skip to main content


Title: Downscaled‐GRACE Data Reveal Anthropogenic and Climate‐Induced Water Storage Decline Across the Indus Basin
Abstract

GRACE (Gravity Recovery and Climate Experiment) has been widely used to evaluate terrestrial water storage (TWS) and groundwater storage (GWS). However, the coarse‐resolution of GRACE data has limited the ability to identify local vulnerabilities in water storage changes associated with climatic and anthropogenic stressors. This study employs high‐resolution (1 km2) GRACE data generated through machine learning (ML) based statistical downscaling to illuminate TWS and GWS dynamics across twenty sub‐regions in the Indus Basin. Monthly TWS and GWS anomalies obtained from a geographically weighted random forest (RFgw) model maintained good consistency with original GRACE data at the 25 km2grid scale. The downscaled data at 1 km2resolution illustrate the spatial heterogeneity of TWS and GWS depletion within each sub‐region. Comparison with in‐situ GWS from 2,200 monitoring wells shows that downscaling of GRACE data significantly improves agreement with in‐situ data, evidenced by higher Kling‐Gupta Efficiency (0.50–0.85) and correlation coefficients (0.60–0.95). Hotspots with the highest TWS and GWS decline rate between 2002 and 2023 were Dehli Doab (−442, −585 mm/year), BIST Doab (−367, −556 mm/year), Rajasthan (−242, −381 mm/year), and BARI (−188, −333 mm/year). Based on a general additive model, 47%–83% of the TWS decline was associated with anthropogenic stressors mainly due to increasing trends of crop sown area, water consumption, and human settlements. The decline rate of TWS and GWS anomalies was lower (i.e., −25 to −75 mm/year) in upstream sub‐regions (e.g., Yogo, Gilgit, Khurmong, Kabul) where climatic factors (downward shortwave radiations, air temperature, and sea surface temperature) explained 72%–91% of TWS/GWS changes. The relative influences of climatic and anthropogenic stressors varied across sub‐regions, underscoring the complex interplay of natural‐human activities in the basin. These findings inform place‐based water resource management in the Indus Basin by advancing the understanding of local vulnerabilities.

 
more » « less
Award ID(s):
2114701
PAR ID:
10523668
Author(s) / Creator(s):
 ;  ;  ;  
Publisher / Repository:
DOI PREFIX: 10.1029
Date Published:
Journal Name:
Water Resources Research
Volume:
60
Issue:
7
ISSN:
0043-1397
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    GRACE satellite data are widely used to estimate groundwater storage (GWS) changes in aquifers globally; however, comparisons with GW monitoring and modeling data are limited. Here we compared GWS changes from GRACE over 15 yr (2002–2017) in 14 major U.S. aquifers with groundwater‐level (GWL) monitoring data in ~23,000 wells and with regional and global hydrologic and land surface models. Results show declining GWS trends from GRACE data in the six southwestern and south‐central U.S. aquifers, totaling −90 km3over 15 yr, related to long‐term (5–15 yr) droughts, and exceeding Lake Mead volume by ~2.5×. GWS trends in most remaining aquifers were stable or slightly rising. GRACE‐derived GWS changes agree with GWL monitoring data in most aquifers (correlation coefficients,R = 0.52–0.95), showing that GRACE satellites capture groundwater (GW) dynamics. Regional GW models (eight models) generally show similar or greater GWS trends than those from GRACE. Large discrepancies in the Mississippi Embayment aquifer, with modeled GWS decline approximately four times that of GRACE, may reflect uncertainties in model storage parameters, stream capture, pumpage, and/or recharge rates. Global hydrologic models (2003–2014), which include GW pumping, generally overestimate GRACE GWS depletion (total: approximately −172 to −186 km3) in heavily exploited aquifers in southwestern and south‐central U.S. by ~2.4× (GRACE: −74 km3), underscoring needed modeling improvements relative to anthropogenic impacts. Global land surface models tend to track GRACE GWS dynamics better than global hydrologic models. Intercomparing remote sensing, monitoring, and modeling data underscores the importance of considering all data sources to constrain GWS uncertainties.

