Space Data Atlas

Challenges

Dancing with the SARs

Intermediate, Advanced. Earth Science.

Use radar data from the NASA-ISRO NISAR mission to track and visualize surface change, such as wetland loss, wildfire scars, earthquakes, farming or glacier motion, somewhere on Earth.

Read the official challenge on spaceappschallenge.org

NISAR products in CMR are labeled provisional or beta; Sentinel-1 OPERA RTC is the validated fallback.

Your first hour

  1. Create an Earthdata Login and run the NISAR GCOV starter for your area of interest.
  2. Preview coverage in the GIBS NISAR layer or ASF Search before downloading any large HDF5 files.
  3. Pick one change story (flooding, burn scar, crop cycle, ground motion) and gather before and after scenes.

Suggested datasets

In order of how useful they are likely to be. These are suggestions, not the official resource list. Tap a badge to see what it means.

No login
Open to anyone.
Free API key
Sign up for a free key first.
Searching is open; downloading needs a free Earthdata Login account.
Works in browser
Can be called straight from a web page.

NISAR L2 Geocoded Polarimetric Covariance (GCOV), Provisional

NASA ASF DAAC

Map-projected L-band radar backscatter from the NASA-ISRO NISAR mission, the simplest NISAR product for spotting surface change such as flooding, crop growth, forest loss or burn scars. Released as a provisional (pre-validation) product, so treat values with care; a beta version and urgent-response products also exist in CMR.

CMR short name NISAR_L2_GCOV_PROVISIONAL_V1, concept ID C2854338529-ASF, GIBS layer NISAR_L2_Geocoded_Polarimetric_Covariance

NISAR L2 Geocoded Unwrapped Interferogram (GUNW), Provisional

NASA ASF DAAC

Interferograms built from pairs of NISAR radar passes, showing centimeter-scale ground motion from earthquakes, subsidence, landslides or glacier flow. Provisional product; expect large HDF5 files and some InSAR background before using it.

CMR short name NISAR_L2_GUNW_PROVISIONAL_V1, concept ID C2854335566-ASF

OPERA Radiometric Terrain Corrected Backscatter from Sentinel-1 (RTC-S1)

NASA ASF DAAC / JPL OPERA

Analysis-ready C-band radar backscatter from Sentinel-1 as Cloud Optimized GeoTIFFs, validated and much easier to open than raw SAR. A good fallback or comparison layer for NISAR, with a longer archive.

CMR short name OPERA_L2_RTC-S1_V1, concept ID C2777436413-ASF, GIBS layer OPERA_L2_Radiometric_Terrain_Corrected_SAR_Sentinel-1

OPERA Dynamic Surface Water Extent from HLS (DSWx-HLS)

NASA PO.DAAC / JPL OPERA

30 m maps of open and partial surface water derived from Landsat and Sentinel-2, useful for tracking wetland loss, flooding and reservoir change, and for ground-truthing radar water maps.

CMR short name OPERA_L3_DSWX-HLS_V1, concept ID C2617126679-POCLOUD, GIBS layer OPERA_L3_Dynamic_Surface_Water_Extent-HLS

OPERA Land Surface Disturbance Alert from HLS (DIST-ALERT)

NASA LP DAAC / JPL OPERA

30 m alerts flagging where vegetation cover has dropped compared with a historical baseline, which picks up wildfire scars, clearing and storm damage. Pairs well with radar change maps for validation.

CMR short name OPERA_L3_DIST-ALERT-HLS_V1, concept ID C2746980408-LPCLOUD, GIBS layer OPERA_L3_DIST-ALERT-HLS_Color_Index

Harmonized Landsat Sentinel-2 (HLS) Landsat 30 m (HLSL30) v2.0

NASA LP DAAC

Landsat surface reflectance resampled to match Sentinel-2 on a common 30 m grid, so fields can be tracked every few days. The Sentinel-2 half is HLSS30; both come as Cloud Optimized GeoTIFFs.

