H22C-01 10:20h
High-resolution regional recovery and validation of GRACE hydrological signals
We investigate a new method and its results to estimate the GRACE time-variable gravity field with enhanced temporal and spatial resolutions. The method is based on a regional inversion of in situ (on-orbit) gravitational potential difference estimated through the conservation of energy principle [Jekeli, 1999; Han, 2004]. Unlike the contemporary method to extract the time-variable gravity signal from the GRACE Level-2 (L2) data product in the form of monthly geopotentials, our method utilizes the GRACE Level-1B (L1B) data (KBR range-rate, precise orbit, accelerometry, and attitude data) and geophysical inversion. It provides the regional time-variable gravity signals with simultaneous adjustment of other parameters such as inter-satellite orbital position and velocity vectors. As a result, we improve both the spatial and temporal resolutions of the gravity estimates and thus reduce the temporal aliasing effect. We present the processing results as a demonstration based on four months of GRACE L1B data, and the analysis based on the recently available two years of data. Using the determined gravitational potential difference profiles, the water mass redistributions of two of the world's largest river basins, the Amazon and Mississippi, are studied. The terrestrial water mass variations are estimated every 2 degrees in latitude and longitude, and for every 15 days. Results are presented to assess the accuracy and resolutions of the regional variations with the L2 data products and hydrological models such as the NOAA CPC and other available in situ data and models. In particular, river gauges and other data will be used to validate the observed regional GRACE hydrological measurements.
H22C-02 10:35h
ICESat Measurements of Rivers and Lakes
Although ICESat was designed to measure ice, land and atmosphere characteristics, the Geoscience Laser Altimeter System (GLAS) aboard ICESat has gathered valuable range data over oceans, lakes, and rivers. For example, the path of the laser spot along the ground intersects the confluence of the Amazon and Tapajos rivers in Brazil, and follows Rio Tapajos for about 150 km upstream. The laser spot measurements repeat with a 170 m separation at 40 Hz, and have an approximate 70 m diameter. The noise level (Root-Mean-Square) of the 40-Hz Rio Tapajos elevations is about 3 cm, after accounting for downhill slope. Further examples from GLAS are shown for elevations over lakes and other rivers having a north-south orientation (e.g, Mississippi). ICESat demonstrates the potential for gathering a globally consistent set of hydrological elevation measurements with known accuracies.
H22C-03 INVITED 10:50h
RIVER AND LAKE FROM RADAR ALTIMETRY SAMPLE PRODUCTS FOR HYDROLOGISTS
Recent research in the application of altimetry for monitoring river and lakes levels showed the advantage of using data derived from satellite with a global coverage and repeated sampling in addition to ground data. The European Space Agency has initiated a "River and Lake" project to provide the scientific community with easy-to-use, effective and accurate river and lake height measurements from both ERS and ENVISAT satellite altimeters. De Montfort University (DMU, UK) has developed a system to obtain an estimation of the River and the Lake height from both ERS and ENVISAT data. The ambition is to obtain 10 years of data of worldwide coverage of large rivers and lakes and to make available to Hydrologists River and Lake height products in near real time, i.e., in less than 3 hours after the measurement. Sample data will be distributed on CD-ROM at the conference as well as tools to visualize them.
H22C-04 INVITED 11:05h
Surface waters monitoring by satellite altimetry
Water level measurement by satellite altimetry has been developed and optimized for open oceans. Nevertheless, the technique is now applied to obtain water levels of extensive inland seas, lakes, rivers, floodplains and wetlands. Several satellite altimetry missions have been launched since the early 1990s : ERS-1 (1991-1996), Topex/Poseidon (1992- ), ERS-2 (1995- ), GFO (2000- ), Jason-1 (2001- ) and ENVISAT (2002- ). We have developed a global data base of water level time series over lakes and rivers based on satellite altimetry. Most of the water level time series of the data base are constructed using the Topex/Poseidon GDRs data, but a number of lakes water levels are based on ERS, GFO and Envisat data. To construct water level time series on rivers, we need to define virtual stations corresponding to the intersection of the satellite track with the river. For that purpose, we select a rectangular window taking in all available along track 10Hz (for T/P) altimetry data over the river area corresponding to at least one orbital cycle. The coordinate of the virtual station is defined as the barycenter of the selected data within the window. After rigourous data editing, all available 10 Hz data of a given cycle are combined into a single measurement through simple numerical averaging. At the time of writing, about 50 lakes and 50 virtual stations on rivers worldwide are available. The data base will be regularly fed by new rivers and lakes water level time series. In addition to revealing the spatial and temporal signature of climate variability on water levels, systematic use of satellite altimetry in large river basins might support initialization and verification of models used in forecasts of hydrological variability, and, possibly, estimates of river discharge where rating curves can be established by surface-based methods.
