Session: B.5 Hydrology
(Convener: )

Title: EGSIEM's Near Realtime Mass Transport Products For Monitoring Of Hydrological Extreme Events
Presenter: Flechtner, Frank
Co-Authors: A. Kvas, C. Gruber, T. Mayer-Gürr, A. Güntner, B. Gouweleeuw

Abstract: The nominal time delay of the GRACE Level-1 instrument data (11 days) and of the derived monthly global Level-2 gravity field products (60 days) makes the application of GRACE for monitoring of e.g. hydrological extremes difficult. Flood forecast models need, e.g. near-real time (NRT) information to estimate the probable development of the event in terms of flood stage or river discharge with typical lead times of a few days for larger river basins.

To enable the application of GRACE (and later GRACE-FO) mass redistribution data for rapid monitoring of hydrological extreme events, the EU funded project EGSIEM (European Gravity Service for Improved Emergency Management) has established a NRT and Regional Service, that aims to reduce the time delay of mass transport products to less than 5 days, to increase the time resolution from one month to one day, and to improve the quality by providing regional solutions based on alternative representations of the gravity field, e.g. space-localizing radial base functions.

The quality of the NRT mass transport products is tested using GNSS loading as well as hydrological flood events. An operational test run of the NRT Service has started on April 1. Here, the NRT products are provided on a daily basis to the EGSIEM Hydrological Service which derives NRT flood indicators to be used within DLR´s Center for Satellite-based Crisis Information.

The presentation will focus on the proof of concept of the NRT service, will show some statistics for historical flood events and will summarize results gained throughout the operational test run.

Back to top

Title: From Drought to Recovery in California:Insights Using GRACE, Models and In Situ Observations
Presenter: Famiglietti, Jay
Co-Authors: A. McEvoy, P. Liu, D. Stampoulis, JT Reager, D. Wiese, F. Landerer

Abstract: Over the life of the GRACE mission, California has experienced 5 distinct hydrometeorological phases. These include modest wetting from 2002-2006; a drought phase between 2006 and 2010; a modest wet period during the mild El Niño of 2010-211; the major drought of 2011-2015; and a major recovery phase since 2015. In this presentation, we explore the drought of 2011-2015, as well as the recovery since 2015. In particular we investigate whether the recent recovery has satisfied the GRACE-based drought deficit estimate which peaked in 2015, as well as whether surface water reservoirs and groundwater levels have been significantly restored.

Back to top

Title: A Global Gridded Dataset of GRACE Drought Severity Index
Presenter: Meng, Zhao
Co-Authors: A. Geruo, I. Velicogna, J. Kimball

Abstract: We developed a new satellite-based drought severity index (DSI) using time-variable terrestrial water storage (TWS) estimate from the Gravity and Climate Experiment Mission (GRACE). The GRACE-DSI is fully observational and provides globally continuous drought monitoring, allowing monitoring also in regions where sparse ground observations (especially precipitation) constrain the use of traditional model-based monitoring methods. The GRACE-DSI enables drought feature comparison across regions and periods, it is unaffected by uncertainties associated with soil water balance models and meteorological forcing data, and it incorporates water storage changes from human impacts including groundwater withdrawals that modify land surface processes and impact water management.

During the past 14-years the GRACE-DSI captures major global drought events and compares favorably with traditional drought monitoring tools such as the United States Drought Monitor (USDM), Palmer DSI (PDSI), standardized precipitation and evapotranspiration index (SPEI), satellite-derived vegetation index and surface soil moisture estimates, and in-situ groundwater observations. The GRACE-DSI has a quantified uncertainty that facilitates its incorporation into decision making and operational applications.

