InSAR measurements of volcanic deformation at Etna – forward modelling of atmospheric errors for interferogram correction
Rachel Holley(1) , Geoff Wadge(1)
, and Min Zhu(1)
University of Reading,
3 Earley Gate, Whiteknights Road,
Berkshire, RG6 6AL,
Sicily’s Mount Etna is one of the most active volcanoes in the world, and has been the subject of numerous InSAR deformation studies. However the high relief results in strong atmospheric effects, both from hydrostatic changes in the water vapour distribution with height, and from the strong turbulence in the wake of the summit. These are exacerbated by the area’s coastal climate. A variety of techniques have been proposed to mitigate atmospheric effects, however the majority of these require a large catalogue of previous data (for example PS and stacking), or ground control data such as GPS wet delay. To develop InSAR as a fully operational method for volcanic monitoring, particularly for sudden events in remote areas, correction techniques must not require a large set of previous data or ground-based control. Forward modelling of the atmospheric water vapour fields can potentially satisfy both these criteria.
This study uses Envisat ASAR data obtained for ascending and descending passes over Etna every thirty-five days, from October 2004. The atmospheric conditions at the time of each image acquisition are modeled numerically using the UK Met Office’s Unified Model, using the ‘nested mode’ to initialise a high resolution run from the global model data. The final model produces a 1km horizontal resolution vapour field, and the difference between the vapour fields for a pair of acquisition dates can be used to estimate and remove the atmospheric signal in the interferogram.
The MERIS sensor on board Envisat provides a total columnar water vapour product at 300m resolution, at the same time as the ASAR data are collected. The MERIS can be used for direct correction of the interferograms, however the products are valid only for completely cloud free conditions – less than 15% of potential images over Sicily and a particular problem over the summit itself. The partially cloud-free areas of MERIS data can, however, be compared to the atmospheric model for validation purposes. We present the initial model validation results and corrected interferograms for data acquired in the first year of the study.