Filtering down spectral bands

[jparal]

Hi, I am trying to download Harmonized Landsat Sentinel (HLS) dataset from earthdata using earthaccess. A single granule contains several spectral bands and mask(s) but for our use case, I only need a subset.
Is there a way for me to select a subset of the bands either using search_data API call or manually after getting the result of the search and before passing it onto download()?

Thx, J

[MattF-NSIDC]

What collection are you looking at?

I did a search and looks like there's a collection containing band-specific cloud-optimized geotiffs for this dataset (collection 10.5067/HLS/HLSS30.002) : https://data.lpdaac.earthdatacloud.nasa.gov/lp-prod-protected/HLSS30.020/HLS.S30.T01KBU.2023340T220911.v2.0/HLS.S30.T01KBU.2023340T220911.v2.0.B10.tif.

For this collection you could use the granule_name argument to search_data to select the bands you want before downloading.

[betolink]

Hi @jparal, if you know the name of the bands you want to filter you can definitely do it, here is an example after we get our results from search_data()

filtered_urls = []

for granule in results:
    links = granule.data_links()
    for link in links:
        if any(pattern in link for pattern in ["B02", "B01"]):
            filtered_urls.append(link)
            
earthaccess.download(filtered_urls, "HLS_DATA")

This notebook from LP DAAC does something similar, https://nasa-openscapes.github.io/earthdata-cloud-cookbook/tutorials/Observing_Seasonal_Ag_Changes.html#finding-hls-data-using-earthaccess Let us know if this helps!

[jparal]

Thank you, this is helpful. I could do it manually as you suggested and wrap the results in some regex filter. Thx.