CMR short name HLSL30, concept ID C2021957657-LPCLOUD, GIBS layer HLS_L30_Nadir_BRDF_Adjusted_Reflectance

Starter code

Search NISAR GCOV scenesPythonOpen in Colab
import earthaccess

# Search needs no login. Bounding box: (west, south, east, north)
results = earthaccess.search_data(
    short_name="NISAR_L2_GCOV_PROVISIONAL_V1",
    bounding_box=(-122.6, 37.0, -121.5, 38.2),  # San Francisco Bay
    temporal=("2025-10-01", "2026-10-04"),
    count=5,
)
print(len(results), "granules")
for g in results:
    print(g["umm"]["GranuleUR"])

# Next step (downloads need a free Earthdata Login):
# earthaccess.login()
# files = earthaccess.download(results[:1], "./nisar")
Find low-cloud HLS scenes over a farmPythonOpen in Colab
import earthaccess

# 30 m HLS Landsat scenes over a single farm area, low cloud only
results = earthaccess.search_data(
    short_name="HLSL30",
    bounding_box=(-93.7, 41.9, -93.5, 42.1),  # near Ames, Iowa
    temporal=("2026-05-01", "2026-09-30"),
    cloud_cover=(0, 20),
    count=5,
)
print(len(results), "scenes")
for g in results:
    print(g["umm"]["GranuleUR"])

# Each scene has one Cloud Optimized GeoTIFF per band.
# earthaccess.login()
# earthaccess.download(results[:1], "./hls")
Search NISAR interferogramsPythonOpen in Colab
import earthaccess

# Search needs no login. Bounding box: (west, south, east, north)
results = earthaccess.search_data(
    short_name="NISAR_L2_GUNW_PROVISIONAL_V1",
    bounding_box=(-122.6, 37.0, -121.5, 38.2),  # San Francisco Bay
    temporal=("2025-10-01", "2026-10-15"),
    count=5,
)
print(len(results), "granules")
for g in results:
    print(g["umm"]["GranuleUR"])

# Next step (downloads need a free Earthdata Login):
# earthaccess.login()
# files = earthaccess.download(results[:1], "./nisar")
Search OPERA radar backscatter tilesPythonOpen in Colab
import earthaccess

# Search needs no login. Bounding box: (west, south, east, north)
results = earthaccess.search_data(
    short_name="OPERA_L2_RTC-S1_V1",
    bounding_box=(-95.8, 29.5, -95.0, 30.1),  # Houston
    temporal=("2026-09-01", "2026-10-15"),
    count=5,
)
print(len(results), "granules")
for g in results:
    print(g["umm"]["GranuleUR"])

# Next step (downloads need a free Earthdata Login):
# earthaccess.login()
# files = earthaccess.download(results[:1], "./opera_rtc")
# import rasterio; img = rasterio.open(files[0]).read(1)
Search OPERA surface water mapsPythonOpen in Colab
import earthaccess

# Search needs no login. Bounding box: (west, south, east, north)
results = earthaccess.search_data(
    short_name="OPERA_L3_DSWX-HLS_V1",
    bounding_box=(-81.2, 26.6, -80.6, 27.2),  # Lake Okeechobee, Florida
    temporal=("2026-09-01", "2026-10-15"),
    count=5,
)
print(len(results), "granules")
for g in results:
    print(g["umm"]["GranuleUR"])

# Next step (downloads need a free Earthdata Login):
# earthaccess.login()
# files = earthaccess.download(results[:1], "./opera_dswx")
# import rasterio; img = rasterio.open(files[0]).read(1)
Search OPERA land disturbance alertsPythonOpen in Colab
import earthaccess

# Search needs no login. Bounding box: (west, south, east, north)
results = earthaccess.search_data(
    short_name="OPERA_L3_DIST-ALERT-HLS_V1",
    bounding_box=(-63.5, -10.5, -62.5, -9.5),  # Rondonia, Brazil
    temporal=("2026-08-01", "2026-10-15"),
    count=5,
)
print(len(results), "granules")
for g in results:
    print(g["umm"]["GranuleUR"])

# Next step (downloads need a free Earthdata Login):
# earthaccess.login()
# files = earthaccess.download(results[:1], "./opera_dist")
# import rasterio; img = rasterio.open(files[0]).read(1)

Try a live search

Ask NASA's Common Metadata Repository for the five newest files in NISAR L2 Geocoded Polarimetric Covariance (GCOV), Provisional. This runs in your browser and needs no login.