What data to access from nasadem planetary computer?

Hello!

I was able to understand the overall code to access the data. But I am not sure which areas of interest to access. Are these areas of interests for Los Angeles, Delhi, Taipei? How to get the right co-ordinates?

@kar95ar Yes, the areas of interest are the South Coast Air Basin (encompassing LA and other areas), Delhi, and Taipei. You can get the relevant coordinates by using the grid_metadata.csv.

Thank you for clarifying. Each entry has a different coordinates. How do I work around this issue?
I’m trying to understand overall datasets and how to do I connect them.

Also, in the pm25 metadata dataset provided by you, has a column named split. Do we consider test train split based on the split column provided by you?

Each entry has a different coordinates. How do I work around this issue?

I think there are a few ways, but one thing you can do to simplify things is to create a bounding box by getting the min and max latitudes and longitudes for each location, and then pass that to the intersects kwarg of the pystac client.

Alternatively, you can just pass all the polygons as a multipolygon to the intersects kwarg.

Example of how to define an area_of_interest to pass to intersects halfway down the page here: Planetary Computer

Example of how to define a multipolygon: https://dev.socrata.com/docs/datatypes/multipolygon.html#,

Also, in the pm25 metadata dataset provided by you, has a column named split. Do we consider test train split based on the split column provided by you?

These are the splits we’ve defined for scoring the competition, i.e. test corresponds to files that fall within the test period defined by the competition. For training, you are free to split the training data however you like.

@cszc Thank you so much for clarification. That really helps.

1 Like