Process data

[mirestrepo]

• [x] First 2 weeks of August to focus on running the full pipeline with new data

[hollandjg]

We ran the floe size distribution processing on the dataset provided by the lab:

The best comparison at this stage will be the original domain that Ellen used for the paper. The region name is “Beaufort Sea”. We do want it to run with both aqua and terra. [...]

Dateranges we can examine: I picked out a couple dates that will be good test cases. 2006-05-05 and 2008-07-14. To test the date range functionality, you can just take the two dates I listed and make the range go from the day before to a day later.

In polar stereographic coordinates, the bounding box is defined by left=-2334051 bottom=-414418 right=-1127689 top=757892

Longitude/latitude bounding box: Top left: (-152.9892, 67.6236) Bottom left: (-124.9319, 68.3670) Bottom right: (-114.8220, 78.9424) Top right: (-168.9041, 77.5051)

After these are done, I have a set of randomly sampled small scenes that we can run. These are defined using a set of CSV files with information in each row - the same way that the sample file is made in the IFT pipeline. The case definition files are here: https://github.com/danielmwatkins/ift_cal-val/tree/main/data/ift_data/ift_case_definitions

@danielmwatkins is reviewing those data.