H22C-05 11:20h
Calculation of River Discharge and Prediction of Lake Height from Satellite Radar Altimetry: Example for the Lake Chad Basin
The application of satellite radar altimetry to the determination of lake and river elevations has been used in numerous projects, and is well validated. Here, we show that with the aid of ground-based information, this technique can be extended to determine river discharge and predict downstream lake and marsh height. The Lake Chad basin provides an ideal case study due to its well-known hydrology, complex lake and marsh morphology, and because prediction of lake and marsh height has been identified as potentially useful to people living in the region. Altimetric stage measurements from the TOPEX/POSEIDON satellite, at the Chari/Ohuam confluence estimate river discharge about 500 km downstream at N'Djamena 10-days in advance (r$^{2}$ = 0.9611). Via simple linear correlation methods, the stage measurements successfully estimate the height of the permanent waters of the lake (600km downstream) 39 days in advance (r$^{2}$ = 0.9297). Predicting the water height on the western marshes of the lake-bed is poorer (r$^{2}$ = 0.7958) due to a change in response time of the local stage to the seasonal floods coincident with an observed increase in mean water level in the latter half of the 1990s. Before 1997 a 96-day phase lag results in the best fit (r$^{2}$ = 0.6463). After 1997 the best fit is obtained with a 66-day phase lag (r$^{2}$ = 0.8139). The excellent river discharge and lake height predictions show that the altimetry is a useful tool where ground-based data is difficult to obtain and where rapid water resource assessment is desirable.
H22C-06 INVITED 11:35h
Assimilation of remote observations of surface water into large-scale hydraulic models
Satellite observations of discharge and surface water storage in a given basin are usually widely spaced in time because of the long overpass repeat times for sensors designed to give global coverage. In addition, along-track instruments, such as radar altimeters, may only give a partial coverage of the river basin in space. Such data give an important, but ultimately limited, understanding of surface water dynamics in large river basins. Two-dimensional hydraulic models provide a potential method to interpolate discrete observations in both time and space to provide a more highly resolved view of flow that can be used to drive other process models (for example, models for biogeochemical fluxes). However, the application of such codes to large basins has been problematic because of a lack of: (1) computationally efficient two-dimensional algorithms capable of application at global basin scales and (2) topographic data sufficient to parameterize such codes. Both these constraints have been overcome recently through numerical developments and the release of consistent global topographic data from the Shuttle Radar Topography Mission (SRTM). In this paper, we describe a modelling scheme capable of using the SRTM data and its application to a 280km reach of the Amazon in Brazil at fine spatial resolution (270m). Remote observations of discharge and water surface elevation can then be assimilated into the code and used to improve predictions of flow and inundation at fine space and time scales. We then demonstrate the potential of this model to characterise the value of remotely sensed observations of water level and inform process studies in large river basins.
H22C-07 11:50h
A Virtual Surface Water Satellite Mission: Identifying Key Science Issues
Here we present results from a virtual mission (VM) to explore the feasibility of remotely measuring surface water heights, slopes and inundation extent. The purpose of this presentation is to demonstrate how the VM can guide mission design in order to address key questions regarding surface water dynamics and Earth System interactions; and additionally, to assess how proposed mission specifications affect uncertainty in addressing those science issues and help identify new ones. The VM group has simulated inputs to the LISFLOOD floodplain model using the Variable Infiltration Capacity (VIC) land surface model, and has generated dynamic floodplain storage simulations for analyses. LISFLOOD output was further processed to simulate interferometric response given specified orbits. The resulting floodplain height images were explored with a view towards spatial and temporal measurement frequencies required to address a series of key issues for monitoring surface water hydrology. In particular, we examined tradeoffs between measurement frequency and accuracy for monitoring water heights, slopes, storage changes and inundation extent within selected regions of floodplain. We further investigated how the height and slope information could be used with ancillary data to predict river discharge. Results will have implications for a global-scale surface water monitoring mission, and related issues in biogeochemical and global water cycles.
H22C-08 INVITED 12:05h
Measuring Surface Water From Space
The requirements for the measurement of surface water from space have been reviewed by Alsdorf and Lettenmaier (Science, 301, 2003). These requirements include monitoring the water level to centimetric accuracy with a spatial resolution on the order of 100 meters; imaging of fresh water bodies with a similar spatial scale; determining the slope rivers with an accuracy of about 1cm/1km; temporal sampling which varies from weekly, for arctic rivers, to a few weeks, for tropical rivers like the Amazon. In this work, we review the ability current remote sensing technology to meet these requirements. Among the technologies examined will be radar altimeters, such as the NASA TOPEX altimeter, lidars, such as IceSat, and interferometric radars, such as SRTM or the forthcoming Wide-Swath Ocean Altimeter. We conclude in our study that the requirement for suitable temporal and spatial coverage of heights leads to a choice of either a swath instrument or a constellation of nadir profiling instruments. If, in addition, one requires co-registered delineation of water bodies and water topography plus the ability to observe through clouds, only an interferometric radar system is currently able to meet all of the desired science requirements. In the final part of this study, we present a design for a Ka-band interferometric system which is feasible with currently available technology and which has been optimized for the measurement of surface water. We discuss in detail the error sources and calibration of such an instrument. We also present a mission and orbit scenario capable of meeting the stringent space time sampling requirements detailed above. Finally, we present the results of an instrument simulator which is able to produce data typical of what could be obtained from a spaceborne mission and demonstrate the measurement capabilities over realistic river drainages.