Batch_Cloud_Masking.md

This tutorial will show you how to use a code to apply cloud masks to ECOSTRESS data products. This code shows an example of applying cloud masks to Land Surface Temperature (LST) products, but it can be modified for other ECOSTRESS products.

Table of Contents

What is a Cloud Mask?

Applying a Cloud Mask to a Batch of Images

What is a Cloud Mask?

A cloud mask is an image used to help determine if there is cloud presence in remotely sensed imagery. The mask is binary, meaning it either indicates the presence of a cloud or it does not. If it does indicate the presence of a cloud, that pixel can be removed from the remotely sensed image to improve the accuracy of the overall image.

Tip: Make sure you have Cloud Mask files downloaded in addition to your ECOSTRESS product files. If you do not know how to download these files, see the Downloading from AppEEARS tutorial.

Applying a Cloud Mask to a Batch of Images

1. Download the Batch_Cloud_Mask code from the https://github.com/ECOSTRESS-Tutorials/ECOSTRESS-Batch-Cloud-Mask.

2. Open your finder. Create a project folder to store all the files for this project by right clicking and selecting New Folder. Name your new folder so that you know it is the main project folder.

3. Move the downloaded code file into the project folder.

4. Move the folder with your downloaded ECOSTRESS data into the project folder.

5. Open the project folder and create a new sub folder to store the completed cloud masked file. To do this, go inside the project folder, right click, and select New Folder. Then name the folder so that you know it is for the outputs.

6. Open Visual Studio Code and use File > Open Folder… to get connected to the main project folder that contains the downloaded ECOSTRESS files, the Batch_Cloud_Mask code, and the output subfolder.

| | | |----|----|

7. In the EXPLORER tab, find the Batch_Cloud_Mask code and click on it to open it.

Tip: If you want to know more about what each line of the code does, read the comments in the code. Comments in the code are identified by #. These comments do not actually change how the code runs, but they can be helpful to put notes on how the code works for yourself or other users. This can also be helpful if you want to customize the code because it will guide you to which parts you may want to change!

Examples of comments (green text following the #):

8. Then, find the section of the code titled Set Constants for the Project. Find the variable called input_directory. Change the text that says "Replace_this_text_with_folder_path" to the path of the main folder where your ECOSTRESS files are stored.

1. To copy the folder path, use the EXPLORER panel on the left side of Visual Studio Code to find the folder you are interested in. Once you have found it, right click on it and select Copy Path. Now you can paste the path into your code. Make sure it is still wrapped in quotes and has r outside the first quote.

9. Then, find the variable called output directory. Change the text that says "Replace_this_text_with_folder_path" to the path of the folder where you want the output files to be stored. Make sure it is still wrapped in quotes and has r outside the first quote.

Example Directory Set-up:

10. Now the code should be set up to be run with your images. Scroll back to the top to the section titled Import the Libraries we Need to Apply the Cloud Mask. This is the first block of code we want to run. Click into the box with the library importing code and press Shift+Return to run it.

11. At the top of the window, a pop up will appear prompting you to select a kernel to run your code with. Click on Python Environments …

12. Select the ECOSTRESS environment that you created, or another one if you have a different one you want to use.

Tip: If you do not have an ECOSTRESS environment set up, follow the Creating an Environment tutorial to make one.

13. Let the code run for a few seconds. You will see the seconds counting up in the bottom left of the cell. You will know it is done when a green check mark appears.

14. Continue this process of running each block of code, in order from top to bottom, by clicking into the module with the code and pressing Shift+Return.

    1. The Collect File Names section of the code will return a table with the date and time of the image, LST image file path, the associated Cloud Mask file path, and the output file path that the masked image will be saved to.

Example:

2. The Apply the Cloud Mask to the LST Image section of the code will return the name of the masked file and its missing proportion. If the missing proportion is too high, it will give you a message saying “Low quality scene. Skipping this file.” and that file will not be saved to the output folder.

Example:

15. Once your code has run, check your outputs folder to make sure the files have been saved there correctly.

Tip: *Read if you do not want to skip files OR you used a polygon shapefile*

If you used a non-rectangular polygon shapefile instead of a rectangle selection or shapefile to download ECOSTRESS data, you will need to adjust your code so that it does not skip all of your images. Part of the code filters for high quality images by skipping files that have more than 50% of the pixels missing. However, when your image is clipped to the shapefile you uploaded, it treats the surrounding area as missing pixels. This will cause your code to flag all of your images as having a high proportion of missing pixels, and thus they will not be saved. Alternatively, you may want to keep all your images, regardless of their missing proportions. To fix this, look in the Apply the Cloud Mask to the LST Image section, and find this part of the code:

Highlight that section of the code and press Command+/ to comment it out. This means it will be ignored when the rest of the code runs. You will know you have commented it out when it looks like this:

Then you can proceed with the code as normal!

You have now cloud masked a batch of ECOSTRESS images!