Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Geocoded Polarimetric Covariance (GCOV)

Product Specification Find Data

Product Overview

Geocoded Polarimetric Covariance (GCOV) products provide calibrated backscatter measurements corrected for both radiometric and terrain distortions, presented as gamma-0 power values. For users familiar with other corrected SAR amplitude products, these data can be used like Normalized Radar Backscatter (NRB) or Radiometrically Terrain Corrected (RTC) products.

Due to the side-looking nature of NISAR, topography can significantly affect the magnitude of backscatter, with areas facing the sensor becoming artificially “brighter” (returning higher backscatter) than expected, which biases covariance measurements. To mitigate the effect of topography, an area-based RTC is applied to the covariance terms, normalizing the backscatter. The reference digital elevation model (DEM) for processing and RTC is based on the Copernicus 30-m and 90-m DEM. These normalized terms are then geocoded to the output grid using area-based adaptive multi-looking.

GCOV products are projected to the appropriate UTM zone for their location and have a pixel spacing of 10 or 20 meters, depending on the acquisition location. They can be used directly in geospatial analysis platforms and workflows. Each product includes a covariance variable for each polarization available from the acquisition.

Product Specification

A complete description of NISAR GCOV products is available in Shiroma (2024).

Data Layers

A GCOV data product includes the following raster data sets. Complete descriptions of these data layers are available in Shiroma (2024, Section 4.3).

Geocoded Polarimetric Covariance Terms

/science/LSAR/GCOV/grids/frequency[A|B]/HHHH
/science/LSAR/GCOV/grids/frequency[A|B]/HVHV
/science/LSAR/GCOV/grids/frequency[A|B]/HHHV
/science/LSAR/GCOV/grids/frequency[A|B]/HHVH
/science/LSAR/GCOV/grids/frequency[A|B]/HHVV
/science/LSAR/GCOV/grids/frequency[A|B]/HVVH
/science/LSAR/GCOV/grids/frequency[A|B]/HVVV
/science/LSAR/GCOV/grids/frequency[A|B]/VVVV
/science/LSAR/GCOV/grids/frequency[A|B]/VHVH
/science/LSAR/GCOV/grids/frequency[A|B]/VHVV
/science/LSAR/GCOV/grids/frequency[A|B]/RHRH
/science/LSAR/GCOV/grids/frequency[A|B]/RVRV
/science/LSAR/GCOV/grids/frequency[A|B]/RHRV

The primary data elements of the GCOV product are the images of the geocoded polarimetric covariance terms. Visit Polarimetry (2025) to learn more about NISAR polarimetry.

The diagonal terms shown in Figure 3 (HHHH, HVHV, VVVV, VHVH, RHRH, RVRV) are real-valued and describe the intensity of the radar backscatter for the given polarization.

The remaining terms shown in Figure 3 (HHHV, HHVH, HHVV, HVVH, HVVV, VHVV, RHRV) are complex-valued and describe the intensity and phase between different polarizations. These off-diagonal terms are only included for quad-pol data acquisitions. Users who require off-diagonal values for dual-pol acquisitions will need to generate these layers manually from GSLC products.

The frequencies and polarizations available in a particular GCOV data product will vary based on the acquisition mode used to collect the data.

Diagram of the GCOV polarizations

Figure 3:Matrix of the possible polarizations available in NISAR GCOV products. Products provided for any of the combinations on the diagonal (highlighted in dark grey) are formatted with real-valued float values, while any of the products provided from above the diagonal (highlighted in light grey) are formatted with complex values.

Number of Looks

/science/LSAR/GCOV/grids/frequency[A|B]/numberOfLooks

The GCOV terms are obtained from the geocoding of the RSLC product polarimetric covariance terms using an adaptive area-based multi-looking algorithm. The multi-looking window and number of averaged looks vary with the topography and radar geometry.

The number of looks layer provides the number of looks used for computing each GCOV term sample.

Mask

/science/LSAR/GCOV/grids/frequency[A|B]/mask

The mask layer provides information about the averaging ensemble of radar samples that originated the GCOV terms through adaptive multi-looking.

Radiometric Terrain Correction Gamma-to-Sigma Factor

/science/LSAR/GCOV/grids/frequency[A|B]/rtcGammaToSigmaFactor

The RTC Gamma-to-Sigma factor provides factors to normalize the backscatter normalization convention of the GCOV matrix from gamma0 to sigma0.