     
    more » « less
  2. Abstract

    Global Navigation Satellite System (GNSS) vertical displacements measuring the elastic response of Earth's crust to changes in hydrologic mass have been used to produce terrestrial water storage change (∆TWS) estimates for studying both annual ∆TWS as well as multi‐year trends. However, these estimates require a high observation station density and minimal contamination by nonhydrologic deformation sources. The Gravity Recovery and Climate Experiment (GRACE) is another satellite‐based measurement system that can be used to measure regional TWS fluctuations. The satellites provide highly accurate ∆TWS estimates with global coverage but have a low spatial resolution of ∼400 km. Here, we put forward the mathematical framework for a joint inversion of GNSS vertical displacement time series with GRACE ∆TWS to produce more accurate spatiotemporal maps of ∆TWS, accounting for the observation errors, data gaps, and nonhydrologic signals. We aim to utilize the regional sensitivity to ∆TWS provided by GRACE mascon solutions with higher spatial resolution provided by GNSS observations. Our approach utilizes a continuous wavelet transform to decompose signals into their building blocks and separately invert for long‐term and short‐term mass variations. This allows us to preserve trends, annual, interannual, and multi‐year changes in TWS that were previously challenging to capture by satellite‐based measurement systems or hydrological models, alone. We focus our study in California, USA, which has a dense GNSS network and where recurrent, intense droughts put pressure on freshwater supplies. We highlight the advantages of our joint inversion results for a tectonically active study region by comparing them against inversion results that use only GNSS vertical deformation as well as with maps of ∆TWS from hydrological models and other GRACE solutions. We find that our joint inversion framework results in a solution that is regionally consistent with the GRACE ∆TWS solutions at different temporal scales but has an increased spatial resolution that allows us to differentiate between regions of high and low mass change better than using GRACE alone.

     
    more » « less
  3. Abstract. We investigate the interannual and interdecadalhydrological changes in the Amazon River basin and its sub-basins duringthe 1980–2015 period using GRACE satellite data and a physically based, 2 kmgrid continental-scale hydrological model (LEAF-Hydro-Flood) that includes aprognostic groundwater scheme and accounts for the effects of land use–landcover (LULC) change. The analyses focus on the dominant mechanisms thatmodulate terrestrial water storage (TWS) variations and droughts. We findthat (1) the model simulates the basin-averaged TWS variations remarkablywell; however, disagreements are observed in spatial patterns of temporaltrends, especially for the post-2008 period. (2) The 2010s is the driestperiod since 1980, characterized by a major shift in the decadal mean comparedto the 2000s caused by increased drought frequency. (3) Long-term trends in TWSsuggest that the Amazon overall is getting wetter (1.13 mm yr−1), but itssouthern and southeastern sub-basins are undergoing significant negative TWSchanges, caused primarily by intensified LULC changes. (4) Increasingdivergence between dry-season total water deficit and TWS release suggests astrengthening dry season, especially in the southern and southeasternsub-basins. (5) The sub-surface storage regulates the propagation ofmeteorological droughts into hydrological droughts by strongly modulatingTWS release with respect to its storage preceding the drought condition. Oursimulations provide crucial insight into the importance of sub-surface storagein alleviating surface water deficit across Amazon and open pathways forimproving prediction and mitigation of extreme droughts under changingclimate and increasing hydrologic alterations due to human activities (e.g.,LULC change). 
    more » « less
  4. Abstract

    The rapid decline of groundwater resources in South Asia due to excessive irrigation during dry season requires awareness of optimal on‐field water requirements. Such information is currently provided on farmer cellphones through an operational Irrigation Advisory System (IAS). To minimize the cost of sending such irrigation advisory texts to farmers while maximizing impact of IAS on groundwater sustainability, we integrated Gravity Recovery and Climate Experiment (GRACE) data with Landsat Thermal Infrared (TIR) Imagery to target regions in greater need of the IAS service. We demonstrated the concept of an improved IAS over eight irrigation districts of the Ganges and Indus basins. The Surface Energy Balance Algorithm for Land (SEBAL) was used to monitor on‐field water consumption (evapotranspiration‐ET) over cropped areas using Landsat TIR data at plot‐scale spatial resolution. Comparison of SEBAL ET with crop water demand from Penman‐Monteith (FAO56) technique quantified the extent of overirrigation at the plot scale and provided a tangible pathway to microtarget the IAS service only to farmers with the largest groundwater use footprint, thereby improving the impact of the IAS service further. Our results suggested that an operational IAS that integrates GRACE and Landsat TIR data on average can save about 85% (80 million m3) of groundwater per dry season for irrigation districts of Northern India and 87% (or 150 million m3) per year for irrigation districts of Eastern Pakistan.

