New Sudan data

[ivanzvonkov]

This PR:

• Adds new Sudan data
• Adds GEE script for generating crop mask for region
• Reruns intercomparison

GEE Script snapshot: https://code.earthengine.google.com/2f5d1453e238e6789f64a0fecabcb3eb

Note: Currently exporting the map doesn't work because it's too large. Recommended next step is to either use higher maxPixel parameter or add code to split region in two since middle region is not relevant.

Output:

Updated intercomparison: https://github.com/nasaharvest/crop-mask/blob/New-Sudan-data/maps/Sudan_Gedaref_Darfur_AlJazirah_2022/intercomparison.ipynb

Issue 1: ROI specified by GEE code does not match sample allocation

@Gedeon-m-gedus GEE code:

var border = ee.FeatureCollection("FAO/GAUL/2015/level1").filter(
  ee.Filter.inList('ADM1_NAME', 
    ['Gadaref', 'Western Darfur', 'Southern Darfur', 'Al Jazeera']
  )
);

Training pts plotted do not fall within boundaries

Issue 2: Intercomparison has some bug which duplicates samples A LOT

@hannah-rae

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ivanzvonkov]

@Gedeon-m-gedus is it possible to get the ROI from FAO or no?

[Gedeon-m-gedus]

@Gedeon-m-gedus is it possible to get the ROI from FAO or no?

@ivanzvonkov I think the FAO names are kinda different from the names our stakeholders gave. FAO boundaries are different from the real boundaries we want.

[ivanzvonkov]

I have not paid close enough attention with other datasets, investigating duplicates now and will push fix in another PR. @hannah-rae

[ivanzvonkov]

@Gedeon-m-gedus I updated the GEE script to use your provided ROI and some more points are included, however there are still several points in the North East that are not in the ROI. Let me know what you think of this.

[ivanzvonkov]

Going to merge, can update the ROI in separate PR if need be.