G23A-01 INVITED 13:40h
GRACE Mission Status and Gravity Field Product Improvement Plans
Approximately two years of gravity field data products from the Gravity Recovery And Climate Experiment were recently made available to the user community. These gravity field products are dubbed Release 01, and are associated with a specific set of background models & data processing methods used in their creation. Work is presently underway within the GRACE Science Data System, aimed at updating the Level\-1 data products, and the Level\-2 background models and processing methods within the next several months, eventually to result in improved Release 02 data products. Within this context, this paper presents a brief overview of the mission status, followed by a discussion of near-term processing plans for upgrading the gravity field products from GRACE. Product quality enhancements anticipated in the next release, including improved accuracy and spatial resolution, as well as prospects for improvements in temporal sampling of the gravity field are discussed.
G23A-02 13:55h
Development and Assessment of GRACE Derived Gravity Field Monthly Solutions
Recent gravity field satellite missions like CHAMP and particularly GRACE have given us the opportunity to solve for Stokes' coefficients of the Earth gravity field on a monthly basis up to higher resolutions than before, when they were limited to very low harmonic degrees using SLR data. Nevertheless, pertinent results may depend on many particularities of data processing, on combining GPS, Kb-range, Kb-range-rate and even SLR measurements, on their relative weighting, on a priori models chosen, on the strategy applied for "instrumental" parameters relative to the GRACE K-band system or to (Super-)STAR accelerometer. The estimation of J2 is not accurate with CHAMP and GRACE data, and LAGEOS data may still be necessary to remedy this problem. Moreover, it becomes difficult to assess the accuracy of the generated models based on the different tests available to us: formal errors, internal consistency, and comparison to external data or models. We propose to synthesize some recent developments made in CNES/GFZ concerning both the strategy of processing and the assessment of models, the barotropic ocean model in particular. To that purpose, monthly gravity field solutions are computed using GRACE data, incorporating CHAMP data and LAGEOS data as well.
G23A-03 14:10h
Time Variable Gravity from Local Mascon Analysis of GRACE Data
We have analyzed GRACE Level 1-B data in 2003 and assessed a new approach for extracting time variable gravity that isolates the gravity signal in both time and space. The Level-1B satellite-to-satellite range rate (KBRR) data and accelerometry are processed in daily arcs using the precise orbit products produced by the GRACE team from GPS to calibrate both the accelerometer and KBRR data. We then adjusted select components of the intersatellite baseline vector for each data segment isolated to the region of interest. Herein, we solved for mass anomalies in 4$^\deg$ x 4$^\deg$ blocks over the Amazon and nearby Atlantic Ocean and estimate mass flux in units of cm of water over each block. We show with this approach that we can recover mass anomalies on a submonthly basis with 10 to 15 day temporal resolution. We discuss the important issues related to this solution, including the size of the mascon blocks, the weight given to the temporal and spatial constraint used to stabalize the solutions, as well as the optimal correlation in time and distance. We compare the the mascon results with solutions obtained from the more standard approach using spherical harmonics and with independent hydrology models and lake data. This technique demonstrates that sub-monthly medium wavelength mass flux phenomena are well sensed by the hyper-precise line of sight velocity data produced from GRACE.
G23A-04 14:25h
Time-Variable Gravity From Satellite Laser Ranging and Doppler Measurements: A look at the interannual and annual variations and comparison with GRACE results
Satellite Laser Ranging (SLR) derived long wavelength gravity time series analysis has focused principally on the effects of the recent large changes in the Earth's zonals, particularily J2, or the long-term secular rates, and the potential causes. However, it is also possible to estimate the shorter wavelength coefficients, including non-zonals, over monthly time scales, and to connect these with known geophysical signals. While the formal uncertainty of these terms is significantly higher than that for J2, it is also clear that there is useful signal to be extracted. We will present updates on recent gravity field variations, including the J2 evolution, as well as other components of the interannual variations of the gravity field and geophysical and climatic connections. These results will also be compared to data from the GRACE mission, which provides an independent means of assessing the SLR and Doppler derived time series. Work is presently in progress to process the SLR and GRACE data using a common set of background models within GEODYN to allow a more complete assessment. Initial comparisons have shown reasonable agreement in terms of annual and semiannual variations, as well as some component time series, such as the C2,1 and S2,1 terms.