     
    more » « less
  5. These datasets accompany a publication in Geophysical Research Letters by Martens et al. (2024), entitled: "GNSS Geodesy Quantifies Water-Storage Gains and Drought Improvements in California Spurred by Atmospheric Rivers." Please refer to the manuscript and supporting information for additional details.

    Dataset 1: Seasonal Changes in TWS based on the Mean and Median of the Solution Set

    We estimate net gains in water storage during the fall and winter of each year (October to March) using the mean TWS solutions from all nine inversion products, subtracting the average storage for October from the average storage for March in the following year. One-sigma standard deviations are computed as the square root of the sum of the variances for October and for March. The variance in each month is computed based on the nine independent estimates of mean monthly storage (see “GNSS Analysis and Inversion” in the Supporting Information).

    The dataset includes net gains in water storage for both the Sierra Nevada and the SST watersheds (see header lines). For each watershed, results are provided in units of volume (km3) and in units of equivalent water height (mm). Furthermore, for each watershed, we also provide the total storage gains based on non-detrended and linearly detrended time series. In columns four and five, respectively, we provide estimates of snow water equivalent (SWE) from SNODAS (National Operational Hydrologic Remote Sensing Center, 2023) and water-storage changes in surface reservoirs from CDEC (California Data Exchange Center, 2023). In the final column, we provide estimates of net gains in subsurface storage (soil moisture plus groundwater), which are computed by subtracting SWE and reservoir storage from total storage.

    For each data block, the columns are: (1) time period (October of the starting year to March of the following year); (2) average gain in total water storage constrained by nine inversions of GNSS data; (3) one-sigma standard deviation in the average gain in total water storage; (4) gain in snow water equivalent, computed by subtracting the average snow storage in October from the average snow storage in March of the following year; (5) gain in reservoir storage (CDEC database; within the boundaries of each watershed), computed by subtracting the average reservoir storage in October from the average reservoir storage in March of the following year; and (6) average gain in subsurface water storage, estimated as the average gain in total water storage minus the average gain in snow storage minus the gain in reservoir storage.

    For the period from October 2022 to March 2023, we also compute mean gains in total water storage using daily estimates of TWS. Here, we subtract the average storage for the first week in October 2022 (1-7 October) from the average storage for the last week in March 2023 (26 March – 1 April). The one-sigma standard deviation is computed as the square root of the sum of the variances for the first week in October and the last week in March. The variance in each week is computed based on the nine independent estimates of daily storage over seven days (63 values per week). The storage gains for 2022-2023 computed using these methods are distinguished in the datafile by an asterisk (2022-2023*; final row in each data section).

    Dataset 1a provides estimates of storage changes based on the mean and standard deviation of the solution set. Dataset 1b provides estimates of storage changes based on the median and inter-quartile range of the solution set.

    Dataset 2: Estimated Changes in TWS in the Sierra Nevada

    Changes in TWS (units of volume: km3) in the Sierra Nevada watersheds. The first column represents the date (YYYY-MM-DD). For monthly solutions, the TWS solutions apply to the month leading up to that date. The remaining nine columns represent each of the nine solutions described in the text. “UM” represents the University of Montana, “SIO” represents the Scripps Institution of Oceanography, and “JPL” represents the Jet Propulsion Laboratory. “NGL” refers to the use of GNSS analysis products from the Nevada Geodetic Laboratory, “CWU” refers to Central Washington University, and “MEaSUREs” refers to the Making Earth System Data Records for Use in Research Environments program. The time series have not been detrended.

    We highlight that we have added changes in reservoir storage (see Dataset 8) back into the JPL solutions, since reservoir storage had been modeled and removed from the GNSS time series prior to inversion in the JPL workflow (see “Detailed Description of Methods” in the Supporting Information). Thus, the storage values presented here for JPL differ slightly from storage values pulled directly from Dataset 6 and integrated over the area of the Sierra Nevada watersheds.

