Sentinel-1 SAR Backscatter

Date modified: 21 April 2021

This dataset is expected to become available to Digital Earth Africa users in the second quarter of 2021.

Product overview

Background

The Sentinel-1 mission, composed of a constellation of two C-band Synthetic Aperture Radar (SAR) satellites, are operated by European Space Agency (ESA) as part of the Copernicus Programme. The mission currently collects data every 12 days over Africa at a spatial resolution of approximately 20 m.

Radar backscatter measures the amount of microwave radiation reflected back to the sensor from the ground surface. This measurement is sensitive to surface roughness, moisture content and viewing geometry. DE Africa provides Sentinel-1 backscatter as Radiometrically Terrain Corrected (RTC) gamma-0 (γ0) where variation due to changing observation geometries has been mitigated.

The dual polarisation backcastter time series can be used in applications for forests, agriculture, wetlands and land cover classification. SAR’s ability to ‘see through’ clouds makes it critical for mapping and monitoring land cover changes in the wet tropics.

DE Africa’s Sentinel-1 backscatter product is developed to be compliant with the CEOS Analysis Ready Data for Land (CARD4L) specifications.

Specifications

Coverage and metadata for the Sentinel-1 backscatter can be viewed on DE Africa’s Metadata Explorer.

Table 1: Sentinel-1 product specifications

Specification

Product name

Sentinel-1 Radiometric Terrain Corrected SAR Backscatter

Instrument

C-SAR

Frequency

C-band (5.405 GHz)

Acquisition mode

Interferometric Wide swath (IW)

Polarisation

VV + VH

Number of bands

5

Cell size - X (degree)

0.0002

Cell size - Y (degree)

0.0002

Coordinate reference system

EPSG: 4326

Temporal resolution

Every 12 days

Temporal range

2017-01-01 – present

Parent dataset

Sentinel-1 GRD

Update frequency

Daily

Table 2: Sentinel-1 measurements

Band ID

Description

Data type

No data value

vv

Linear backscatter intensity in VV polarization

float32

NaN

vh

Linear backscatter intensity in VH polarization

float32

NaN

mask

Data mask

uint8

0

area

Normalized scattering area

float32

NaN

angle

Local incidence angle (degree)

uint8

255

Table 3: Sentinel-1 data mask definition

Value

Description

0

Nodata

1

Valid

2

Invalid (in/near radar shadow)

Processing

DE Africa Sentinel-1 dataset is processed by Sinergise Sentinel Hub using ESA Sentinel-1 GRD as input. Radiometric terrain correction is applied using Copernicus DEM, following method developed in D. Small 2011.

Media and example images

False color backscatter over Okavango

Figure 1: Sentinel-1 false color composite over an area in Okavango Delta, Botswana.

References

  1. Small, “Flattening Gamma: Radiometric Terrain Correction for SAR Imagery,” in IEEE Transactions on Geoscience and Remote Sensing, vol. 49, no. 8, pp. 3081-3093, Aug. 2011, doi: 10.1109/TGRS.2011.2120616.

License

Access to Sentinel data is free, full and open for the broad Regional, National, European and International user community. View Terms and Conditions.

Data access

Amazon Web Services S3

DE Africa Sentinel-1 dataset is available in AWS S3, sponsored by the Public Dataset Program.

Table 4: AWS data access details.

AWS S3 details

Bucket ARN

arn:aws:s3:::deafrica-sentinel-1

Product name

s1_rtc

The bucket is in the AWS region af-south-1 (Cape Town). Additional region specifications can be applied as follows:

aws s3 ls --region=af-south-1 s3://deafrica-sentinel-1/

The file paths follow the format <product_name>/<tile_id>/<year>/<month>/<day>/<datatake_id>/.

The filenames follow the format <product_name>_<datatake_id>_<tile_id>_<year>_<month>_<day>_<band>.<extension>.

Table 5: AWS file path and name convention.

File path element

Description

Example

productname

Product name

s1_rtc

tile_id

1x1 deg tile labeld by southwest corner coordinates

N00E041

year

Observation year

2021

month

Observation month

02

day

Observation day

02

datatake_id

Sentinel-1 mission datatake ID (000001-FFFFFF)

0445E6

band

band ID, VV, VH, MASK, ANGLE, AREA, metadata or userdata

VV

extension

.tif for image files, .json or .xml for metadata files

.tif

For each Sentinel-1 tile, two metadata files are provided:

  • _metadata.json is a STAC compliant file.

  • _metadata.xml is a CARD4L compliant file.

OGC Web Services (OWS)

This product is available through DE Africa’s OWS.

Table 6: OWS data access details.

OWS details

Name

DE Africa Services

Web Map Services (WMS) URL

https://ows.digitalearth.africa/wms?version=1.3.0

Web Coverage Service (WCS) URL

https://ows.digitalearth.africa/wcs?version=2.1.0

Layer name

s1_rtc

Digital Earth Africa OWS details can be found at https://ows.digitalearth.africa/.

For instructions on how to connect to OWS, see this tutorial.

Open Data Cube (ODC)

The Sentinel-1 dataset can be accessed through the DE Africa ODC API, which is available through the DE Africa Sandbox.

ODC product name: s1_rtc

Specific bands of data can be called by using either the default names or any of a band’s alternative names, as listed in the table below. ODC Datacube.load commands without specified bands will load all bands.

Table 7: Sentinel-1 ODC band names.

Band name

Alternative names

vv

VV

vh

VH

mask

MASK

area

AREA, normalised_scattering_area

angle

ANGLE, local_incidence_angle

Band names are case-sensitive.

For examples on how to use the ODC API, see the DE Africa example notebook repository.

Technical information

Radiometric terrain correction

Radar backscatter can be provided as Beta Naught (β0), Sigma Naught (σ0) or Gamma Naught (γ0), depending on the reference area used to normalise the backscatter as shown in Figure 2.

  • β0 is normalised to an area defined in slant range plane

  • σ0 is normalised to ground area

  • γ0 is normalised to an area perpendicular to the line of sight

Normalization areas for SAR backscatter

Figure 2: Normalization areas for SAR backscatter. Source: Small 2011.

When local terrain information is not available, an ellipsoidal Earth model can be used to define the reference areas. However, this may lead to significant inaccuracy in the normalised backscatter.

Using knowledge of sensor geometry and a digital elevation model, local scattering (or illuminated) area can be esimated to achieve a radiometric terrain correction and minimise the impact of terrain on the measured backscatter. Such correction allows data collected with different viewing geometry to be combined or compared to detect real change on Earth surface.

Backscatter with and without RTC

Figure 3: Beta0 (left) and RTC Gamma0 (right) over an area near Lake Edward, Democratic Republic of the Congo. Backscatter variation due to terrain is minimized in RTC and contrast between different land cover types is increased.

Geometric accuracy

Absolute geometric accuracy of this product has been estimated to be within 6 metres (RMSE), using data processed over Corner Reflectors in Queensland Australia.