GRACE Science Team Meeting

Session B.5-b: Hydrology

(An Overview of the 2nd GRACE Hydrology Workshop)
Matthew Rodell

(The US-Southeast drought observed with GRACE)
Felix Landerer

(Dwindling groundwater in northern India: an application of fitting mascons to GRACE harmonics)
Virendra Tiwari

(Application of GRACE Water Storage for Water Resources Management: Case Study, High Plains Aquifer, US)
Laurent Longuevergne

(Assessing High Latitude Winter Precipitation Using GRACE)
Sean Swenson

(Mass variations in Siberia from GRACE and their relation to permafrost changes)
Holger Steffen

(Using GRACE total water storage data to constrain the dynamics of surface flow in the Community Land Model routing scheme in the Amazon basin)
Caroline de Linage

(GRACE estimates of water storage over Australia, with a note on data filtering)
David Garcia-Garcia

(The role of rivers in the seasonal variations of terrestrial water storage over global basins )
Hyungjun Kim

(Mass rates over Tibet-plateau region from GRACE temporal gravity field)
Zhicai Luo


(Assessing regional mascon GRACE solutions over Australia)
Joseph Awange

(Recent hydrological behavior of the East African Great Lakes region inferred from GRACE, satellite altimetry and rainfall observations)
Melanie Becker

(Groundwater component of land water storage in Asia from GRACE and hydrological modelling)
Melanie Becker

Session: B.5.a - Theme: 2nd Hydrology Workshop
Title: An Overview of the 2nd GRACE Hydrology Workshop
First Author: Matthew Rodell
Presenter: Matthew Rodell

Abstract: The 2nd GRACE Hydrology Workshop that was held just prior to this GRACE Science Team Meeting in Austin, Texas. The goal of this workshop was to share recent results in three areas: (1) water cycle science; (2) integration of GRACE data with models and other observations; (3) water resources and other applications. An overview of the main results will be presented.

Back to top

Session: B.5.b - Hydrology Applications
Title: The US-Southeast drought observed with GRACE
First Author: Felix Landerer
Presenter: Felix Landerer
Co-Authors: Jean O. Dickey

Abstract: The southeast continental US has experienced severe drought conditions over the last several years, resulting in water shortages and user restrictions for the agricultural and private sector. GRACE measurements provide a unique measurement of total terrestrial water storage anomalies, and with nearly 7 years of data, long-term trends can be reliably assessed and compared with hydrological models and standard drought indices. However, one difficulty in the interpretation of GRACE data in particular over land is the integral character of the measurement: from GRACE alone, one cannot distinguish between the sources of anomalies, so additional data needs to be considered. Our analysis of long-term changes over the southeast US indicates that GRACE observes a significantly stronger water loss over this region than hydrological models simulate. This may be related to reductions in groundwater, which is not adequately captured in hydrological model simulations. We also discuss the consistency of the observed GRACE signal with observed precipitation variations, and potential implications for the regionŐs water resources. Ultimately, the goal of this work is to assess GRACE observations as an additional tool for drought monitoring and enhance its value in this regard.

Back to top

Session: B.5.b - Papers are solicited on advances in hydrological applications of GRACE data products
Title: Dwindling groundwater in northern India: an application of fitting mascons to GRACE harmonics
First Author: Virendra Tiwari
Presenter: John Wahr
Co-Authors: John Wahr, Sean Swenson

Abstract: Northern India and its surroundings, home to roughly 600 million people, is probably the most heavily irrigated region in the world. GRACE measurements reveal a steady, large-scale mass loss in this region that we attribute to excessive extraction of groundwater. We use output from hydrological models to remove the contributions from natural water storage variability, and thus to isolate the anthropogenic component. We derive numerical estimates for the groundwater loss by subdividing the region into relatively small, non-overlapping blocks, and simultaneously fitting amplitudes for all blocks to the GRACE spherical harmonic coefficients. This approach is similar in spirit to the more familiar mascon approach used by other groups, but here we are fitting to GRACE harmonic coefficients rather than to level-1 data. Our results suggest that this region lost groundwater at a rate of 54 ± 9 km3/yr between April, 2002 (the start of the GRACE mission) and June, 2008. This is probably the largest rate of groundwater loss in any comparable-sized region on Earth. Its likely contribution to sea level rise is roughly equivalent to that from melting Alaskan glaciers.