We established a framework for combining the GRACE-DSI with other drought metrics to better characterize hydrological drought, including its propagation and persistence, and recovery. The GRACE- DSI quantifies the overall variation in terrestrial water storage, which complements the PDSI which is limited to the surface layers. Comparison between GRACE-DSI and PDSI helps monitor the evolution of a meteorological drought to a hydrological drought. We use the SPEI to quantify the anomalies in climatic water balance on different time-scales. The times-scales at which the SPEI metrics agree with the GRACE-DSI provide an estimate of the persistence of a hydrological drought event. We linked the time-scales of the maximum correlation between GRACE-DSI and SPEI to the depth of the water storage components that dominate the regional hydrologic cycle: maximum correlation occurs for longer period SPEI when deeper water storage changes dominate and for short period SPEI when shallower water storage dominate.

Back to top

Title: Investigation of the intense mass depletion trend in northwest China
Presenter: Rodell, Matthew
Co-Authors: J.S. Famiglietti, D.N. Wiese, J.T. Reager, H.K. Beaudoing, F.W. Landerer, and M.-H. Lo

Abstract: GRACE has detected rapid mass loss in a relatively small area of northwest China. There are many confounding factors that complicate diagnosis of this apparent trend, including glacier melt, coal mining, inter-basin transfers of surface water, and population growth and associated increases in irrigation. In this presentation we attempt to explain how and to what extent all of these factors contribute to the observed trend.

Back to top

Title: Sustained water loss in California's mountain ranges during severe drought from 2012 through 2015 inferred from GPS
Presenter: Argus, Donald
Co-Authors: F. Landerer, D. Wiese, H. Martens, Y. Fu, J. Famiglietti, B. Thomas, T. Farr, A. Moore, M. Watkins

Abstract: Drought struck California during 7 of the 9 years from 2007 through 2015, reducing the state's available water resources. Pumping of Central Valley groundwater has produced spectacular land subsidence. Uplift of the adjacent Sierra Nevada mountains has been proposed to be either tectonic uplift or solid Earth's elastic response to unloading of Central Valley groundwater. We find that, of the 24 mm of uplift of the Sierra Nevada from October 2011 to October 2015, just 5 mm is produced by Central Valley groundwater loss, less than 2 mm is tectonic uplift, and 17 mm is solid Earth's elastic response to water loss in the Sierra Nevada. We invert GPS vertical displacements recording solid Earth's elastic response to infer changes in water storage across the western U.S. from January 2006 through June 2017. We find water changes to be sustained over periods of drought or heavy precipitation: the Sierra Nevada lost 15 ±19 km3 of water during drought from October 2006 to October 2009, gained 18 km ±14 km3 of water during heavy precipitation from October 2009 to October 2011, and lost 45 ±21 km3 of water during severe drought from October 2011 to October 2015 (95% confidence limits). Such large changes are not in hydrology models: snow accumulation in October is negligible and long-term soil moisture change is small. We infer there must be large loss of either deep soil moisture or groundwater in river alluvium and in crystalline basement in the Sierra Nevada. The results suggest there to be parching of water in the ground during the summer of years of drought and seeping of melting snow into the Sierra Nevada in the spring of years of heavy precipitation.

Back to top

Title: Groundwater withdrawals under drought: reconciling GRACE and models in the United States High Plains Aquifer
Presenter: Nie, Wanshu
Co-Authors: B. Zaitchik, S. Kumar, M. Rodell

Abstract: Advanced Land Surface Models (LSM) offer a powerful tool for studying and monitoring hydrological variability. Highly managed systems, however, present a challenge for these models, which typically have simplified or incomplete representations of human water use, if the process is represented at all. GRACE, meanwhile, detects the total change in water storage, including change due to human activities, but does not resolve the source of these changes. Here we examine recent groundwater declines in the US High Plains Aquifer (HPA), a region that is heavily utilized for irrigation and that is also affected by episodic drought. To understand observed decline in groundwater (well observation) during a recent multi-year drought, we modify the Noah-MP LSM to include a groundwater pumping irrigation scheme. To account for seasonal and interannual variability in active irrigated area we apply a monthly time-varying greenness vegetation fraction (GVF) dataset to the model. A set of three experiments were performed to study the impact of irrigation with groundwater withdrawal on the simulated hydrological cycle of the HPA. The results show that including the groundwater pumping irrigation scheme in Noah-MP improves model agreement with GRACE mascon solutions for TWS and well observations of groundwater anomaly in the southern HPA, including Texas and Kansas. Results for the HPA in Nebraska are mixed, likely due to the model’s failure to capture major groundwater recharge in winter during drought period. This presentation will highlight the value of the GRACE constraint for model development and the importance of considering the applicability of different solution based GRACE products for model evaluation, present estimates of the relative contribution of climate variability and irrigation to declining TWS in the HPA under drought, and identify opportunities to integrate GRACE-FO with models for water resource monitoring in heavily irrigated regions.

