Exploring the Benefits of Stratified Tropospheric Delay Corrections using Global Atmospheric Models in InSAR
Jolivet, Romain1; Agram, Piyush1; Simons, Mark1; Doin, Marie-Pierre2; Lin, Nina1; Peltzer, Gilles3
1California Institute of Technology, UNITED STATES; 2ISTerre, CNRS, FRANCE; 3University of California, Los Angeles, UNITED STATES

Phase delays accumulated in the atmosphere present a major limitation in our ability to measure ground deformation using SAR interferometry. Spatial and temporal variations of pressure, temperature and water vapor content introduce significant delay in the interferometric phase that can overwhelm any deformation signal and introduce bias in estimates of strain rates. Estimation of Atmospheric Phase Screens (APS) is therefore essential when focusing on slowly deforming areas, where Line-Of-Sight displacements rates do not exceed several millimeters per year.
Several methods have been developed in the past 20 years to mitigate both stratified and turbulent component of the APS. The stratified delay results from temporal variation in the average stratification of the lower troposphere and therefore shows a correlation with the elevation, while the turbulent delay is considered as random in space and time. Applying time-dependent processing schemes such as stacking or time series analysis can mitigate the impact of turbulent delay. However, due to the uneven temporal sampling of SAR acquisitions, there is a potential for bias in the estimation of strain rates when the stratified delay is not corrected for [Doin et al., 2009].
An empirical approach to estimating and correcting for the effects of stratified troposphere uses the phase information in the interferograms to estimate a relationship between topography and phase [e.g. Cavalié et al., 2008; Elliott et al., 2008; Lin et al., 2010]. However, such methods cannot be used when the targeted deformation signal also correlates spatially with the topography and do not account for lateral variations of the stratification. Alternatively, predictive methods take advantage of external data sets, such as local atmospheric data collections [Delacourt et al., 1998], GPS zenith delays [e.g. Li et al., 2006] or multispectral imagery [e.g. Li et al., 2012], to produce forward models of the stratified delay. These methods depend on external data sets that are not always available. Finally, some studies use outputs from dynamic modeling of the troposphere to derive the atmospheric phase screen [e.g. Foster et al., 2006]. Such methods are successful but require in situ atmospheric data to better constrain the prediction.
There are also benefits to using global atmospheric re-analysis to remove the contribution related to the stratification of the troposphere [Doin et al., 2009; Jolivet et al., 2011]. These models provide subdaily and systematic estimates of vertical profiles of temperature, pressure and water vapor content. Jolivet et al., (2011) developed a simple and efficient method to derive delay maps coincident with SAR acquisitions from a spatial and temporal interpolation of atmospheric variables provided by global atmospheric models. We build on this study to demonstrate the importance of such corrections on both single interferograms and time series analysis products. We present multiple examples of corrections, spanning different geographic and tectonic contexts, highlighting the benefits of this method.
There are also benefits to using global atmospheric re-analysis to remove the contribution related to the stratification of the troposphere [Doin et al., 2009; Jolivet et al., 2011]. These models provide subdaily and systematic estimates of vertical profiles of temperature, pressure and water vapor content. Jolivet et al., (2011) developed a simple and efficient method to derive delay maps coincident with SAR acquisitions from a spatial and temporal interpolation of atmospheric variables provided by global atmospheric models. We build on this study to demonstrate the importance of such corrections on both single interferograms and time series analysis products. We present multiple examples of corrections, spanning different geographic and tectonic contexts, highlighting the benefits of this method.
Finally, we focus on dense time series of Envisat SAR data for Mt Etna, Italy. By systematically correcting interferograms from stratified tropospheric delay, we remove a 5 mm/yr bias in the line-of-sight deformation rates between the top and the bottom of the volcano. The inferred bias spatially correlates with the local topography (i.e. therefore with the expected deformation field) and is due to seasonal oscillations in the interferometric phase that are accurately estimated using the predictions from global atmospheric models.
We recently released the software used to predict the stratified delay, called PyAPS [Agram et al., 2013]. This python-based package is free, documented, and available on line at http://earthdef.caltech.edu/.

References:
Agram, P.S., Jolivet, R., Riel, B., Lin, Y. N., Simons, M., Hetland, E., Doin, M.-P. and Lasserre, C., 2013. New Radar Interferometric Time Series Analysis Toolbox Released. Eos, 94(7), pp.69-76.
Cavalié, O., Lasserre, C., Doin, M.-P., Peltzer, G., Sun, J., Xu, X. and Shen, Z.K., 2008. Measurement of interseismic strain across the Haiyuan fault (Gansu, China), by InSAR. Earth and Planetary Science Letters, 275(3-4), pp.246257.
Delacourt, C., Briole, P. and Achache, J., 1998. Tropospheric corrections of SAR interferograms with strong topography. Application to Etna. Geophysical Research Letters, 25(15), pp.2849-2852.
Doin, M.-P., Lasserre, C., Peltzer, G. Cavalié, O. and Doubre, C., 2009. Corrections of stratified tropospheric delays in SAR interferometry: Validation with global atmospheric models. Journal of Applied Geophysics, 69(1), pp.35-50.
Elliott, J.R., Biggs, J., Parsons, B. and Wright, T., 2008. InSAR slip rate determination on the Altyn Tagh Fault, northern Tibet, in the presence of topographically correlated atmospheric delays. Geophysical Research Letters, 35(12).
Foster, J., Brooks, B., Cherubini, T., Shacat, C., Businger, S. and Werner, C. L. , 2006. Mitigating atmospheric noise for InSAR using a high resolution weather model. Geophysical Research Letters, 33(16), p.L16304.
Jolivet, R., Grandin, R., Lasserre, C., Doin, M.-P. and Peltzer, G., 2011. Systematic InSAR tropospheric phase delay corrections from global meteorological reanalysis data. Geophysical Research Letters, 38(17).
Li, Z., Fielding, E., Cross, P. Muller, J.-P., 2006. Interferometric synthetic aperture radar atmospheric correction: GPS topography-dependent turbulence model. Journal of Geophysical Research-Solid Earth, 111(B2), p.B02404.
Li, Z., Xu, W. B., Feng, G. C., Hu, J., Wang, C. C., Ding, X. L. and Zhu, J. J., 2012. Correcting atmospheric effects on InSAR with MERIS water vapour data and elevation-dependent interpolation model. Geophysical Journal International, 189(2), pp.898-910.
Lin, Y.-N.N., Simons, M., Hetland, E., Muse, P. and DiCaprio, C., 2010. A multiscale approach to estimating topographically correlated propagation delays in radar interferograms. Geochemistry Geophysics Geosystems, 11(9), p.Q09002.