Back to top

Session: B.5.b - Hydrology Applications
Title: Application of GRACE Water Storage for Water Resources Management: Case Study, High Plains Aquifer, US
First Author: Laurent Longuevergne
Presenter: Laurent Longuevergne
Co-Authors: B.R. Scanlon, C.R. Wilson, Z.L. Yang, G. Strassberg, J. Chen, G.Y. Niu

Abstract: The Gravity Recovery and Climate Experiment (GRACE) satellites provide an opportunity to monitor regional water storage variations for hydrological studies. Reliable estimates of water storage variations in a basin require a balance between GRACE noise reduction and maximum spatial resolution. The objective of this study was to quantitatively investigate processing choices and their impacts on bias, leakage, GRACE noise reduction, and estimated total error, allowing solution of the trade-offs. Different approaches to data concentration were evaluated and the ability of localizing global spectra using optimal spatiospectral localization was examined. This optimal concentration method allows adaptation of the effective spherical harmonics bandwidth and was found to be a powerful method to achieve balance between spatial and spectral concentration. This method is particularly suited to retrieval of basin-scale water storage variations. The effect of several filtering methods on concentration was interpreted in terms of bias and potential leakage, which can be corrected using a-priori hydrological models. The processing techniques were applied to the intensively monitored High Plains aquifer (450,000 km2 area), where application of the appropriate optimal processing method allowed retrieval of water storage variations at subbasin scales of about 200,000 km2.

Back to top

Session: B.5.b - Hydrology Applications
Title: Assessing High Latitude Winter Precipitation Using GRACE
First Author: Sean Swenson
Presenter: Sean Swenson

Abstract: Gauge estimates of precipitation can be biased low due to wind-induced undercatch. This problem can be especially large during winter when solid precipitation is present.

In this study we use GRACE to provide the first remotely sensed estimates of precipitation gauge undercatch during the cold-season at high latitudes.

Back to top

Session: B.5.b - Hydrology Applications
Title: Mass variations in Siberia from GRACE and their relation to permafrost changes
First Author: Holger Steffen
Presenter: Holger Steffen
Co-Authors: J. Mueller, J. Boike, N. Peterseim

Abstract: After 7 years in orbit, the Gravity Recovery and Climate Experiment (GRACE) satellite mission now also allows detecting smaller secular trends of mass variations as well as long-periodic signals. In this study, we focus on changes of the permafrost regime in Siberia, Russia, using GRACE monthly solutions from the three main analysis centres GFZ, CSR and JPL. The results show that observed positive trends of mass changes are related to large Siberian rivers such as Ob, Lena and Yenisei. Two major trends of about 0.7 microGal/year can be clearly identified. The first concerns the upper Ob River. It includes, depending on the specific GRACE solution centre, the Angara River drainage basin, which is part of the Yenisei River system. The second trend is centered on the upper Lena River north-east of Lake Baikal and is also clearly determined, but with different solution-dependent values. All these significant trends seem to be caused by long-term hydrological changes, especially since no other geophysical explanation is found yet. Similar features can be found in the trend of the GLDAS hydrology model. Removing the hydrological contribution positive mass changes of about 0.8 microGal/year appear in the Central Siberian Plateau and the Kolyma River drainage basin, which may be related to changes in permafrost. However, further investigations are needed to really understand such mass changes and attribute them to the related physical processes.

Back to top

Session: B.5.b - Hydrology Applications
Title: Using GRACE total water storage data to constrain the dynamics of surface flow in the Community Land Model routing scheme in the Amazon basin
First Author: Caroline de Linage
Presenter: Caroline de Linage
Co-Authors: Min-Hui Lo, James Famiglietti

Abstract: The surface water component is a major contributor to total water storage in the Amazon basin. Global models that do not account for surface water routing often show poor agreement with the GRACE total water storage observations in this basin. We show in this study the impact of introducing a routing scheme of the discharge in the land surface model CLM version 3.5. The comparison with the GRACE observations is based on the more significant modes of a Principal Component Analysis of both data sets in order to take into account the spatio-temporal variations of the total water storage at the seasonal scale. We use the GRACE observations to estimate an effective velocity for the entire basin assuming a spatially and temporally constant velocity. The results of different simulations with a spatially variable surface flow velocity are also shown and discussed.

Back to top

Session: B.5.b - Hydrology Applications
Title: GRACE estimates of water storage over Australia, with a note on data filtering
First Author: David Garcia-Garcia
Presenter: David Garcia-Garcia
Co-Authors: Victor Zlotnicki

Abstract: Time-Variable Gravity data from the Gravity Recovery And Climate Experiment (GRACE) mission are filtered by combining two published filters (Swenson & Wahr, 2006, and Chen et al, 2006), with no further filtering step. The result is more effective than when the filters are used separately or followed by a typical gaussian. The filtered data are then used to study total water content over Australia. The annual signal explains 64% of the variance of the data, in good agreement with two independent estimates of the same quantity. The first one is made from the GLDAS hydrological model, and the second one from estimates of the water flux via precipitation, evaporation and river runoff. All three methods show a spatially varying signal, both in variance and in phase. Maximum variance is observed in the north, reaching peak amplitude in March, 4-5 months ahead than in the Southeast. The range of the annual cycle shows variability with an approx. 2-year cycle; sources of this signal, including possible teleconnections, are explored.

Back to top

Session: B.5.b - Papers are solicited on advances in hydrological applications of GRACE data products
Title: The role of rivers in the seasonal variations of terrestrial water storage over global basins
First Author: Hyungjun Kim
Presenter: Hyungjun Kim
Co-Authors: Pat J. -F. Yeh, Taikan Oki, and Shinjiro Kanae

Abstract: Fluvial transport is an important process in terrestrial hydrological circulations, and it is still not clear how significant rivers would contribute to the TWS variations under different climates. However, most of the previous TWS related studies and GRACE hydrology applications implicitly assumed that soil moisture and snow water equivalent are the only major TWS components. In the present study, we investigated the role of rivers in total terrestrial water storage (TWS) variations in global basins.

Ensemble land surface simulations using five global precipitation observations are performed, and the Bayesian Model Averaging (BMA) technique is applied with perturbed effective velocities to obtain more reliable streamflow simulations. GRDC gauged discharges are used to determine the weights of effective velocities for each target basins, and river storage variations in this model framework effectively includes downslope movement of shallow groundwater. From the total 178 global river basins, the 29 target basins are selected according to basin area (> 220,000 km2), skills of streamflow simulations (correlate coefficient (CC) > 0.5 and Root-Mean-Squared-Error (RMSE) < 100 mm/month). Also, in order to compare how each individual component contributes to total TWS variations, two new indices, Component Contribution Ratio (CCR) and Component Exchange Intensity (CEI), are devised.

Analysis considering river storage shows substantial error reduction in terms of the CC and normalized RMSE except for the dry basins. It is also found the role of river is more significant in the basins where each water storage component intensively interacts in the transition of seasons, or when the basins contain relatively large amount of water in river networks. According to CCR of river storage and mean ambient temperature, the basins are successfully classified into five groups corresponding to climatic regions; dry, temperate, cold, wet, and polar region. The eight global major basins are selected for the subsequent analysis, and all selected basins show well-reproduced seasonal changes and inter-annual variations in both discharge and TWS simulations. River storage is the most dominant component which explains up to 73% of total TWS variations in wet basins (e.g., Amazon and Orinoco), and it acts as a buffer which smoothes the seasonal variation of total TWS especially in snow dominated basins (e.g., Amur and Yenisei). In temperate basins, soil moisture is the most dominant water storage component, but river and snow water are still important. However, in dry basins, river storage effect is rarely found, and soil moisture explains most of total TWS variations.

Role of rivers which have not received enough attention in TWS studies is not an ignorable component along with soil moisture and snow water variations. Without the representation of river storage component, TWS simulation may not be able to properly reproduce the amplitude and seasonal variation of observed GRACE TWS.

Back to top

Session: B.5.b - Papers are solicited on advances in hydrological applications of GRACE data products
Title: Mass rates over Tibet-plateau region from GRACE temporal gravity field
First Author: Zhicai Luo
Presenter: Zhicai Luo
Co-Authors: Li Qiong

Abstract: Tibet-plateau is located in the Southwestern China, covering all the Tibet autonomous Region and Qinghai Province, western Sichuan, southern Xinjiang Uygur Autonomous Region and the parts of Gansu and Yunnan Provinces. Tibet-plateau is the main distribution area of the modern glacier in China, and covers 83% of the overall glaciers. It is of great significant for ecological environment and global or regional climate changes to study the mass variations over Tibet-plateau. The mass variations over the Tibet-plateau region are recovered from the 60 monthly GRACE (Gravity Recovery and Climate Experiment) earth’s gravity field models for the period from January 2004 to December 2008 in the paper, and then the characteristics of the mass variations are analyzed. The results show that the mass variations in the Tibet plateau region, including water storage changes and snow water variations, have notable seasonal features. The amplitude of the annual variations is approximately 5.3cm in terms of equivalent water height, and achieves the maximum in March and August each year. The long-term mass variation rates of the Tibet-plateau region change from 1cm/year in southeastern to -1cm/year in northwestern. Comparing the mass variations from GRACE to those from CPC (Climate Prediction Center) hydrological model, the similarities are shown well and the annual amplitude difference is less than 0.3cm. Finally, the time series of mass variations in three different areas are analyzed, which are the source of three rivers in eastern Tibet, Ngari in northern Tibet and Brahmaputra in southwestern Tibet.

Back to top

Session: B.5.b - Hydrology Applications
Title: Assessing regional mascon GRACE solutions over Australia
First Author: Joseph Awange
Presenter: Joseph Awange
Co-Authors: I Anjasmara;K Fleming;M Kuhn;WE Featherstone;B Heck

Abstract: Previous studies of temporal gravity change over Australia using global monthly GRACE solutions have found that the small signal expected over Australia and the leakage of signals from the surrounding ocean hinders the use of such data for estimating terrestrial water storage variations. This study evaluates the mass concentration (mascon) method, which has been proposed by the authors as one way of addressing these shortcomings. The mascon time series (10 days) is compared with two other GRACE releases; CSR RL04 (monthly) and CNES/GRGS RL02 (10 days), as well as rainfall (TRMM 3B43 product, monthly), and terrestrial water storage (WaterGap Global Hydrological Model, WGHM, monthly) time series. Spatial and temporal trends are determined using principal component analysis (PCA) and multi-linear regression analysis (MLRA), with the latter involving a temporal model consisting of a linear trend, offset and annual cyclic terms being fitted to the time series.

