HR: 0800h
AN: G51C-0857 [Abstracts]
TI: Using spatially-variable wind fields derived from GPS zenith wet delay timeseries to compensate
atmospheric phase signatures in SAR interferograms
AU: * Onn, F
EM: phae@stanford.edu
AF: Electrical Engineering Department, Stanford University, U.S.A, 322 Packard Building
350 Serra Mall, Stanford, CA 94305
United States
AU: Zebker, H A
EM: zebker@stanford.edu
AF: Electrical Engineering Department, Stanford University, U.S.A, 322 Packard Building
350 Serra Mall, Stanford, CA 94305
United States
AB:
Fluctuations in the distribution of water vapor near the surface of the earth causes variations in neutral atmospheric
refractive index. These fluctuations appear as excess delays in Global Positioning System (GPS) signals and as phase shifts
in Interferometric Synthetic Aperture Radar (InSAR) images. Because the neutral atmospheric medium affects the propagation of
GPS and SAR signals in comparable ways, we use timeseries of observations of zenith wet delay (ZWD) from a network of
continuous GPS stations operating in the area imaged by a spaceborne SAR to correct atmospheric phase signatures observed in
a radar interferogram of the
study area. We interpolate spatial samples of GPS ZWD
to derive maps of atmospheric delay which we subtract from the observed atmospheric phase in the SAR interferogram. We
interpolate ZWD samples recorded in a virtual network of GPS data generated by applying Taylor's ``frozen-flow'' hypothesis
to ZWD measurements recorded before and after the SAR acquisition instances. The ``frozen-flow'' hypothesis
is used in a stochastic transport model for integrated refractivity fields, which assumes that the observed timeseries of ZWD
at a fixed GPS site consists of a superposition of ``frozen-in'' refractivity fields moving across the study area under the
action of a slowly-varying mean wind field. The measured relative time delays between pairs of ZWD timeseries permit
calculation of the wind field as a function of time. These spatially-variable wind field estimates are then used to translate
GPS ZWD measurements observed before and after the radar acquisition times to equivalent spatial samples on the grid of the
radar interferogram. Thus, we infer a denser distribution of ZWD samples from GPS than is possible by considering GPS delay
measurements acquired only at the SAR observation times. We find that the atmospheric delay map generated by
interpolation of the virtual network of GPS ZWD observations
results in lower rms error than the corresponding map generated from GPS data recorded only at the satellite SAR times. We
compare the resulting spatial disribution of GPS ZWD samples with colocated observations InSAR differential to verify our
estimates of wind fields. We also compare our estimates of wind with in situ measurements of surface layer wind obtained from
National Weather Service (NWS) meteorological stations operating in the study area.
DE: 0689 Wave propagation (2487, 3285, 4275, 4455, 6934)
DE: 1220 Atmosphere monitoring with geodetic techniques (6952)
DE: 1240 Satellite geodesy: results (6929, 7215, 7230, 7240)
DE: 1295 Integrations of techniques
DE: 3270 Time series analysis (1872, 4277, 4475)
SC: Geodesy [G]
MN: Fall Meeting 2005