Google Earth Engine Tutorial-58: Soil Moisture Downscaling, using Machine Learning Techniques

[@HassanRezvan]

Dear Amirhossein, thank you for your video; I have a question in the time 54:50. why did you resample and add scale:1000 at this stage, where sm_10days is one of your layers in the collection variable and in the following, collection is the input of your RFR model. Isn't our ultimate goal to reach a spatial resolution of 1 km? I also read the paper and couldn't find my answer. I would really appreciate it if you take me out of the confusion; also if there is an explanation in this regard in the paper, please mention the section. Thank you!

[@amirhosseinahrarigee]

thanks for your comment. it is 1000 meters

[@habibaferchichi6978]

Thank you for you great content!, can you please add how to test the trained model?, and then validate the downscaled SMAP by computing R-squared and RMSE

[@amirhosseinahrarigee]

welcome. will try to make a video in this regard as soon as possible.

[@smrahuseynova3377]

Thanks for great content, I have one question. How we can detect oil spill in land area?

[@amirhosseinahrarigee]

Optical images for land regions work better.

[@smrahuseynova3377]

Good day Sir, I have several questions. I didn't find better source than your channel in this area, so would be glad if u help me. I want to get sea surface temperature map using Landsat 8 dataset, but I got confused which dataset I should use. I have done some research and here are my thoughts: 1. Using raw images. In that case I should convert DN values to TOA radiance first, after that atmospheric correction should be applied with equation that involves upwelling, downwelling, atmospheric transmissivity and emissivity parameters. Then water surface temperature can be calculated using another equation (that involves K1, K2). In this method I don't understand one part. Parameters for atmospheric correction should be obtained from Landsat level 2 dataset, however this dataset is mostly used for land surface temperature analysis. So is it correct to get these parameters from level 2 dataset? 2.Can I use Landsat 8 TOA dataset and in that case do I need to go through all these above steps like DN to TOA conversion, atmospheric correction and getting temperature or is it already processed dataset? Is it level 1 or level 2 dataset? 3. Can Landsat 8 level 2 dataset be used for SST mapping or is it used just for LST mapping? Please give your suggestions, Thanks for great content

[@amirhosseinahrarigee]

Thanks for your comment and glad to hear the tutorials are useful. Landsat LST image is available under collection 2 level 2 data in google earth engine platform and LST manual calculation is no longer needed.

[@smrahuseynova3377]

Thanks for great content, I have one question, can we calculate chlorophyll index values (like MODIS provides) from Sentinel 2 images? I have done chlorophyll index mapping using Sentinel 2 images, but don't know how can I get actual values, when I click specific pixel from inspector it just returns pixel values for rgb channels, would be glad if u answer

[@amirhosseinahrarigee]

Welcome. Yes, you can use chlorophyll related spectral indices. Just make a single layer for visualisation , in case you want to see only one single value.