SAR data are sensitive to many factors and hence require more pre-processing steps than most other remote sensing data sources, in order to generate consistent and comparative data. We know how pre-process SAR data for your purposes, regardless if your goal is to detect centimeter-subsidence due to mining or if you want to detect spring floods. Different techniques, such as interferometry, polarimetric SAR or working with backscatter data require different types of pre-processing. The pre-processing is also dependent on the sensor and its configuration for each image.

A single look complex (SLC) SAR image to the left and the multilooked SLC SAR image to the right.

Within a forest project project funded by the Swedish Space Agency, we found that the backscatter dependence on temperature could be very high and that by including weather data, the backscatter for a forest stand could be predicted with an R2 of 0.69

This and many more operations need to be considered carefully when working with SAR images.

The Sentinel-1 backscatter could be predicted with temperature as explanatory variable, reaching an R2 of 0.69 for a single forest stand in northern Sweden.