Deal with no pixels for MSS to TM regression

[ZZMitch]

Hey,

I am testing LandsatLinkr out for my study area and got an error when completing step 5. The two tasks (medianOffset and mss2TMCoefCol) fail with this error: Error: Error in map(ID=2_2_LM05_017030_19860815): Array: Parameter 'values' is required.

I am not sure if it is helpful but here is my script: https://code.earthengine.google.com/5c7a9d116d2f8af50d421525c1ad1ddd?noload=true

I also went and checked out the image referenced in the error and attached a jpg of it.

Thanks for building this great tool into GEE, I am excited to use it for some of my projects.

[ZZMitch]

Just a quick note that I ran through this code with another WRS1 scene (since I have 4 WRS1 scenes to cover) and got the same error (even same image being referenced). I will try the other two WRS1 scenes and see if that remains the case.

[jdbcode]

Hi Mitchell -- thanks for testing the code!

Given the amount of cloud cover in the image, my guess is that no pixels are being found for developing the relationship between the TM and MSS image after cloud/shadow masking. Let me look into only including images with < given-amount-of cloud-cover for developing the relationship.

Note that this project is low on my priorities right now, so it may take some time if the fix ends up being gnarly.

[jdbcode]

@ZZMitch I added a possible workaround to the module. Add any images that throw this error to the list of images to exclude in the run file. You'll need to do a little detective work to get the Landsat ID based on the image info provided in the error message. Here is your script with the image ID added to the EXCLUDE_IDS list: https://code.earthengine.google.com/b0b962884e600325e0f047d98d0d5e8b?noload=true
Can you try it an see what happens?

[ZZMitch]

@jdbcode I ran the updated code with the extra exclusion and both tasks completed! I will continue the steps and report any other issues if I find them.

[jdbcode]

@ZZMitch Glad to hear the work around worked!
I'm going to reopen the issue and think about a more robust solution that does not include manually adding images to the exclusion list.