G41D-01 INVITED
A Strategic Look at Exploiting Current and Future Satellite Radar Altimetry Missions for Hydrology and the Evolution of the River&Lake Products
Since the launch of the altimeters on-board ERS-1 and TOPEX/POSEIDON 17 years ago, significant
advances in all facets of Radar Altimetry have resulted in a height accuracy over the open ocean to the cm
level. Thanks to advances in the processing of Radar Altimetry data, results are now obtainable over
surfaces for which the instruments were not designed. The Radar Altimeter was designed to operate over the
oceans and continental ice caps; however echoes are now successfully being processed from within the
continental land masses. Over inland water bodies such as Rivers and Lakes, the measurements of both the
Radar Altimeter and Radiometer are degraded by the presence of land; however it is recognized by the
global community that useable results can be obtained in Continental Hydrology by dedicated reprocessing
of the raw altimeter measurements and careful use of environmental corrections. The European Space
Agency has launched a research initiative, "River&Lake", focused on developing two special user products,
one aimed at hydrologists and the other at altimeter specialists.
This paper will report upon the strategic outlook for exploiting the current and future potential of Radar
Altimetry missions. Particular attention is paid to their support to Hydrology, their mission requirements and
the potential evolution of the River&Lake products, currently at the stage of a pilot demonstration
experiment.
The interim outcome of the River&Lake Project, among other international and national activities, was
delivered at the 2nd Space for Hydrology Workshop held in Geneva in November 2007. The objectives of this
event were to review and acknowledge the important progress made recently in using Satellite data for
Hydrology, juxtaposed with in-situ data and the modelling effort. The sessions incorporated characterising the
difficulties of present capabilities, what lessons have been learned from overcoming various issues,
identifying currently unresolved issues, determining actions required in the future in addition to consolidating
the requirement for a dedicated surface water monitoring mission. A Summary of the findings and the
recommendations of the 2nd Space for Hydrology Workshop will be provided in San Francisco.
http://earth.esa.int/riverandlake , http://earth.esa.int/hydrospace07
G41D-02
Global Inland Water Levels From Multi-Mission Satellite Radar Altimetry: A Quantitative Assessment of Current Capabilities and Future Potential
Multi-mission satellite radar altimetry makes a unique contribution to the monitoring of global inland surface water; existing datasets already allow derivation of decadal timeseries over many targets worldwide, and these data are utilised both for climate change research and to inform water resource management. As the number of gauged catchments continues to fall, the importance of a global remote sensing measurement capability becomes ever more critical. The key to unlocking this potential is to retrack the complex waveforms returned from inland water targets, to identify and discard echo components returned from targets not directly beneath the satellite, and to discriminate successfully between wet land and inundated surface. This paper presents a global assessment of current capabilities, utilising an automated methodology to grade more than 50,000 multi-mission altimeter derived time series and hence derive quantitative estimates of the extent to which current and past altimeters can measure the earth's inland water surfaces. The best performance is found to be from the EnviSat RA-2, where 8337 (33 percent of targets acquired by the instrument) return useable timeseries of heights when correctly retracked. The enhancement of this unique capability anticipated from the series of proposed future missions (including CryoSat2 and Sentinel-3) is discussed, and the key contribution to global climate change monitoring is demonstrated.
G41D-03 INVITED
Hydrological Effects in the EarthScope Plate Boundary Observatory
The dense network of 1,100 continuously operating GPS stations in the Plate Boundary Observatory (PBO)
is providing high quality position time series. Data are processed by PBO Analysis Centers at the New Mexico
Institute of Mining and Technology and at Central Washington University. The results are combined by the
Analysis Center Coordinator at the Massachusetts Institute of Technology and are made available from the
UNAVCO Data Center in Boulder. Analysis software of Langbein, 2008, was used to estimate secular trends
and annual variations in the time series. The results were interpreted in terms of hydrological loading and
poroelastic effects, from both natural and anthropogenic changes in water storage. The effects of monument
stability were also considered. The density of PBO observations allows for the identification of spatial
patterns that appear coherent over relatively broad areas. Vertical annual signals of 8-10 mm peak-to-peak
amplitude are evident at stations in the mountains of northern and central California and southern Oregon
showing peak uplift in October and are correlated to hydrological loading. The vertical elastic loading signal,
calculated from the 0.25 by 0.25 degree community Noah land-surface model, fits the annual signal well and
appears also to model the secular trends, although the time duration of ~3 years is still limited. In contrast to
mountainous regions, stations in the valleys of California show greater spatial variability ranging from stations
with almost no detectable annual signal to stations with very large, 20-30 mm, amplitudes with peak uplift in
March. The vertical signals are temporally correlated to ground-water variations caused by pumping for
agricultural irrigation and likely are caused by poroelastic effects in the sediments rather than loading.
