and water budget constraints as described in claim 8, characterized in that, The method further includes constructing a rate of decline and a rate of recovery of water storage based on the anomaly sequence of terrestrial water storage; when identifying drought stages, the rate of decline and the rate of recovery of water storage serve as auxiliary references for trend confirmation and reverse suppression. When the water storage deficit is within a preset range and the rate of decline of the water storage is continuously positive or exceeds the rate of decline threshold, the identification of the drought initiation and development stages is enhanced. When the meteorological recovery rate increases and the water storage recovery rate is continuously positive or exceeds the recovery rate threshold, the confirmation of the drought mitigation phase and the recovery phase is enhanced.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, it implements the drought stage discrimination method that integrates GNSS water storage inversion and water budget constraints as described in any one of claims 1 to 9.

Images