Back to top

Title: GRACE and geochemical–based approach to assess the age, origin, and optimum utilization of fossil aquifers in the Arabian Peninsula
Presenter: Sultan, Mohamed
Co-Authors: K. Fathy, H. Save

Abstract: Observations extracted from the Gravity Recovery and Climate Experiment (GRACE) CSR 1° x 1° mascon solutions, and geochemical and geochronological data, were used to assess the sustainability, origin, age, evolution, and groundwater potential of large fossil aquifers using the Mega Aquifer System (MAS) in Arabia as a test site. Analysis of GRACE solutions and Land Surface Model outputs revealed: (1) the MAS central and northern sections are experiencing high groundwater extraction and depletion rates (TWS and GWS), yet sustainable exploitation of the aquifer could be achieved by reducing the annual extraction by ~ 3 km3, and (2) the MAS southern sections (Rub Al Khali Aquifer System (RAKAS) are experiencing low GWS depletion rates to steady-state conditions. Geochemical, remote sensing, and field investigations over the RAKAS revealed: (1) increasing 36Cl model ages along groundwater flow directions (up to 970,000 years), indicating aquifer recharge in wet Pleistocene periods; (2) progressive depletion in the O and H stable isotopic compositions of aquifers with increasing distance from the Red Sea Hills basement outcrops, indicating modest recharge during prevailing dry conditions; and (3) the presence of relatively fresh (TDS: 800 to 2,800 mg/L) and non-radioactive (226Ra + 228Ra < 0.185 Bq/kg) water in western RAKAS. Findings suggest that sustainable agricultural development is achievable at current extraction rates in western RAKAS and provide a replicable and cost-effective model.

Back to top

Title: Progress in understanding subsurface and surface water links using GRACE
Presenter: Reager, John
Co-Authors: M.J. Tourian, H.E. Macedo, E. Beighley, J. Lucey, N. Sneeuw, J. Famiglietti and C. David

Abstract: GRACE data have been shown to capture the extreme hydrologic conditions associated with floods and droughts as the land surface fills to become saturated or slowly empties to become dry. The capacity of soils to absorb, hold and retain water is important in each of these processes, and represents a fundamentally important hydrologic parameter for modeling and prediction of extremes. Recent research and progress on this topic will be presented, including results and analysis from three recent or ongoing projects: (1) A coupled analysis of GRACE-based land water storage and streamgage-based river discharge to apply a base flow recession approach for the first time using GRACE observations in the Mississippi River Basin; (2) Mapping of the runoff generation process and the contributions of baseflow to streamflow, as well as estimating total water storage in the Amazon and Mississippi River Basins; and (3) a global assessment of the relationship between GRACE terrestrial water storage anomalies and surface water formation from radiometric remote sensing observations of surface water inundation extent. These results suggest the extent to which knowledge of storage conditions can help in flood and drought mechanistic representations, and explores the range of conditions, scales and intensities for which storage-driven flood and drought processes are relevant and observable by satellite gravity.

