OPERA RTC-S1 — analysis-ready backscatter workflow¶

The OPERA Radiometric Terrain Corrected Sentinel-1 product (opera-rtc-s1) is NASA's analysis-ready, terrain-corrected backscatter mirror of the Sentinel-1 mission. Unlike the raw SLC stack, RTC is search-only: it's already a processed output, not an InSAR input. This walkthrough searches the OPERA RTC catalog over a bbox and inspects the available scenes.

Needs the [asf] extra (pip install earthlens[asf]).

Why this product?¶

RTC is meant for amplitude / backscatter analyses — flood mapping, land-cover change, urban growth, forest disturbance — where the interferometric phase is not needed. The terrain correction makes the values directly comparable across acquisitions (and across satellites in the wider OPERA-S1 family) without per-pixel reprocessing.

Because the SLC has been collapsed into amplitude, stacking is not meaningful — the catalog flags this row as stackable: false, and the backend rejects stack mode with a clear error:

In [ ]:

from earthlens.asf import Catalog

row = Catalog().get_product('opera-rtc-s1')
print(f'platform={row.platform!r}  dataset={row.dataset!r}  product_type={row.product_type!r}')
print(f'stackable: {row.stackable}')

from earthlens.asf import Catalog row = Catalog().get_product('opera-rtc-s1') print(f'platform={row.platform!r} dataset={row.dataset!r} product_type={row.product_type!r}') print(f'stackable: {row.stackable}')

Search by bbox + window¶

OPERA RTC selects via DATASET.OPERA_S1 rather than a platform — the backend routes that through geo_search(dataset='OPERA_S1', processingLevel='RTC', ...) transparently.

In [ ]:

from earthlens.earthlens import EarthLens

el = EarthLens(
    data_source='asf',
    variables=['opera-rtc-s1'],          # alias 'opera-rtc' also works
    start='2024-06-01',
    end='2024-06-15',
    lat_lim=[37.0, 37.5],                # San Francisco Bay
    lon_lim=[-122.5, -122.0],
    path='opera_rtc_out',
    max_results=20,
)
products = el.datasource._search()
print(f'{len(products)} RTC product(s) match the request')

from earthlens.earthlens import EarthLens el = EarthLens( data_source='asf', variables=['opera-rtc-s1'], # alias 'opera-rtc' also works start='2024-06-01', end='2024-06-15', lat_lim=[37.0, 37.5], # San Francisco Bay lon_lim=[-122.5, -122.0], path='opera_rtc_out', max_results=20, ) products = el.datasource._search() print(f'{len(products)} RTC product(s) match the request')

Inspect what came back¶

Each RTC scene carries its acquisition time, the satellite (S1A / S1C), and pixel-spacing metadata. We pull a few fields for orientation.

In [ ]:

for product in products[:5]:
    props = product.metadata['product'].properties
    start = props.get('startTime', '?')
    platform = props.get('platform', '?')
    print(f'  {start}  {platform}  {product.metadata["fileName"]}')

for product in products[:5]: props = product.metadata['product'].properties start = props.get('startTime', '?') platform = props.get('platform', '?') print(f' {start} {platform} {product.metadata["fileName"]}')

Companion products¶

Each RTC scene has a static-layer mask companion (opera-rtc-s1-static) carrying topography masks, no-data zones, and similar per-burst metadata. The mask shares the same dataset key but a different product_type; pull it as a separate request when needed for masking analyses.

In [ ]:

static = EarthLens(
    data_source='asf',
    variables=['opera-rtc-s1-static'],
    start='2024-06-01',
    end='2024-06-15',
    lat_lim=[37.0, 37.5],
    lon_lim=[-122.5, -122.0],
    path='opera_rtc_static_out',
    max_results=5,
)
static_products = static.datasource._search()
print(f'{len(static_products)} static-layer mask product(s)')

static = EarthLens( data_source='asf', variables=['opera-rtc-s1-static'], start='2024-06-01', end='2024-06-15', lat_lim=[37.0, 37.5], lon_lim=[-122.5, -122.0], path='opera_rtc_static_out', max_results=5, ) static_products = static.datasource._search() print(f'{len(static_products)} static-layer mask product(s)')

Disturbance alerts: a different OPERA-S1 product¶

If you want change detection rather than raw backscatter, OPERA also publishes opera-dist-alert-s1 — a land-surface disturbance alert layer derived from the RTC time series. It's lower-latency and analysis-ready.

In [ ]:

dist = EarthLens(
    data_source='asf',
    variables=['opera-dist-alert-s1'],
    start='2024-06-01',
    end='2024-06-15',
    lat_lim=[37.0, 37.5],
    lon_lim=[-122.5, -122.0],
    path='opera_dist_out',
    max_results=5,
)
dist_products = dist.datasource._search()
print(f'{len(dist_products)} disturbance-alert product(s)')

dist = EarthLens( data_source='asf', variables=['opera-dist-alert-s1'], start='2024-06-01', end='2024-06-15', lat_lim=[37.0, 37.5], lon_lim=[-122.5, -122.0], path='opera_dist_out', max_results=5, ) dist_products = dist.datasource._search() print(f'{len(dist_products)} disturbance-alert product(s)')

Downloading¶

All three searches above ran anonymously. To pull the products to disk, set EDL credentials and call el.download() (see Authentication). The download path is shared with the InSAR stack: same ASFAuth, same idempotent skip-existing logic.