B.4 - Progress in Hydrological Applications
(Moderator: Matt Rodell)

   Ole Andersen (12)
   Jean-Paul Boy (12)
   Jay Famiglietti (10)
   Andreas Guentner (10)
   Scott Luthcke (12)
   Matt Rodell (12)
   Sean Swenson (12)
   John Wahr (12)
   Discussion (28)

Ole Andersen (12)
Presenter: Ole Andersen
Co-Authors: M. Butts, F. Jakobsen, F. G. Lemoine, S. B. Lutcke, P. Berry, J. Freeman
Title: Flooding in Bangladesh from satellite altimetry and GRACE gravimetry.

Abstract: Bangladesh is heavily exposed to Monsoon flooding as an annual event with river water levels rising and falling slowly during the monsoon season, though with some pulsation. Bangladesh is located at confluence of three major rivers: the Ganges, the Brahmaputra, and the Meghna River.

In this study we investigate the possibility of applying available remote sensing data from GRACE and satellite altimetry as a supplement to existing hydrological studies. The GRACE gravity changes are analysed using a local MASCON approach derived by NASA/GSFC, solving for mass change at 10-day intervals using 4 deg X 4 deg blocks from GRACE level 1B data. Satellite altimetry over the region has been retracked using the EAPRS Expert-retracker System in order to derive height of rivers, in particular the Ganges and Brahmaputra rivers. The EAPRS system has the ability to recovers nearly unterinerrupted time series over these rivers. GRACE derived mass change from 2002 to 2004 have been studied along with altimetry for the same period.

The main forcing factors causing the floods in Bangladesh are high inflows from upstream combined with extreme high rainfall inside the country. ENVISAT has the ability of modeling the changes upstream the rivers, which is a major causing factor. G GRACE gravitmetry can model the integrated amount of ground and river water stored in Bangladesh on monthly to inter-annual scales which is an important parameter as well.

In this study we will validate and investigate the possibly use of these new datasets for studies of these large scale hydrological signals in a relatively small country like Bangladesh.

Jean-Paul Boy (12)
Presenter: Jean-Paul Boy
Co-Authors: Claudia C. Carabajal, Scott B. Luthcke, David D. Rowlands, Frank G. Lemoine and Douglas S. Chinn
Title: Retrieval of the Three Gorges Dam Impoundment Signal and the Yangtze Basin Water Cycle in GRACE Solutions

Abstract: The impoundment of the Three Gorges Dam Reservoir, on the Yangtze river (China), is being done in three episodes, the first occurred in 2003, and subsequently in 2006, to finally reach the maximum desired level in 2009. The water level at the dam increased from 55 meters above mean sea level to 135 meters in 2003, and will reach up to 175 meters in 2009, representing a total volume increase of about 39 km3 (Wang, 2000; Yang et al., 2005). We focus here on the first impoundment completed in June 2003, which corresponds to a volume of water of 17 km3 (Yang et al., 2005), and investigate if this mass change can be retrieved from different types of GRACE solutions (mascon, constrained and unconstrained spherical harmonics). In addition to the water load due to the impoundment, we describe the seasonal water changes retrieved by GRACE, and compare them to various hydrological models, such as GLDAS (Global Land Data Assimilation System) and ECMWF (European Centre for Medium-range Weather Forecasts), and to precipitation data from meteorological stations in the region. Observations show the significant effect that surface water stored within the large number of dams upstream from the Yangtze (having a total volume of about 180 km3) can have in the seasonal water cycle.

Jay Famiglietti (10)
Presenter: Jay Famiglietti
Co-Authors: M. Rodell, J. L. Chen, C. R. Wilson and K.-W. Seo
Title: Progress in GRACE Hydrology Applications from the Basin to Continental Scales

Abstract: In this presentation we review the recent progress of our science team group. Results will emphasize advances in GRACE data processing, analysis of terrestrial and ice sheet water mass changes, and new applications in groundwater remote sensing, land surface modeling and data asssimilation, freshwater discharge, and decomposition of the storage change signal into its surface water, soil moisture and groundwater components.

Andreas Guentner (10)
Presenter: Andreas Guentner
Co-Authors: S. Werth, L. Menzel, P. Dll
Title: Calibration of the WaterGAP Global Hydrology Model WGHM with Water Storage Variations from GRACE

Abstract: The traditional way of calibrating hydrological models is by adjusting them to observed river discharge data. With the availability of GRACE data, temporal water storage variations may be used as an additional calibration variable for large river basins within a multi-criterial model calibration approach.