Back to top

Title: Aquifer Response and Groundwater Flow: Inferences from GRACE, GGMs, Field, and RS data
Presenter: Abdelmohsen, Karem
Co-Authors: M. Sultan, and H. Save

Abstract: Remote sensing, field, and geophysical data were used to investigate (1) the structural control on groundwater flow in large aquifers and (2) aquifer response to wet and dry periods using the less studied Nubian Sandstone Aquifer System (NSAS) of NE Africa as a test site. A three-fold exercise was conducted. Firstly, GOCE-based Global Geopotential Models (GGMs) were compared to terrestrial gravity data to identify the optimum model for deriving Bouguer gravity anomalies. Secondly, structures and uplifts were mapped using hill shade images and their extension in the subsurface were mapped using the Eigen_6C4 model-derived Bouguer anomalies and their Tilt Derivative products (TDR). Thirdly, GRACE CSR 1° x 1° mascon solutions were examined to identify the temporal and spatial mass variations within the study area. Findings include: (1) The Eigen-6C4 shows the lowest deviation from the terrestrial gravity anomalies; (2) the surface expressions of structures matched with their postulated subsurface extensions (extracted from the Eigen-6C4); (3) identified fault systems include: Red Sea rift-related N-S to NW-SE trending grabens formed by reactivating basement structures during Red Sea opening and Syrian arc-related NE-SW trending dextral shear systems; (4) TWS patterns are uniform throughout the identified shear systems; (5) basement uplifts impede or redirect the groundwater flow, and (6) only in wet periods does the Dakhla Aquifer System (DAS) receive substantial replenishment by groundwater flow from the south, (7) the fast response of the aquifer in wet periods can be explained by preferential flow in fractured media, and (8) groundwater discharge (and soil moisture) is increasing in lowlands during wet periods and vice versa in dry periods.

Back to top

Title: Application of Multiplicative Random Cascades to Spatially Downscale Observed Terrestrial Water Storage Anomalies
Presenter: Ukasha, Muhammad
Co-Authors: J.A. Ramirez

Abstract: Gravity Recovery and Climate Experiments (GRACE) satellite mission has been observing terrestrial water storage anomalies (TWSA) at a monthly scale since 2002. Given its coarse spatial resolution (i.e. ≥160,000 km2), TWSAs have been used in numerous hydrological studies at a regional scale. However, TWSA spatial footprint limits its use in understanding small-scale spatial variability of terrestrial water storage in connection with hydrologic, atmospheric, ecological and socio-economic processes. Therefore, spatial downscaling of observed TWSA is of great interest to hydrological community. In this study we explored the possibility of using the well-known, random cascade models to perform downscaling of GRACE TWSA. Using 0.5 degree GRACE MASCONS dataset for the southwest United States, we first analyzed the TWSA for spatial self-similarity. Near mono-fractal behavior (i.e., simple scaling) of TWSA was observed in the process of spatially upscaling the GRACE TWSA observations from 0.5 degree to 4 degree. Given this behavior of TWSA, random cascades can be used to spatially model TWSA at scales ranging from 0.5 to 4 degrees. However, assuming that a similar scaling structure is present for scales below 0.5 degree, we used multiplicative random cascades to downscale TWSA from the large-scale (4 degree) to the small-scale (1/16th degree). Downscaling was performed using two variants of random cascade generators i.e., i) uniformly distributed cascade and ii) beta log-normally distributed cascade. For each variant of cascade, 1000 realizations were performed to downscale TWSA. By comparing the best realization (realization for which Euclidean distance between modeled and observed TWSA images is minimum at 0.5 degree) for each variant of cascade, we found that random cascade based on uniform distribution better models the TWSA at small-scales. In this paper we present the method to downscale TWSA based on its scaling behavior and results of the described methodology.

