S1 GRD product - retrieve soil moisture


as I understand there are no complete algorithm for retrieve soil moisture from GRD product available in s1tbx. Right? Can you suggest links how to process images to soil moisture?
I will be grateful for any help.

How are the incidence_near and incidence_far calculated
How are the incidence_near and incidence_far calculated

I´m on the same, trying to detect a flooding! But Im starting on Sentinel…keep in contact!



GRD contains intensity in one or two polarizations. You can work this into calibrated geocoded backcattering coefficients (sigma_0) using the s1tbx. To retrieve soil moisture, you would need to apply a model that relates soil moisture to sigma_0, taking into account the SAR parameters (frequency, polarization, viewing configuration). This is not available in s1tbx (yet:-). Problem is that these models usually include other parameters, e.g. for soil material, soil surface roughness. You end up having more parameters in the model than “degrees of freedom” related to the measured SAR data. Also, these models assume a bare surface, which is not always realistic. For single SAR images, the problem is under-determined. Unless you make a number of assumptions, e.g. the target you look at is indeed bare and has a known surface geometry, it is more or less impossible to directly estimate soil moisture. If you plug in other data, like a co-timed optical image to confirm that the area is bare, meteo data to have a reasonable idea about actual dry/wet conditions, soil maps to understand local soil type, you should be able to get at least some relative indications on soil moisture. Time series, interferometric coherence and ascending/descending combinations would get you somewhat further (more degrees of freedom), though mostly into relative differences and you have to make additional assumptions, for instance that soil geometry does not change.

This problems is not different from other “inversion” problems. I am always amazed how much can be “read” from simplistic indices like NDVI in optical data. What is great about Sentinels, though, is that we can now work with endless amounts of data sets and complimentary optical and SAR series, getting closer to working solutions.


Thank you for your comprehensive answer.


Hi Govard, did you end up processing your image to soil moisture? if so, can you tell me, how did you do that in SNAP?



hi Govard, could you please tel me the procedure you followed with the S1TBX /SNAP to identify soil moisture.


Did you manage to carry out the process? Could you help me ?, I’m very lost and I need to get the soil moisture. Thank you