Error during soil moisture mapping

[Bikesh Jojiju]

Dear SEPAL users,

I am using Soil Moisture Mapping tool and while I was calculating statistics, it says " ModuleNotFoundError". How can I fix this?

Thank you,

B

[Daniel Guerrero]

Dear SEPAL user, 

Our apologize for the inconvenient in the statistics step, the error was already reported here: https://github.com/openforis/sepal/issues/258 and we are working actively to find the best solution.

We will keep you updated with our progress.

Best,

Daniel G.

SEPAL Team 

[Bikesh Jojiju]

Sorry to hear about the issue. Are there any alternatives I can use for soil moisture mapping?

[Daniel Guerrero]

Dear user, 

If  you have completed steps 1, 2 and 3, it means that  you already have downloaded and applied the filter to the soil moisture images.

You can download the results to your local machine and do the post-processing steps off line, please check how to exchange files with SEPAL.

Note that the results are located in: /home/sepal-user/pysmm_downloads/.

Best,

Daniel G.

[Bikesh Jojiju]

Dear Daniel,

Thanks to your help, I was able to locate and download the processed maps. However, I have some queries, and it would be greatly appreciated if you could address them.

1. Are the maps clipped to the shape files of roi provided because the maps I got were only clipped to the latitude and longitude of the roi (the maps were of rectangular shape)?
2. There were some pixels with no values (blanks). Is it because the pixels with dense vegetation were excluded?
3. What is the resolution of the output maps produced by the module? The resolution of the maps I got was around 94 meters even though it used 10 m sentinel images. Are the maps reclassified to lower resolution for accuracy or other proposes?

Once again thank you for your help.

Best regards,

Bikesh

[Daniel Guerrero]

Dear Bikesh, I'm glad you were able to find the maps.

The output maps are a mixture of images that intersect with your AOI and S1 data. It is normal to have this type of shapes since there may not be an image available for all the specified dates. That is why the last step is implemented: the module merges all images according to the user parameters and calculates the chosen statistic. The result is a single image that hopefully has reduced noise.

According to the documentation, soil moisture maps may contain noise and no-data values. This appears as a grainy or textured pattern in the image. That is one of the main reasons we apply a morphological filter. However, we cannot fully reduce all the noise.

As the model uses several regressor variables, such as GLDAS, LC_maps, elevation model, among others, to perform the prediction, the output resolution will be affected by all of them.

I hope this information is helpful to you.

Best regards,

Daniel G