In this presentation, work steps towards the calibration of the WaterGAP Global Hydrological Model (WGHM) with GRACE data are presented. WGHM is a conceptual global water balance model with 0.5 spatial resolution. So far, it has been calibrated against river discharge for more than 1200 gauging stations worldwide. The major steps towards the use of GRACE data involve:

  1. A sensitivity analysis of the WGHM model was performed (Monte Carlo approach with Latin Hypercube sampling) to identify the most sensitive model parameters with regard to storage change. These are potential calibration parameters when using GRACE data. It is shown that the subsets of sensitive parameters vary regionally with the importance of specific water storage compartments in the respective environments (e.g., snow, soil moisture, surface water).
  2. Considering uncertainty of model parameters and climate input data, the range of possible WGHM simulation results is compared to GRACE data of water storage variations for large river basins. For this purpose, basin filters were applied to both GRACE and WGHM after converting water mass from the hydrological model to a spherical harmonic representation. Several GRACE solutions were used for the comparison (GFZ_RL03, GFZ_RL04, CSR_RL01, CSR_RL04). The results show that GRACE-based storage variations are within the WGHM uncertainty range in many river basins while larger deviations in some other basins point out deficiencies of the model structure. Differences between the GRACE solutions usually are smaller than the differences between GRACE and the hydrological model. WGHM tends to have smaller seasonal water storage variations than GRACE.
  3. As a first approach for model calibration, the best WGHM model runs of the sensitivity analysis were examined. While model performance with regard to GRACE storage variations increased markedly relative to the standard WGHM version for several river basins, only few model parameters show identifiable optimum parameter ranges. Additionally, selected sensitive parameters were calibrated relative to GRACE data. The results indicate that a single-parameter calibration is not sufficient to obtain improved water storage estimates. Evaluating simulation results with regard to both storage change and river discharge data reveals a trade-off in model performance between both variables for most basins. In this way, the results illustrate the value of GRACE-based water storage variations to identify structural deficiencies of hydrological models.

Scott Luthcke (12)
Presenter: Steve Klosko
Co-Authors: D. Rowlands; F. Lemoine; M. Rodell; S. Klosko; D. Chinn
Title: Evaluation and Validation of Mascon Recovery Using GRACE KBRR Data with Independent Mass Flux Estimates in the Mississippi Basin

Abstract: The direct recovery of surface mass anomalies using GRACE KBRR data processed in regional solutions holds great promise for producing mass flux estimates with 10-day temporal resolution. The approach undertaken at GSFC uses a tailored orbit estimation strategy based solely on the KBRR data. We introduce a set of temporal and spatial correlation constraints to stabilize the mass flux estimates. The Mississippi Basin, with its well understood surface hydrological modelling available from the Global Land Data Assimilation System (GLDAS), which uses advanced land surface modeling and data assimilation techniques, and a wealth of groundwater data, provides an excellent opportunity to compare GRACE estimates of the mass flux in the entire hydrological column with those available from independent and reliable sources.

We have provided estimates showing that the mass flux anomalies, as represented by surface layer of water within a regional cell and have accuracies of 2-3 cm. In this paper, through comparisons with these independent estimates, we evaluate the effect on the solution of changing correlation times and distances in the constraints, altering the parameter recovery for areas external to the Mississippi Basin, and changing the relative strength of the constraints with respect to the KBRR data. The accuracy and stability of the mascon solutions are thereby assessed. The mascon solutions are shown to be very stable, especially with regard to recovery of semi-annual and longer period trends and are of the accuracy predicted. This validation confirms that mascons provide critical environmental data records for a wide range of applications.

We will also show results showing that a standard, gridded mass anomaly GRACE product will work well over all of the continental areas. Such a product is being made available through a related NASA ACCESS activity.

Matt Rodell (12)
Presenter: Matt Rodell
Co-Authors: Ben Zaitchik, Jay Famiglietti
Title: Spatial, temporal, and vertical decomposition of GRACE derived terrestrial water storage

Abstract: The nature of GRACE hydrology observations is a double edged sword. On the one hand, GRACE is the only satellite system capable of monitoring water in all of its forms, down to the base of the deepest aquifer. On the other, most applications demand data on the individual components of water storage, not a vertically integrated quantity. Furthermore, the spatial and temporal resolutions of GRACE are low relative to most other observing systems. I will discuss progress in isolating variations in groundwater and other water stocks from GRACE based terrestrial water storage changes, including the use of auxiliary water storage estimates and data assimilation in a sophisticated land surface model. Data assimilation harnesses the advantages of both observations and numerical models, resulting in high resolution output fields that are valuable for numerous applications.

Sean Swenson (12)
Presenter: Sean Swenson
Co-Authors: P. yeh, J. Wahr, J. Famiglietti
Title: A comparison of GRACE and in situ water storage estimates

Abstract: I'll discuss a comparison of GRACE water storage estimates to estimates derived from soil moisture and well-level measurements in Illinois.

John Wahr (12)
Presenter: John Wahr
Co-Authors: S. Swenson, I. Velicogna, B. Killett
Title: Some applications of post-processed GRACE fields

Abstract: The errors in the GRACE gravity fields tend to appear as north/south-trending stripes in the lat/lon domain. Those stripes can be reduced by post-processing the data, using a method developed by Sean Swenson (Swenson & Wahr, GRL, Apr 25, 2006). The post-processed fields have a much-improved spatial resolution. This talk will focus on some of the applications made possible by this post-processing.

Discussion (28)




Sessions & Abstracts

Meeting Program

Meeting Proceedings

Past Meetings

Getting to UTCSR