Back to top

Title: River flows and snow water equivalent estimated using the GRACE observations
Presenter: Wang, Shusen
Co-Authors: H. Russell

Abstract: Snow mass (or Snow Water Equivalent, SWE) and river flow measurements are difficult and less accurate in cold regions due to isolation, the hash environment and lack of instrumentation. Peak river flows and floods in cold regions are commonly a result of snowmelt during the spring break-up. Quantifying SWE and modelling river flows over cold regions are important in many hydrological, climate and ecological applications (e.g., floods forecasting), but remain challenged due to scarce data and data quality issues in basin-scale snow observations and lack of knowledge for cold region hydrological processes. This study developed a model for estimating basin-scale SWE and river flows using the total water storage (TWS) observations from the GRACE satellite mission. The SWE estimation is based on mass balance approach which is independent of in situ snow gauge observations, it thus largely eliminates the limitations and uncertainties with the current in situ or remote sensing snow estimates. By using the GRACE-based SWE estimates, river flows are forecasted by simulating surface runoff and baseflow from groundwater discharge. The model also quantifies the hysteresis between the snowmelt and the streamflow rates, or the lump time for water travel in the basin. The model has been tested for a number of basins in Canada of varying size from 106 to 105 km2. The predicted river flows were compared with the observed values at hydrometric stations. This study provides a relatively simple method which only needs GRACE and temperature data for SWE and river flow estimates, and flood forecasting. The model can be particularly useful for regions with spare observation networks. The results help better understand the roles of environmental factors in determining floods and their spatial variations across different hydroclimatic regions. This study also provides the basis for downscaling the GRACE TWS data to small scales (104 km2) for watershed/aquifer surface water/groundwater applications.

Back to top

Title: GRACE: A unique tool to advance precipitation estimation over cold regions
Presenter: Behrangi, Ali
Co-Authors: A. Gardner; JT Reager; A. Singh

Abstract: Here we show that GRACE can advance precipitation estimation in high latitudes and high altitudes, where the uncertainties in the current precipitation products are highest. With the lack of high quality precipitation estimates in cold regions and the growing need for more spatiotemporally comprehensive precipitation estimates, GRACE (and in near future GRACE-FO) can provide unique and independent information that has been underutilized for the improvement of precipitation analysis. This is especially timely with significant advancements in the refinement of the GRACE gravity solution. More specifically, by performing case studies over Eurasia and two Endorheic basins in the Tibetan plateau, we demonstrate how GRACE can help constrain precipitation products. GRACE estimates suggest that the global precipitation climatology project (GPCP) likely overestimates precipitation over Eurasia, consistent with CloudSat observations, and underestimates precipitation in winter in the Tibetan plateau. Furthermore, we demonstrate that GRACE is extremely useful to select a proper gauge undercatch correction factor, commonly applied to rain-gauges. The impact of this study is significant as all of the global precipitation products use in-situ data for bias correction.

Back to top

Title: Merging Space-based Gravimetry and Radiometry to Improve Regional-scale Estimates of Terrestrial Water Storage
Presenter: Forman, Bart
Co-Authors: J. Wang, M. Girotto, and R.H. Reichle

Abstract: The goal of this project is to simultaneously merge multi-frequency, multi-polarization passive microwave brightness temperatures that are sensitive to land surface soil and snow water with gravimetric retrievals of terrestrial water storage (TWS) in order to improve land surface model estimates of snow water equivalent (SWE), soil moisture (SM), and groundwater. L-band observations collected by SMOS and SMAP are sensitive to soil moisture whereas X-, K- and Ka-band observations collected by AMSR-E and AMSR2 are sensitive to snow mass. General science questions include:

1. What is the complementary gain via inclusion of L- (1.4 GHz), X- (~10 GHz), K- (~19 GHz), and/or Ka-band (~36 GHz) microwave brightness temperature (Tb) assimilation in conjunction with GRACE TWS assimilation?
2. How do such synergistic effects between different radiometric observations manifest themselves in space and time relative to GRACE TWS assimilation alone?
3. Can a multi-variate assimilation framework effectively downscale GRACE TWS retrievals in space and time thereby adding spatial and temporal resolution to the GRACE TWS retrievals that currently does not exist?

