OLCI Data Release

Download OLCI data

OLCI/Sentinel-3A Instrument

The Ocean and Land Color Instrument (OLCI) onboard the Sentinel-3A satellite is a medium-resolution and multi-spectral imager. The OLCI instrument is the follow-on mission to the Medium Resolution Imaging Spectrometer (MERIS) instrument flown on Envisat. The Sentinel-3A is on a near-polar, sun-synchronous orbit with a descending node equatorial crossing at ~10:00h Local Time (LT). The OLCI instrument adopts the “push-broom” method to scan the Earth’s surface covering the solar spectral range from 400 nm to 1040 nm, and it has a 68.5° field of view around nadir and covers a swath with of 1270 km (Sentinel-3 OLCI User Guides, https://sentinel.esa.int/web/sentinel/user-guides/sentinel-3-olci). The primary objective is to screen the ocean and land surface to obtain information related to biology, meanwhile OLCI observations are also can be used for obtaining valuable information for atmospheric aerosol and surface characterization.

Retrieval specifications

In the framework of EUMETSAT funded study to support Copernicus Programme, the GRASP (Generalized Retrieval of Atmosphere and Surface Properties) algorithm (Dubovik et al., 2011, 2014) is adapted on the development of OLCI/Sentinel-3A aerosol and surface products. The details of the project information are on https://www.eumetsat.int/science-studies/

Specifically, GRASP/Models approach (Chen et al., 2020; Lopatin et al., 2021; Dubovik et al., 2021) is used. The OLCI Reduced Resolution Level-1B (L1B RR) TOA radiances at 9 spectral channels (Oa02: 412.5 nm, Oa03: 442.5 nm; Oa04: 490 nm; Oa05: 510 nm; Oa06: 560 nm Oa08 665 nm; Oa12: 753 nm; Oa17: 865 nm; Oa21: 1020 nm) are inverted using the GRASP multi-pixel concept. In order to deal with OLCI single viewing measurements, which have no or limited sensitivity to angular features of BRDF, the OLCI/GRASP retrieval constrain the angular properties of underlying surface strongly: over ocean Fresnel reflection is fixed using a priori wind speed, and over land BRDF retrieval is constrained using a priori values for Ross-Li volumetric and geometric terms. The isotropic reflectance is retrieved both over land and ocean under mild spectral smoothness constraints.

OLCI/GRASP product description

The OLCI/GRASP V1.0 aerosol and surface products are processed based on GRASP (v1.0.0) for one year (June 2018 to May 2019) (We are going to extend our processing in the future).  The daily product is contained in files typically named as:

Level 2 (10 km):          GRASP_OLCI_L2_YYYYMMDD.full.nc

Level 3 (0.1 degree):   GRASP_OLCI_L3_YYYYMMDD.01degree.nc

Level 3 (1 degree):      GRASP_OLCI_L3_YYYYMMDD.1degree.nc

The file format is NetCDF-4. The Level 2 data is created from the raw GRASP output after merging them into global daily files and applying general filter for coastal line and pixels with high fitting residuals. The Level 2 file is at 10 km resolution under sinusoidal projection and Level 3 is created as regrided products at 0.1 and 1 degrees resolutions in equirectangular projection of data available in Level 2.

Levels 2 and 3 data are publicly available. Climatological monthly and seasonal aggregations are also provided along with the daily files.

The product post-processing procedure (i.e. quality flag) is under investigation and development.

In the files, the main aerosol parameter is AOD (). Meanwhile, we also diagnosed aerosol parameters such as: AAOD() and SSA(), which are of interest for further considerations. The main surface product is NDVI, white sky albedo DHR () and black sky albedo BHR_ISO (), and we keep diagnosed full BRDF parameters in the product file. We also provided supplementary information such as LandPercentage, relative fitting residual (ResidualRelative), absolute fitting residual (ResidualAbsolute) etc.