    Dataset 3: Estimated Changes in TWS in the Sacramento-San Joaquin-Tulare Basin

    Same as Dataset 2, except that data apply to the Sacramento-San Joaquin-Tulare (SST) Basin.

    Dataset 4: Inversion Products (SIO)

    Inversion solutions (NetCDF format) for TWS changes across the western US from January 2006 through March 2023. The products were produced at the Scripps Institution of Oceanography (SIO) using the methods described in the Supporting Information.

    Dataset 5: Inversion Products (UM)

    Inversion solutions (NetCDF format) for TWS changes across the western US from January 2006 through March 2023. The products were produced at the University of Montana (UM) using the methods described in the Supporting Information.

    Dataset 6: Inversion Products (JPL)

    Inversion solutions (NetCDF format) for TWS changes across the western US from January 2006 through March 2023. The products were produced at the Jet Propulsion Laboratory (JPL) using the methods described in the Supporting Information.

    Dataset 7: Lists of Excluded Stations

    Stations are excluded from an inversion for TWS change based on a variety of criteria (detailed in the Supporting Information), including poroelastic behavior, high noise levels, and susceptibility to volcanic deformation. This dataset provides lists of excluded stations from each institution generating inversion products (SIO, UM, JPL).

    Dataset 8: Lists of Reservoirs and Lakes

    Lists of reservoirs and lakes from the California Data Exchange Center (CDEC) (California Data Exchange Center, 2023), which are shown in Figures 1 and 2 of the main manuscript. In the interest of figure clarity, Figure 1 depicts only those reservoirs that exhibited volume changes of at least 0.15 km3 during the first half of WY23.

    Dataset 8a includes all reservoirs and lakes in California that exhibited volume changes of at least 0.15 km3 between October 2022 and March 2023. The threshold of 0.15 km3 represents a natural break in the distribution of volume changes at all reservoirs and lakes in California over that period (169 reservoirs and lakes in total). Most of the 169 reservoirs and lakes exhibited volume changes near zero km3. Datasets 8b and 8c include subsets of reservoirs and lakes (from Dataset 8a) that fall within the boundaries of the Sierra Nevada and SST watersheds.

    Furthermore, in the JPL data-processing and inversion workflow (see “Detailed Description of Methods” in the Supporting Information), surface displacements induced by volume changes in select lakes and reservoirs are modeled and removed from GNSS time series prior to inversion. The water-storage changes in the lakes and reservoirs are then added back into the solutions for water storage, derived from the inversion of GNSS data. Dataset 8d includes the list of reservoirs used in the JPL workflow.

    Dataset 9: Interseismic Strain Accumulation along the Cascadia Subduction Zone

    JPL and UM remove interseismic strain accumulation associated with locking of the Cascadia subduction zone using an updated version of the Li et al. model (Li et al., 2018); see Supporting Information Section 2d. The dataset lists the east, north, and up velocity corrections (in the 4th, 5th, and 6th columns of the dataset, respectively) at each station; units are mm/year. The station ID, latitude, and longitude are listed in columns one, two, and three, respectively, of the dataset.

    Dataset 10: Days Impacted by Atmospheric Rivers

    A list of days impacted by atmospheric rivers within (a) the HUC-2 boundary for California from 1 January 2008 until 1 April 2023 [Dataset 10a] and (b) the Sierra Nevada and SST watersheds from 1 October 2022 until 1 April 2023 [Dataset 10b]. File formats: [decimal year; integrated water-vapor transport (IVT) in kg m-1 s-1; AR category; and calendar date as a two-digit year followed by a three-character month followed by a two-digit day]. The AR category reflects the peak intensity anywhere within the watershed. We use the detection and classification methods of (Ralph et al., 2019; Rutz et al., 2014, 2019). See also Supporting Information Section 2i.

    Dataset 10c provides a list of days and times when ARs made landfall along the California coast between October 1980 and September 2023, based on the MERRA-2 reanalysis using the methods of (Rutz et al., 2014, 2019). Only coastal grid cells are included. File format: [year, month, day, hour, latitude, longitude, and IVT in kg m-1 s-1]. Values are sorted by time (year, month, day, hour) and then by latitude. See also Supporting Information Section 2g.

     
    more » « less