G23A-05 14:40h
Recent Changes in the Dynamic Oblateness (J2) of the Earth: Contributions from the Earth's Subsystems and Their Implications
Earth's dynamic oblateness (J2) has been decreasing due to post-glacial rebound (PGR). However, an increase began in 1997 (Cox and Chao, Science, 2002) indicating a pronounced change in the global-scale mass redistribution process. Dickey et al. (Science, 2002) have determined that the observed increases in J2 were caused primarily by a surge in sub-polar glacial melting and mass shifts in the Southern Pacific, and Indian Oceans. Recently, the geodetic J2 series has been decreasing rather than increasing, which motivates the further study. Data series have been extended, in particular ocean loading derived from ECCO bottom pressures and hydrology (Milly, pers. comm.). Both glacial and oceanic sources had enhanced upward trends around 1998. The sums of the glacial, oceanic, atmospheric and continental hydrological contributions to J2 are compared with geodetic observations (Cox and Chao, 2002; Cox, pers. comm., 2003). The modeled excitation captures the rise of J2 in the late 1990's as well as its subsequent decrease, even though J2 contributions due to glacial melting maintain a positive slope. Thus it is not necessary for the glaciers to switch to a net positive mass-balance, in order to match the decrease in the observed J2. Special emphasis will be placed on the recent glacier results.
G23A-06 14:55h
Independent Determinations of Low Degree Gravitational Change
We examine low degree gravitational variations of C21, S21, and C20 observed by the Gravity Recovery and Climate Experiment (GRACE) satellites during the first 2 years of this gravity mission. The GRACE observations are compared with independent estimates from accurately measured Earth rotational changes and predictions from atmospheric, oceanic, and hydrological models. The 18 GRACE monthly gravity solutions, covering the period April 2002 to March 2004, show strong seasonal variability in the C21, D21, and C20 time series, and generally agree with Earth rotation-derived changes and geophysical model estimates, in particular in S21 and C20. Earth rotation-derived results agree very well with geophysical model predictions from advance atmospheric, oceanic, and hydrological models, in particular in the two non-zonal terms (i.e., C21 and S21). The differences between GRACE results and Earth rotation-derived estimates, especially in C21 are likely caused by the errors in the background models used in GRACE data processing.
G23A-07 INVITED 15:10h
Subarctic Bottom Pressure Oscillation observed by GRACE, linked to ENSO
The Gravity Recovery and Climate Experiment (GRACE) satellites have observed strong signals during 2003-04 in the meridional differences in ocean bottom pressure (OBP) in the high-latitude N. Pacific (50N minus 25N) and in the Indian Ocean sector of the Antarctic Circumpolar Current (55S minus 25 S). `Strong' means peak-to-peak amplitudes in the annual variability of 6 mbars, when averaged zonally over about 4,000 km. Reasonableness of the signal is determined by comparison with outputs from a non-Boussinesq model that does not assimilate data, as well as with a version of the ECCO model with altimetric data assimilation. Focussing on the N. Pacific signal, there is a strong hint of interannual trend in both data and model. The non-Boussinesq model is then used to analyze a 50 year long time series of model output. We find that the N. Pacific signal leads the Niño3.4 index by more than 12 months with a correlation of 0.52 in the 99% significant interval. The model also shows a moderate east-west OBP oscillation in the tropical Pacific. Unlike the subarctic Pacific oscillation, the bottom pressure difference between the eastern (Niño3.4 region) and the western tropical Pacific (warm-pool region) is in phase with the Niño3.4 index with a correlation 0.82 and has a characteristics of eastward mass-shift during El Niño events, indicating their direct effect on deep oceans, however, the GRACE satellites cannot `see' this much smaller E-W signal. Lower latitude applications of GRACE data, with signals of ~1-2 mbar, need to wait for expected improvements in GRACE fields after one or more reprocessing cycles. The E-W difference also amplifies the zonal errors in GRACE data, while the N-S differences then to cancel them.
G23A-08 15:25h
GRACE Observations over the Ocean
Monthly gravity field coefficients from the GRACE mission for the period from August 2002 to August 2004 are used to create maps of long-wavelength barotropic sea level variations (equivalent water levels) over the ocean. These maps are compared to long-wavelength maps derived from altimetry corrected for steric (non-barotropic) sea level variations. Since the steric sea level variations are based on a monthly climatology, only the seasonal cycle is examined. The altimery - steric data are referenced to the same background barotropic model used in the GRACE processing and are shifted to the same reference frame as GRACE in order to make the comparisons. Similarities and differences are discussed, with a goal to quantify where GRACE ocean observations appear realistic.