Multi-year, “synthetic” experiments across the Volga River basin in Russia along with “real” experiments across North America highlight the measurable improvements associated with AMSR-E brightness temperature, SMOS brightness temperature, and GRACE TWS retrieval assimilation. For example, X-, K-, and Ka-band assimilation improves estimates of SWE, but has relatively little influence on groundwater estimates. Alternatively, assimilation of GRACE TWS improves model estimates of groundwater, but has relatively little influence on modeled SWE. Simultaneous assimilation of radiometry and gravimetry presents a host of new challenges, including competing effects between different observation types as well as oversampling in the GRACE TWS retrievals. This talk will highlight some of the benefits (and detractions) associated with multi-sensor assimilation and the merger of both radiometric and gravimetric observations into an advanced land surface model.

Back to top

Title: Combining GRACE water storage and SMOS soil moisture retrievals to improve water and vegetation forecast
Presenter: Tregoning, Paul
Co-Authors: S. Tian

Abstract: Spatial and temporal variations of soil moisture content and groundwater storage strongly affect flooding, solute transport and vegetation productivity. However, an accurate assessment of root-zone soil moisture and groundwater storage change is difficult due to an inability to monitor with ground-based point measurement techniques at an appropriate spatial resolution, and the uncertainty associated with land surface model predictions. Global retrievals of terrestrial water storage (TWS) change and soil moisture (SM) from satellites provide an opportunity to better disaggregate root-zone soil moisture and groundwater from TWS through data assimilation. Data assimilation is an effective approach to optimally combine information from both model predictions and observations, and capture the unmolded activities. For the first time, GRACE TWS and SMOS near-surface SM estimates were jointly assimilated into a water balance model (W3 model). The performance of joint assimilation was assessed against open-loop model simulations and the assimilation of either GRACE TWS anomalies or SMOS SM alone. The joint assimilation typically led to more accurate water storage profile estimates with improved surface SM, root-zone SM, and groundwater estimates against in-situ observations. The assimilation successfully downscaled GRACE-derived integrated water storage horizontally and vertically into individual water stores at the same spatial scale as the model and SMOS, and partitioned monthly averaged TWS into daily estimates. These results demonstrate that satellite TWS and SM measurements can be jointly assimilated to produce improved water balance component estimates.

Back to top

Poster Title: Assimilation of GRACE/GRACE-FO and AMSR-E/AMSR-2 into the NASA Catchment Land Surface Model: Towards Year-round Estimation of Terrestrial Water Storage over Snow-Covered Terrain
Presenter: Wang, Jing
Co-Authors: Y. Xue, B.A. Forman, M.Girotto, and R.H. Reichle

Abstract: The Gravity and Recovery Climate Experiment (GRACE) has revolutionized large-scale remote sensing of the Earth’s terrestrial hydrologic cycle and has provided an unprecedented observational constraint for global land surface models. However, the coarse-scale (in space and time) and vertically-integrated measure of terrestrial water storage (TWS) limits GRACE’s applicability to smaller scale hydrologic applications. In order to effectively downscale (in space and time) TWS retrievals while simultaneously enhancing model-based estimates of TWS (and its constituent components), a multi-variate, multi-sensor data assimilation framework is presented. The framework will simultaneously assimilate gravimetric retrievals of TWS from GRACE along with C- and X-band passive microwave (PMW) brightness temperature (Tb) observations over snow-covered terrain. The goal is to merge the multi-sensor observations with an advanced land surface model in order to provide an estimate that is superior to either the observations or model alone.

The framework employs the NASA Catchment Land Surface Model (Catchment) and an ensemble Kalman filter (EnKF). A synthetic assimilation experiment is presented for the Volga river basin in Russia where snow dominates the hydrologic cycle. The skill of the output from the assimilation of synthetic observations is compared with that of model estimates generated without the benefit of assimilation. Preliminary results are shown that the EnKF framework improves modeled estimates of TWS, snow depth, and snow mass (a.k.a. snow water equivalent). The data assimilation routine produces a conditioned (updated) estimate that is more accurate and contains less uncertainty during both the snow accumulation phase of the snow season as well as during the snow ablation season.