Naming Convention

NISAR GCOV product names will conform to the following naming convention:

NISAR_IL_PT_PROD_CYL_REL_P_FRM_MODE_POLE_S_StartDateTime_EndDateTime_CRID_A_C_LOC_CTR.EXT

Hover over a term listed above to see its description, or click on a specific segment to skip to its description below. Refer to Figure 5 in the Naming Conventions section to see a diagram.

NISAR
5 characters for the Mission

  • NISAR

I
1 character for the Instrument used for the acquisition

  • L: L-band SAR

  • S: S-band SAR

L
1 character for the product processing Level (1/2/3)

  • GCOV products are Level 2

PT
2 characters for the Processing Type

  • PR: Production

  • UR: Urgent Response

  • OD: Science On-Demand

PROD
4 characters for the PRODuct Identifier

  • GCOV

CYL
3 characters for the CYcLe number in the mission

  • each cycle represents 12 days

  • zero padded

  • starting at 001

REL
3 characters for the RELative orbit track number within a cycle

  • resets to 1 with a cycle number increment

  • zero padded

  • valid values: 001-173

P
1 character for the Orbit Direction (direction of movement of the satellite at the time of imaging)

  • A: Ascending

  • D: Descending

FRM
3 characters for the track FRaMe number

  • a segment of an orbital track corresponding to the product

  • zero padded

  • valid values: 001-176 on each track

MODE
4 characters for the Bandwidth MODE Code of the Primary and Secondary Bands

  • first two characters for primary band, last two for secondary band

  • Possible values: 40, 20, 77, 05, or 00

    • 00 only used if the secondary band is missing

POLE
4 characters for the Polarization of the data for the primary and secondary bands respectively. Each band uses a two character code among the following:

  • SH = HH – Single Polarity (H transmit and receive)

  • SV = VV – Single Polarity (V transmit and receive)

  • DH = HH/HV – Dual Polarity (H transmit)

  • DV = VV/VH – Dual Polarity (V transmit)

  • CL= LH/LV – Compact Polarity (Left transmit)

  • CR = RH/RV – Compact Polarity (Right transmit)

  • QP = HH/HV/VV/VH – Quad Polarity

  • NA if band does not exist

For example:

  • a single-pol H-primary polarization mode would be SHSH

  • a dual-pol V-primary polarization mode would be DVDV

  • a “quasi-quad” polarization mode would be noted DHDV

  • a “quasi-dual” polarization mode would be noted SHSV

S
1 character for the Source of data for the product

  • A: Acquired source of the observation, single mode

  • M: Mixed source of observations, mixed mode

StartDateTime
15 chars for Radar Start Date/Time of the data processed as zero Doppler

  • Format: YYYYMMDDTHHMMSS, UTC

EndDateTime
15 chars for Radar End Date/Time of the data processed as zero Doppler

  • Format: YYYYMMDDTHHMMSS, UTC

CRID
6 characters for the Composite Release IDentifier

  • indicates the version of the workflow used to process the data, incorporating a number of software and project versions

  • Format of EPMMmm

  • MMmm indicates the major and minor version numbers

A
1 character for the Product Accuracy or Fidelity of the Orbit Ephemeris and Radar Pointing:

  • P: Precise

  • M: Medium

  • N: Near Real-Time

  • F: Forecast

C
1 character indicating Coverage of the frame

  • F: Full

  • P: Partial

LOC
1 character to represent the Location of the Science Data System

  • J: JPL

CTR
3 characters for product CounTeR

  • zero padded

EXT
1 to n characters for Extension

  • h5

  • met

  • log

Finding Data

GCOV products can be found in Vertex by setting the Dataset filter to “NISAR” and the Science Product filter to “GCOV (Geocoded Polarimetric Covariance)”: Find Data in Vertex

GCOV products can be found in Earthdata Search by searching for “NISAR_L2_GCOV_”: Find Data in Earthdata Search

References
  1. Shiroma, G. H. X. (2024). Level-2 Geocoded Polarimetric Covariance (NASA SDS Product Specification JPL D-102274 (Revision E)). NASA Jet Propulsion Laboratory. https://nisar.asf.earthdatacloud.nasa.gov/NISAR-SAMPLE-DATA/DOCS/NISAR_D-102274_RevE_NASA_SDS_Product_Specification_L2_GCOV_Nov8_2024_w-sigs.pdf
  2. Polarimetry. (2025). https://science.nasa.gov/mission/nisar/polarimetry/
Geocoded Products (Level 2)
GUNW
Range-Doppler Products (Level 1)
RSLC