Annual vertical signals in southern California, where not obviously influenced from localized ground-water
fluctuations, are small with ~2 mm amplitude and may be due to unmodeled tropospheric effects on the GPS
signal. Overall, the results illustrate the potential value of using dense GPS arrays to assess and improve
hydrological modeling.
http://www.unavco.org
G41D-04
CGPS detected hydrological and anthropogenic signals in California and Nevada from NASA REASoN project
The continuous GPS (CGPS) network data products of the NASA REASoN (Research, Education and Applications Solution Network) project combine loosely constrained daily coordinate solutions from SOPAC and JPL analysis centers to generate daily a single solution under a unified reference frame (ITRF2005) for stations in Western North America. These new solutions exhibit consistent good quality and can be used to detect subtle tectonic signals, which were buried in the noise of earlier solutions. In particular, the CGPS vertical solutions reveal mixed signals of strong hydrological and anthropogenic variations and moderate tectonic variations. Separating the vertical hydrological signals from the tectonic signals becomes a critical procedure for both tectonic and hydrological applications. The new solutions also improve the reliability of the vertical terrestrial reference frame and the error model assessment of the geodetic solutions. We discuss the vertical crustal velocity field in California and Nevada from 11.5 years (1997.0-2008.58) of daily CGPS coordinate time series. After regional filtering, the vertical velocity field displays moderate (< 3 mm/yr) background deformation and isolated strong (> 3 mm/yr, up to 15 mm/yr) deformation 'spots'. For this presentation, we study the nature of the abnormal vertical motion spots. We find that most of these strong motion spots display good agreement with the published InSAR results and with studies involving aquifer volume variations. The nature of these spots is identified as hydrological and anthropogenic signals, such as groundwater volume undulation, compaction of aquifer skeleton, magma intrusion and retreat. Thus, the GPS-derived vertical station coordinate solutions provide effective and reliable input for groundwater monitoring and management. Furthermore, the vertically detected hydrological and anthropogenic signals can be used to correct non-tectonic biases in horizontal velocity solutions.
G41D-05 INVITED
Four-dimensional snowmelt distribution in the Sierra Nevada imaged with ground-based Tripod LiDAR
Studies involving the modeling of snow pack and snowmelt distribution under current conditions and potential climate change conditions often use MODIS snow cover maps for the snow cover extent (satellite imagery product with 500 m pixel postings), point data from snow pillows that provide snow water equivalent (SWE), and snow course data that provide estimates of SWE from composite measurements over a site. Systematic measurements of spatially varying snow depth as it relates to the distribution of SWE as scales 30-100 m are needed to adequately parameterize hydrologic-process models, such as the Basin Characterization Model. Repeat ground-based LiDAR (Terrestrial Laser Scanning-TLS) was used to develop sub-decimeter (3-5 cm) 3D time-series of snow melt distribution at four study sites in the Sierra Nevada: 1) the U.C. Berkeley Central Sierra Snow Lab, 2) the Castle Creek snow course, 3) an site west of Conway summit (Highway 395 north of Lee Vining), and 4) the Mammoth Mountain Snow Science Lab. Baseline TLS data at site were collected during Fall 2007 with periodic surveys during the snow accumulation and snowmelt periods through Spring 2008. The spatially dense TLS imagery (3-5 cm spot spacing) were aligned for each epoch and were used to calculate spatially varied snow surfaces, the snow volume, and snow volume change. Snow-core derived SWE measurements were used to compute the spatially varying water volume for each site. Ground-based LiDAR proved to be an excellent tool for measuring sub-decimeter scale snow surfaces and the resulting snow and equivalent water volumes for targets roughly the size of one to four MODIS pixel cells.