For Australia as a whole, the PCAŐs 1st mode shows a consistent picture with all data sets displaying a general north-south varying spatial pattern and strong annual signal. As previously found by the authors, the CSR release is more strongly affected by leakage from the north, while remnant stripping in the CNES/GRGS results is apparent. The annual signal is also strong in the PCAŐs 2nd mode for all datasets, which when combined with the 1st mode explains between 60% to 80% of the total signal in each dataset, a consequence of the dominance of the Northern Australian contribution. Examining Southwest Western Australia and the Murray-Darling River Basin in the southeast, the Southwest sees 60% to 90% of the GRACE signals and 75% of the water storage signal being explained by the 1st mode, but not so for the rainfall (47%). For the Murray-Darling basin, while 50% to 60% of the GRACE and water storage signals are accommodated by the 1st mode, only around 35% of the rainfall signal is explained, with the 2nd mode contributing between 13% and nearly 30%, a higher proportion than for Australia as a whole, suggesting more complicated temporal processes. This is to be expected as the Murray-Darling basin has been seriously afflicted by drought for much of the past decade. With regards to the MLRA, the resulting linear trends reveal unrealistic values arising from the CNES/GRGS solutions, result of the remaining stripping. The mascon, TRMM and WGHM datasets display similar linear trends for much of Australia, especially in the north, the Southwest and the Murray-Darling basin, whereas the CSR results are only similar in the Murray-Darling basin. For the annual cyclic term, the mascon, TRMM and WGHM show similar behavior, while CNES/GRGS and CSR only show the Northern response.

In summary, the 10 day temporal resolution of the mascon time series is an obvious advantage, and when compared to the CNES/GRGS solutions (which has the same temporal resolution) and CSR, gives linear and annual cyclic trends that are closer to those of TRMM and WGHM. This suggests that GRACE solutions produced by the mascon method have the potential to monitor Australian stored water changes at high temporal and spatial resolution.

Back to top

Session: B.5.b - Hydrology Applications
Title: Recent hydrological behavior of the East African Great Lakes region inferred from GRACE, satellite altimetry and rainfall observations
First Author: Melanie Becker
Presenter: Melanie Becker
Co-Authors: W. Llovel; A. Cazenave; A. Guentner; Cretaux J.F.

Abstract: We have jointly analysed space gravimetry data from the GRACE space mission, precipitation and satellite altimetry data over the East African Great Lakes region, in order to study the spatio-temporal variability of hydrological changes. We find that precipitation and terrestrial water storage (TWS) from GRACE show a common mode of variability at interannual time scale, with a minimum in late 2005, followed by rise in 2006-2007. We argue that this event is due to forcing by the strong 2006 Indian Ocean Dipole on East African rainfall. We also show that GRACE TWS is related to the El Nino-Southern Oscillation cycle. Combination of the altimetry-based lake volume with TWS from GRACE over the lakes drainage basins allows estimating soil and underground water volume variations, with results in good agreement with the WGHM hydrological model outputs.

Back to top

Session: B.5.b - Hydrology Applications
Title: Groundwater component of land water storage in Asia from GRACE and hydrological modelling
First Author: Melanie Becker
Presenter: Melanie Becker
Co-Authors: : J.P. Vergnes; A. Cazenave; V. Tiwari ; A. Guentner

Abstract: Total water storage (TWS) variations from GRACE have been analyzed since mid-2002 over 6 large river basins in Asia (Indus, Ganges, Brahmapoutra, Irrawady, Mekong and Yangtze). Over the 6-year time span of analysis, a negative trend in TWS is reported for several basins. As proposed by two recent studies (Rodell et al., Nature, 2009 and Tiwari et al., GRL, 2009), over the river basins of North India, ground water depletion for domestic needs and irrigation may be mostly responsible for the large negative trend reported by GRACE. Here we develop a similar approach (although with different hydrological modelling to separate the ground water component from TWS) and extend the area of investigation to additional river basins of Asia. In addition to specific regions of the Ganges basin, negative trends in TWS are found along the northern portion of the Indus basin and Mekong. We next discuss groundwater change and the respective contributions in TWS of climate variability and anthropogenic forcing for each of the 6 basins considered in this study.

Back to top




UT Sponsored Travel

Sessions & Abstracts

Meeting Program

Meeting Proceedings

Past Meetings