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


Tropospheric Correction Techniques in repeat-pass SAR Interferometry

Zhenhong Li(1) and Paul Cross(1)

(1) University College London, Gowere Street, London, WC1E 6BT, United Kingdom


Tropospheric effects, mainly excess path delays introduced by water vapour, are well known as one of the major limits of repeat-pass SAR interferometry (InSAR). Three distinct correction methods are discussed and tested in this paper: the empirical model [Saastamoinen, 1972], the use of contemporaneous Near Infrared (IR) water vapour measurements from MODIS or MERIS and the use of a network of contemporaneous GPS measurements. The test site examined is Mount Etna on the island of Sicily in Italy. The data-sets employed include surface level meteorological measurements, MODIS near IR PWV (Precipitable Water Vapour) and GPS data.

The empirical technique which has been employed previously by Hassen & Feijt [1996] and Delacourt et al. [1998] was applied to ERS InSAR measurements on Groningen and Mt Etna respectively. It is based on the use of ground-level meteorological measurements of pressure, temperature, and humidity and the model assumes that pressure exponentially decreases and temperature linearly decreases with height. For concentric features such as Mt Etna, it predicts a series of concentric rings around Mt Etna. However, evaluation of ERS InSAR interferograms show no significant correlation with the predicted fringes.

The second approach used the NASA Terra MODIS (Moderate Resolution Imaging Spectroradiometer) data with near coincident ERS passes on 28 June 2000 and on 2 August 2000 which were available within 30 minutes of each other for the first time. Unfortunately no ESA ENVISAT MERIS data was available for ERS, no ESA ENVISAT ASAR interferometric data is available to date (due to the lack of a FRINGE database) and no MERIS FR or RR products are available prior to July 2003 for use with ASAR. The difference of MODIS slant wet delays in the line-of-sight direction varied from 13 cm to 13 cm with an average of 3.0 cm. After correction, the residue of the unwrapped phase is believed to be within the quadric sum of the effects introduced by the uncertainties of the DEM, MODIS near IR water vapour, and orbit. This means that MODIS near IR water vapour is possibly applicable in some cases. Limitations of this approach include MODIS near IR water vapour accuracy (5-10%) and the requirement for cloud-free observations coincident with ERS.

The third approach utilized GPS water vapour products, The WAter Vapour Extraction Software (WAVEs), developed at UCL, was applied to reconstruct GPS 3D water vapour distributions. A first demonstration and test of WAVEs is described in Muller et al. [2003, MERIS workshop]. Application to Mt Etna showed significant reductions in residual differential InSAR fringes. However, high frequency fringes cannot be removed because of the density of the GPS network. Such an approach does have a significant advantage over the other two methods in that the 3D WV field can be continuously monitored irrespective of cloud coverage and there is no additional uncertainty associated with the conversion of water vapour to wet delays. Although national and regional continuous GPS networks are in place in a few seismically active places around the globe, there is no such network in place covering Mt Etna and many other active volcanoes around the world.

Acknowledgements. ERS data was supplied under ESA ENVISAT data grant 853. GPS data was kindly supplied by Prof. Geoff Wadge and Dr. Peter Webely (ESSC, Reading).


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