G41D-06 INVITED
Hydrodynamics of the Groundwater-Fed Sian Ka'an Wetlands, Mexico, From InSAR and SAR Data
The 5300 km2 pristine Sian Ka'an wetland in Mexico is fed entirely by groundwater from the karst aquifer of the Yucatan Peninsula. The area is undeveloped and hence difficult to access. The inflow through underground rivers and karst structures is hard to observe resulting in difficulties to understand, quantify and predict the wetland dynamics. Remotely sensed Interferometric Synthetic Aperture Radar (InSAR) and Synthetic Aperture Radar (SAR) data offer new opportunities to get hydrodynamic information, which is useful for wetland management. InSAR data produces temporal phase-changes of the backscattered radar signal, which can be related to the water level changes in vegetated wetlands. SAR data reveals information of surface properties such as the degree of flooding through the amplitude of the backscattered signal. We used RADARSAT-1 InSAR and SAR data to form 36 interferograms and 13 flooding maps with 24 to 48 day intervals covering the time span of October 2006 to March 2008. The dataset has a high spatial resolution of ca. 20 to 60 m. Sian Ka'an consists of a mosaic of freshwater sloughs, canals, floodplains and brackish tidally-influenced areas. Throughout most of the year, water level changes in the wetland are almost uniform, resulting in a very low fringe signal in the InSAR-observations. However, during periods of maximum water levels in the wetland, steeper gradients of water level changes are observed in the wetland's sloughs, more than in the surrounding floodplains. The data reveal that two sloughs and a canal-shaped feature are main source areas feeding water into the wetland. The maximum relative water level changes observed in the wetland are 48 cm. Tide-induced water level changes appear to occur in 3 separate areas, with a maximum relative change of 24 cm, corresponding well with tidal predictions. The interferograms also reveal information about surface water flow directions and local-scale flow divides in the wetland, which are important for understanding the wetland hydrology. These flow directions and divides are supported by structures visible on Landsat imagery. Finally, the radar backscatter amplitude and its standard deviation reveal the spatio-temporal changes in the flooding extent of the wetland. Using this method we detected that the largest flooding extent (3020 km2) occurred in September 2007 after Hurricane Dean had passed through the area. In the case of no hurricanes, maximum extent of flooding occurs in November-December, 1-2 months after the end of the rainy season (2640 to 2980 km2). The smallest degree of flooding occurs in May (1350 km2) at the end of the dry season. This information along with relative water level change maps may be used for improved hydrological modelling of the wetland.
G41D-07
Helmand River Hydrologic Studies Using InSAR and Altimetry
The Helmand River wetland represents the only fresh water resources of southern Afghanistan, and one of the largely uncharted water basins in the world. Helmand River drains approximately 40 percent of Afghanistan's fresh water which is generated primarily by melting snow, storm and rainfall. The wetland is relatively narrow with mostly marshland covers located by surrounding dry lands. Previous remote-sensing studies of wetlands (Florida, Amazon, Louisiana) involve much larger water volume, and covered by swarm forests, allowing double-bounced scattering to establish InSAR coherence to measure water level changes. Here we demonstrate the use of the JAXA ALOS PALSAR L-Band InSAR to detect small changes of Helmand River water level. 18-Hz Enivsat radar altimetry observed geocentric water level height is used, along with InSAR, to establish a link between the two sensors. Altimetry observed water level shows seasonal amplitudes on the order of 2 m, while the 2-pass DInSAR detects about 14-24 mm of Helmand River water level change over an InSAR time difference of two months. There is good agreement between InSAR and ENVISAT altimetry observed water level change over the InSAR data span. Finally, we use the satellite water level height, 30-m SRTM DEM, TRMM precipitation, NCEP soil moisture, and the case-based reasoning (CBR) approach, to predict hydrologic climatology in the study region.
G41D-08
GPS Multipath as a Soil Moisture Sensor
Measurements of soil moisture are required to study the water and carbon cycles. A global network of in situ soil moisture stations would be helpful to supplement datasets from satellite sensors, but such a network does not currently exist. In contrast, the GPS community has installed thousands of GPS receivers in soil, and is likely to install many more in the next decade. Could some of these GPS sites be used to measure soil moisture? We show results from an Earthscope Plate Boundary Observatory site (P041) where GPS reflected signals are strongly correlated with in situ soil moisture sensors buried at a depth of 2.5 cm. Limitations of the technique will be discussed.