Downloading landsat data and uploading shapefile

[vikhyatgupta19]

Hi All,

I am very new to MPC and currently playing around with it. I was following the documentation of Reading Data from the STACK API : https://planetarycomputer.microsoft.com/docs/quickstarts/reading-stac/. I would like to know if it's possible to download a particular data to my local drive.

Like in my case I am trying to download a single Landsat image by using urllib library however, it's not working and am getting error: Item object has no attribute href.

Here is the complete code :

``
import pystac_client
import planetary_computer

catalog = pystac_client.Client.open("https://planetarycomputer.microsoft.com/api/stac/v1",
modifier=planetary_computer.sign_inplace,)

time_range = "2023-04-01/2023-04-30"
bbox = [76.640625, 28.844674, 81.188965, 31.774878]

search = catalog.search(collections=["landsat-c2-l2"], bbox=bbox, datetime = time_range, query={"platform":{"in": ["landsat-8"]}})
items = search.get_all_items()
len(items)

import geopandas
df = geopandas.GeoDataFrame.from_features(items.to_dict(), crs="epsg:4326")
df

selected_item = min(items, key=lambda item: item.properties["eo:cloud_cover"])
print(selected_item)

from IPython.display import Image
Image(url=selected_item.assets["rendered_preview"].href, width=500)

import urllib.request
filename, _ = urllib.request.urlretrieve(selected_item.href)
``
I would also like to know if there is a way to upload custom data like admin boundaries shapefile. I was hoping to use a shapefile and clip the Landsat imagery according to it.

Additionally, are there any particular resources that one can use to get started with MPC.

[TomAugspurger]

Thanks for the reproducible example!

The first thing to note is that the URLs to download data from are in the assets of an item. Each item (in this case a Landsat scene) has multiple assets (in this case, different bands and some extra metadata files). So you would want something like

filename, _ = urllib.request.urlretrieve(selected_item.assets["red"].href)

to download a single file.

@gadomski put together a little library for downloading assets of a STAC item at https://github.com/gadomski/stac-asset. (Note that our landsat STAC items currently have an issue where we have an asset for a JSON file that doesn't actually exist in blob storage).

I would also like to know if there is a way to upload custom data like admin boundaries shapefile.

Not currently.

I was hoping to use a shapefile and clip the Landsat imagery according to it.

I've used rioxarray.clip for this before. It works well.

Additionally, are there any particular resources that one can use to get started with MPC.

Our quickstarts and tutorials are the best Planetary Computer resources. If you're looking for a general introduction to the scientific Python ecosystem, Project Pythia is a great resource.

One quick note: we generally recommend computing on the data in-place and not downloading the data locally. It's fine to download small amounts of data, especially for prototyping. But if you want to run a large workload processing a lot of data, moving data out of Azure will quickly become a bottleneck. For these medium- or larger-scale workloads, you're better of putting your compute in Azure next to the data.