Recommendation for OLCI/GRASP data users

  • The main OLCI/GRASP aerosol product is AOD (). To simply select high quality product, the threshold ResidualRelative < 0.01 over land and ResidualRelative < 0.1 over ocean are recommended.
  • The Angstrom Exponent (AE) is generally of higher confidence over ocean than over land. The AE over bright land surface is of relatively low confidence due to the lack of information content.
  • Even though the diagnosed SSA is of interest, we suggest to filter low aerosol abundance, i.e. AOD 550<0.5, together with ResidualRelative.

We perform one-year preliminary validation of L3 product with AERONET and MAN maritime network. The grided product is averaged in a 5×5 window centered over the AERONET sites, meanwhile the available AERONET products are averaged within a time window (30 mins for direct Sun product, and 360 mins for MAN measurements):

OLCI/GRASP spectral AOD validation over land (AERONET):

OLCI/GRASP spectral AOD validation over ocean (AERONET):

OLCI/GRASP spectral AOD validation over ocean (MAN):

 

Spatial distribution OLCI/GRASP one-year (June 2018 – May 2019) global AOD 550 nm:

Please, contact us for further information.

Any publication, presentation, or other derivative work based on these results shall acknowledge its use.

Public files:

Level 2: Data after general filter at 10 km resolution under sinusoidal projection

  • Daily
  • Monthly
  • Seasonal

Level 3: Regrid level 2 at 0.1 and 1 degrees resolutions in equirectangular projection

  • 1 degree or 1 degree
    • Daily
    • Monthly
    • Seasonal
Category Parameters Description
Aerosol and Surface AOD () spectral aerosol optical depth
AAOD () spectral aerosol absorption optical depth
SSA () spectral single scattering albedo
AE Ångström exponent
Ross_Li_BRDF_iso () isotropic term of Ross-Li BRDF
Ross_Li_BRDF_vol volumetric term of Ross-Li BRDF
Ross_Li_BRDF_geo geometric term of Ross-Li BRDF
Cox_Munk_iso () isotropic water leaving radiance
DHR () directional hemispherical reflectance (black-sky albedo)
BHR_ISO () isotropic bihemispherical reflectance (white-sky albedo)
NDVI Normalized Difference Vegetation Index
Supplementary ResidualRelative relative fitting residual
ResidualAbsolute absolute fitting residual
Latitude
Longitude
MASL metres above sea level
LandPercentage Land percentage (%)
Pixel_count_surface  Number of aggregated pixels in datasets
Pixel_count_surface_land Number of aggregated pixels in surface-land datasets
Pixel_count_surface_ocean Number of aggregated pixels in surface-ocean datasets
x x-coordinate in projected space
xs x-coordinate in segmenting space
y y-coordinate in projected space
ys y-coordinate in segmenting space

λ = 412, 442, 490, 510, 560, 665, 753, 865 and 1020 nm

Download of GRASP products is available via the current Products page. You have two options: 
• Interactively via the website and the normal GRASP Single-Sign-On (click the ‘Download’ button above).
• Using HTTPS Basic Authentication for downloading data in a batch script (useful for big amount of files). Click ‘Download’ above, the wget command appears in the bottom of the webpage. When a new directory is accessed, the wget command will be automatically updated to download all the data files that fall under this directory.

For longer running downloads we advise to use a tool like screen or tmux. This would allow you to detach from the running process, leaving the download in the background.
Note: GRASP Cloud currently offers only a HTTPS Basic Authentication download possibility. FTP / FTPS / SFTP is not supported. Unfortunately tools like Filezilla will not work.
Popular alternatives to wget include curl, python requests and many more.

  • The current filter from raw output to Level 2 is generally by fitting residual over land and ocean. Next step, we need to use different procedure over different surface types (NDVIs) to make sure to screen bad pixels.
  • Due to the land surface climatology is over latitude [60S, 60N], we do not process land area with latitude higher than 60 degrees.
  • Some gaps over ocean are related with the loading of windspeed.
  • Other unknows …

Please, contact us if you identify anything else unusual.