G23A-09 15:55h
Verification of basin-scale ocean mass variations from GRACE using altimetry and heat storage
Empirical orthogonal analysis of the initially released GRACE gravity fields reveals that roughly half of the variance can be attributed to a single spatial mode with a seasonal signal. Tapley et al. [2004] and Wahr et al. [2004] demonstrated that the seasonal signal in the GRACE fields is consistent with the gravity signal expected from ocean and hydrological models. Recently, the preliminary GRACE observations have been used to derive an estimate of global ocean mass variations. Chambers, Wahr, and Nerem [2004] compared this time-series to a mean climatology determined from satellite altimeter measurements of global mean sea level corrected for the steric variation. The GRACE observations show a seasonal exchange of water mass with the continents of the same magnitude and phase as the steric-corrected altimetry. We present estimates of ocean mass variations using the latest GRACE fields for each ocean basin. In addition to using a mean climatology to steric-correct TOPEX/POSEIDON and Jason altimetry data, we also derive a thermal steric height variation from monthly upper ocean heat storage estimates from the Joint Environmental Data Analysis Center (JEDAC). The JEDAC heat storage fields have been generated from observed temperature profiles (e.g. expendable bathythermograph data). We also present basin-scale comparisons using ECCO model output to steric-correct altimetry data. Finally, preliminary results of secular variations of ocean mass will be discussed.
G23A-10 INVITED 16:10h
Water Storage Variations from GRACE: Results and Comparisons
GRACE time-variable gravity measurements are a new and innovative data type for hydrology. At approximately monthly intervals, the GRACE Project constructs a new, monthly-averaged estimate of the Earth's gravity field. Month-to-month changes in these GRACE gravity fields are primarily due to changes in the vertically-integrated water column, both on land and in the ocean. We have developed techniques that convert the GRACE gravity data into spatially-averaged water storage estimates. Applying these techniques to recently-released gravity fields, we have derived regional estimates of water storage variations for specific regions of the world. We compare these estimates to those produced by macroscale hydrologic models, as well as remotely-sensed estimates of snow water. In addition to providing direct estimates of water storage change, GRACE data can be used to estimate other components of the water cycle as residuals by combining GRACE water storage estimates with estimates of other quantities, such as precipitation and river discharge, in the appropriate water balance equation. We also compare estimates of precipitation minus evapotranspiration made using this approach to those obtained from both hydrologic models and global circulation models.
G23A-11 16:25h
Mass change of the Greenland ice sheet from a climatic-glaciological model and GRACE
We use a model of melting and refreezing of the Greenland ice sheet to estimate the gravitational change signals as a function of increased temperatures on the Greenland ice sheet. The results of the model are compared to actually observed mass changes from analysis of monthly GRACE data. The overall observed GRACE 2002-4 mass change shows good agreement with mass loss estimated from the 1990's NASA PARCA project.
G23A-12 16:40h
Feasibility of Using GRACE Gravity Observations to Constrain the Annual Mass Variation of Glacier Complexes Near the Gulf of Alaska
Mountain glaciers make a significant contribution to the global hydrological cycle though they only constitute 3 to 4% of the Earth's glaciated land area. Unfortunately, determining the total annual and secular contribution from the large glacier complexes is difficult using standard glaciological and aerial altimetric methods. To obtain the annual mass balance, glaciological methods generally sample a small number of geographic sites and difference measurements taken at two times during the year. Aerial altimetry improves the geographic sampling but differences measurements taken once during the same period each year. We explore the use of the GRACE monthly fields in combination with other geodetic techniques (e.g., GPS) to examine the annual component and to better constrain the annual mass balance for the glacier complexes near the Gulf of Alaska. This data set allows for much better temporal sampling of the mass change and is an integrated measure of the mass balance across all of the glacier complexes in the region. We use a technique presented by Davis et al. [2004 Fall AGU, Session G12], which only uses spherical harmonic coefficients with a statistically significant annual component, as well as methods using Gaussian filters [e.g., Swenson and Wahr, 2002] to estimate the annual amplitude of the total mass variation over the region including most of the glaciers near the Gulf of Alaska. We present the observations as well as results from forward models processed in the same manner and discuss the ability of these approaches to obtain the annual mass variation of the glacier complexes.