Future work will include the multi-variate assimilation of L-band radiometry observations of soil moisture in conjunction with GRACE TWS retrievals and C- and X-band PMW Tb observations. The eventual goal is to extend the EnKF framework so that it can be applied in both snow-dominated and non-snow-dominated basins across the globe.

Back to top

Poster Title: A comparison of seasonal variations in hydrologically-driven vertical deformations of the Earth’s surface as estimated by GRACE, ground-based GPS, and a hydrologic loading model
Presenter: Yin, Gaohong
Co-Authors: B.A.Forman; B.D.Loomis; S.B. Luthcke

Abstract: Seasonal variations in terrestrial water storage (TWS) can cause deformation of the Earth’s crust. In turn, observations of surface displacement can be used in the study of terrestrial hydrology. After accounting for non-hydrologic loading factors in the observations of surface displacement, the remaining signal can be used to infer the movement and timing of water across the terrestrial environment.

This study compares estimates of vertical deformation from Gravity Recovery and Climate Experiment (GRACE), ground-based Global Positioning System (GPS) observations, and hydrologic loading estimates derived from the NASA Catchment Land Surface Model (Catchment) over the United States. A particular focus is given to snow-dominated basins. For GRACE, the mass concentration (mascon) solution after the removal of non-hydrologic contributions was first converted to equivalent vertical displacement. Similarly, GPS observations collected by the Plate Boundary Observatory (PBO) are used to estimate hydrologically-relevant vertical displacements. Utilization of the PBO observations involved removal of the mean seasonal cycle, the effects of atmospheric loading, non-tidal ocean loading, and glacier isostatic adjustment (GIA). A low pass filter was then applied to the GPS data to remove high frequency noise. For the hydrologic loading model derived from Catchment output, the TWS anomaly on a 25-km EASE grid was adopted and then converted to vertical displacement using Green’s function.

A number of study locations across the United States are used in this analysis. For each location, vertical displacements are calculated by integrating the displacements associated with all pixels located within a 200-km radius. Statistics between computed displacements of: 1) GRACE-derived versus GPS-derived, 2) model-derived versus GPS-derived, and 3) GRACE-derived versus model-derived are presented. The cross-comparison aims to explore the consistency between the three different methods. Results show that the three methods agree well across much of the western United States except in the Central Valley of California where active groundwater pumping activities occur that are not represented in the Catchment model. The relatively good consistency between the GRACE and GPS results suggests the potential for a merger of GPS and GRACE observations via Bayesian conditioning in an effort to add more spatial and temporal resolutions to the GRACE TWS retrievals.

Back to top

Poster Title: Using GRACE Satellite Gravimetry for Assessing Large-Scale Hydrologic Extremes
Presenter: Sun., Alexander Y.
Co-Authors: B. Scanlon, A. Aghakouchak, Z. Zhang

Abstract: Global assessment of spatiotemporal variability in terrestrial total water storage anomalies (TWSA) in response to hydrologic extremes is critical for water resources management. Using TWSA derived from Gravity Recovery and Climate Experiment (GRACE) satellites, this study systematically assessed the skill of the TWSA-climatology (TC) approach and breakpoint (BP) detection method for identifying large-scale hydrologic extremes. The BP detection method, which dissects time series at abrupt change points (BPs), was informed by empirical mode decomposition (EMD) that can isolate the signatures of hydrologic extremes. A nonlinear event coincidence analysis measure was then adopted to cross-examine abrupt changes detected by both methods with those detected by the Standardized Precipitation Index (SPI). The methods were demonstrated over the 35 largest global river basins. Results showed that our EMD-informed BP procedure is a promising tool for extreme extraction. Abrupt changes detected by the BP method cross-reference well with SPI anomalies and with documented hydrologic extreme events, whereas event timings obtained by TC were ambiguous for a number of river basins, largely due to the limited length of GRACE record. The BP approach demonstrates a robust wet-dry anomaly detection capability, which will be important for applications with the upcoming GRACE Follow-On mission.

Back to top