satellite-remote-sensing

Satellite Remote Sensing

A skill for processing and analyzing satellite imagery for earth science research. Covers data acquisition from major satellite platforms, preprocessing workflows, spectral index computation, land cover classification, and change detection using Python geospatial tools.

Satellite Data Sources

Major Earth Observation Missions

Mission	Operator	Resolution	Revisit	Key Bands	Access
Landsat 8/9	USGS/NASA	30m (MS), 15m (pan)	16 days	11 bands, OLI+TIRS	Free (USGS EarthExplorer)
Sentinel-2	ESA	10m-60m	5 days	13 bands, MSI	Free (Copernicus Open Access Hub)
MODIS	NASA	250m-1km	1-2 days	36 bands	Free (NASA LAADS DAAC)
Sentinel-1	ESA	5-20m	6 days	C-band SAR	Free (Copernicus)
GOES-16/17	NOAA	0.5-2km	5-15 min	16 bands, ABI	Free (NOAA CLASS)

Programmatic Data Access

import planetary_computer
import pystac_client
import rioxarray

# Search Sentinel-2 imagery via Microsoft Planetary Computer
catalog = pystac_client.Client.open(
    "https://planetarycomputer.microsoft.com/api/stac/v1",
    modifier=planetary_computer.sign_inplace,
)

# Search for cloud-free imagery over a region
search = catalog.search(
    collections=["sentinel-2-l2a"],
    bbox=[11.0, 46.0, 12.0, 47.0],  # Tyrol, Austria
    datetime="2025-06-01/2025-08-31",
    query={"eo:cloud_cover": {"lt": 10}},
)

items = search.item_collection()
print(f"Found {len(items)} scenes with <10% cloud cover")

# Load a specific band as xarray DataArray
item = items[0]
red = rioxarray.open_rasterio(item.assets["B04"].href)
nir = rioxarray.open_rasterio(item.assets["B08"].href)

Preprocessing Pipeline

Atmospheric Correction

Raw satellite data (Level-1) must be atmospherically corrected to obtain surface reflectance (Level-2):

• Sentinel-2: Use Sen2Cor processor (ESA) or download pre-processed L2A products
• Landsat: Collection 2 Level-2 products include surface reflectance
• Custom correction: Use 6S radiative transfer model via Py6S