I would like to explore whether we need to do the radiometric terrain correction for time series analysis of Sentinel-1 data or whether we can do the time series analysis in the radar geometry and do the conversion to ground geometry only with the end results.
The main idea is, that the RTC algorithm is quite costly and therefore we would like to skip it if possible. When we use a time series of S1 data from the exact same orbit the differences in the looking geometry shouldn't be so large, that we would need to correct for them, when we are only interested in larger changes in the backscatter time series.
To test for that we would need to be able to align the different Sentinel-1 GRD datasets from the same relative orbit with slight changes in the satellite position to a common radar geometry grid.
@aalling93 this is the topic which we discussed at our flight back from the azores.
I would like to explore whether we need to do the radiometric terrain correction for time series analysis of Sentinel-1 data or whether we can do the time series analysis in the radar geometry and do the conversion to ground geometry only with the end results.
The main idea is, that the RTC algorithm is quite costly and therefore we would like to skip it if possible. When we use a time series of S1 data from the exact same orbit the differences in the looking geometry shouldn't be so large, that we would need to correct for them, when we are only interested in larger changes in the backscatter time series.
To test for that we would need to be able to align the different Sentinel-1 GRD datasets from the same relative orbit with slight changes in the satellite position to a common radar geometry grid.
@aalling93 this is the topic which we discussed at our flight back from the azores.