You must have a javascript-enabled browser and javacript and stylesheets must be enabled to use some of the functions on this site.
 
 

 

Modeling of Atmospheric Effects in InSAR measurements with the Method of Stochastic Simulation

Xiaoli Ding(1) and Zhiwei Li(2)

(1) The Hong Kong Polytechnic University, Hung Hom, KLN, Hong Kong
(2) Hong Kong Polytechnic University, Hung Hom, Hong Kong, Hong Kong

Abstract

Atmospheric effect is an important error source in InSAR measurements. Approaches that have been proposed for mitigating the atmospheric effect include, (1) selecting SAR image pairs acquired under favorable atmospheric conditions, (2) averaging SAR interferograms and (3) calibration with external data sources. The first two methods are not preferred as some images may be wasted or the temporal resolution of measurements is reduced. As for the method of calibration, since in general the spatial resolutions of the external data are lower than that of the SAR images, interpolation of the data to the grid of the SAR image space is required. The quality of the calculated atmospheric effects directly depends on the quality of the interpolator.

Among the several interpolators available, the Kriging interpolator has been preferred by a number of researchers. However, the Kriging interpolator only provides a local best estimate, but not the global best. The method often overestimates the small values while underestimates the large values (the so-called smoothing effect). It does not reflect the spatial variability of sampled data as modeled by the covariance or semivariogram.

The method of stochastic simulation however reproduces the spatial variability and statistics of the sampled data evenly over the study area. It provides the global best estimate. It is proposed to use the method to generate the grid of atmospheric delays to calibrate SAR images/interferograms. The approach is tested with two ERS images covering part of Hong Kong. The atmospheric data used in the study include those calculated from 6 continuous GPS (CGPS) tracking stations, 27 evenly distributed automatic meteorological stations and one radiosonde station. The radiosonde data are first used to evaluate the empirical tropospheric models and select one that is most appropriate for the region. The tropospheric zenith delays (ZNDs) at each of the meteorological stations are then calculated with the model. In addition, the tropospheric ZNDs at the CGPS stations along with other geodetic parameters such as the station coordinates are resolved from the CGPS observations. The ZNDs estimated from the GPS stations and the ground meteorological observations are then combined to estimate the grid ZNDs in the two acquisitions based on the method of stochastic simulation. The DEMs generated from the InSAR results with and without implementing the models are compared with each other and with an existing DEM. Statistical analysis and testing are carried out to examine the significance of the atmospheric errors in the results.

 

Workshop presentation

Full paper

Keywords: ESA European Space Agency - Agence spatiale europeenne, observation de la terre, earth observation, satellite remote sensing, teledetection, geophysique, altimetrie, radar, chimique atmospherique, geophysics, altimetry, radar, atmospheric chemistry