Sentinal 1 data, soil moisture

I am not doing any ground moisture content test as i am interested in detecting the percentage change in the moisture from month to month for 2 years, so how can i interpret the sigma0_vv_db data into moisture percentage, i am getting the values from 0 to 1.i have the monthly sentinal data for past 2 years.

There is no direct relationship between soil moisture and backscatter intensity. You first have to eliminate the impact of roughness and other factors by field measurements or long time series.

I think this one is essential: Soil moisture identification -Sentinel

1 Like

Thank you so much for your guidance sir , as i am trying to collect the ground data and also i did change detection by subtracting the before monsoon data to after monsoon data as a result i am getting the variation in the moisture scale interms of pixel value.

Hello sir as i got the ground soil data of 30 training points from theta probe in terms of soil moisture % and i tried to correlate with the sentinal 1 data VV(bd) , i am getting zero corellation. please help me in this as i am confused

you have to remark that moisture is just one factor contributing to backscatter intensity. It also depends on roughness and vegetation cover of the surfaces which, as I expect in your case, superimpose the variation of the moisture.

Were the images acquired at the same time as the field data was collected?

@johngan has outlined this topic quite nicely: Soil moisture identification -Sentinel
You have the advantage that you have measurements, but the ambiguity of backscatter intensity remains a problem.

Thank u for the reply sir, yes sir the image is of same data as of ground data taken.

Interesting study, recently published:

1 Like

Thank you for the paper sir.

Great example on the impact of soil moisture on SAR backcatter:

1 Like

Hello sir , my VV (db) values of sentinal 1 is ranging in between 0 to 1 for all the 30 watersheds for different seasons , is it correct or my procesing is going somwhere wrong, as i go through different research papers the values are in negative. Please guide me in this

Capture1

initially, calibrated SAR backscatter is scaled so that 1 represents a fully isotropic scatterer.
However, as most of the values range between 0. and 0.1, the data is translated into log-scale. You can to this by right-clicking the raster and selecting “Convert to/from dB”

Then you will have values ranging between -40 and +5 (roughly) as used in many studies.

Some more details on this: Classification Sentinel-1 problems with MaxVer

Thank you so much sir

values are ranging between -24 to +3

that is fine and within the expected range

thank you soo much sir :slightly_smiling_face:

hello sir , i finalized the parameters like this

, can i use this for another watershed moisture calculation, where i dont have any ground data, please guide me in this sir

That entirely depends on the similarity of these two areas. You can only transfer these observations if they were acquired on bare soil with no vegetation cover and you find the same soil types (texture, depth, tillage…) and climate conditions in your second area. Overdose you are just comparing the backscatter variations of the surface cover, e.g. open and dense shrubs.

You know your study areas better than me :slight_smile:

Thank you for the reply sir, it is acquired in bare soil with no vegetation cover.
But the area i am comparing is for for different 11 watersheds in my state which are randomly selected and have different soil types and textures, so what there is any other emperical model so i can correlate up to some extent, please guide me in this sir :slightly_smiling_face: