Abstract
Sentinel-3A (S3A), carrying the Ocean and Land Colour Instrument (OLCI), was successfully launched on February 16th 2016. It was the first of the series planned by the European Commission (EC) in the frame of COPERNICUS Sentinel program. Sentinel-3B is planned for launch in late 2017, bearing identical instruments, thus improving the global Earth coverage. The OLCI series providing global coverage at 300m resolution will therefore represent a major breakthrough in the family of ocean colour sensors.
For being an operational mission feeding in downstream Copernicus services like CMEMS, it is essential to ensure product quality prior public release. As for most ocean color missions, this supposes the implementation of a system vicarious calibration (SVC). Based on the methodologies applied to MERIS (and historically to older sensors), SVC of OLCI is performed separately for near-infrared (NIR) and visible (VIS) bands.
NIR bands SVC is performed over dedicated oligotrophic targets. Over such waters the marine signal is negligible in this spectral region compared to the contribution of the atmosphere. This avoids the use of in situ colocated data and provides enough sensor observations for the analysis and the computation of vicarious calibration gains. Different methodologies are tested and their intercomparison provides the conclusion that a simple unweighted regression on the NIR aerosol reflectances provides robust consistency in the obtained gains.
On the other hand, VIS bands SVC relies on OLCI observations matching very high quality in situ measurements after calibration of the NIR. For the time being, only two operational stations provide sufficiently high quality data for this purpose: BOUSSOLE in the Mediterranean Sea and MOBY in the Pacific Ocean. The number of accurate vicarious calibration matchups to ensure statistically reliable gains is discussed. To increase the statistics and improve reliability, an alternative procedure based on the use of global daily climatologies is shown to provide consistent additional measurements for the computation of robust SVC gains in the VIS.
In this presentation, the implemented SVC procedures for OLCI are described along with the analysis ensuring their reliability. OLCI product quality improvement brought by SVC is shown through the analysis of individual user products as well as by comparison with in situ data and other sensors.
System Vicarious Calibration of Sentinel-3 OLCI
Sentinel-3A (S3A), carrying the Ocean and Land Colour Instrument (OLCI), was successfully launched on February 16th 2016. It was the first of the series planned by the European Commission (EC) in the frame of COPERNICUS Sentinel program. Sentinel-3B is planned for launch in late 2017, bearing identical instruments, thus improving the global Earth coverage. The OLCI series providing global coverage at 300m resolution will therefore represent a major breakthrough in the family of ocean colour sensors.
For being an operational mission feeding in downstream Copernicus services like CMEMS, it is essential to ensure product quality prior public release. As for most ocean color missions, this supposes the implementation of a system vicarious calibration (SVC). Based on the methodologies applied to MERIS (and historically to older sensors), SVC of OLCI is performed separately for near-infrared (NIR) and visible (VIS) bands.
NIR bands SVC is performed over dedicated oligotrophic targets. Over such waters the marine signal is negligible in this spectral region compared to the contribution of the atmosphere. This avoids the use of in situ colocated data and provides enough sensor observations for the analysis and the computation of vicarious calibration gains. Different methodologies are tested and their intercomparison provides the conclusion that a simple unweighted regression on the NIR aerosol reflectances provides robust consistency in the obtained gains.
On the other hand, VIS bands SVC relies on OLCI observations matching very high quality in situ measurements after calibration of the NIR. For the time being, only two operational stations provide sufficiently high quality data for this purpose: BOUSSOLE in the Mediterranean Sea and MOBY in the Pacific Ocean. The number of accurate vicarious calibration matchups to ensure statistically reliable gains is discussed. To increase the statistics and improve reliability, an alternative procedure based on the use of global daily climatologies is shown to provide consistent additional measurements for the computation of robust SVC gains in the VIS.
In this presentation, the implemented SVC procedures for OLCI are described along with the analysis ensuring their reliability. OLCI product quality improvement brought by SVC is shown through the analysis of individual user products as well as by comparison with in situ data and other sensors.