Level-1 Products

Validation of Level-1 products includes both radiometric and geometric validation activities. These activities occur once the radiometric and the geometric calibration of the MSI instrument and the data processing algorithms have been performed.

The radiometric calibration and image quality monitoring approach rely heavily on the on-board Sun diffuser. As MSI has no secondary or back-up diffuser, it is therefore mandatory to obtain an independent validation of the radiometry in order to:

  • Identify any radiometric bias introduced by uncertainties in the calibration method; this concerns absolute radiometry as well as relative radiometry (in the field, inter-band, and inter-satellite).
  • Monitor the long-term radiometry to prevent any impact of the ageing of the diffuser.
  • Serve as a back-up calibration source in case of failure of the diffuser.

A large number of independent and complementary methods are combined to ensure a robust validation approach:

  • The Pseudo-Invariant Calibration Site (PICS) method: based on a long history of MERIS measurements and has been successfully used for a number of sensors, including MSI-A and B, and OLCI-A and B. (see Bouvet et al, 2014)
  • The Rayleigh method over oceanic sites allows an independent validation for lower radiance levels. It is however restricted to shorter wavelengths.
  • The DCC method can be used for relative radiometric assessments. Because DCC are high-altitude targets, they are less dependent on an estimation of the radiative transfer through the atmosphere. The large number of measurements available is another advantage of the method.
  • In-situ measurements over the Railroad Valley vicarious calibration site (U. of Arizona) are provided regularly to ESA through a collaboration agreement with NASA. The OPT-MPC continue to monitor these results.
  • Inter-sensor comparisons offer an important complement and directly address interoperability issues.

Moreover, some activity includes image quality monitoring and validation activities:

  • Signal to Noise Ratio (SNR) and Fixed Pattern Noise (FPN) short-term performance monitoring using diffuser acquisitions
  • SNR and FPM independent validation using ground images for long-term performance assessment
  • FTM validation.

The geometric validation activities consist in:

  • Absolute geolocation performance validation: evaluate absolute geolocation performance against a set of Ground Control Points using L1C images.
  • Relative Geolocation performance validation: evaluate geolocation performance relative to GRI and trigger a geometric calibration if necessary
  • Multi-temporal registration uncertainty validation: evaluation of L1C multi-temporal performance over random and selected sites using image correlation.
  • Multi-spectral registration uncertainty validation: evaluation of the multi-spectral registration uncertainty for refined and unrefined products.

The status of the geometric and radiometric validation of the Level-1 is reported in the Monthly Level-1 Data Quality Report.

Level-2A Products

Concerning Level-2A products, the outputs of the L2A processor (i.e. Sen2Cor) can be classified in two main domains:

  • Radiometry (RAD), which concerns the Atmospheric Correction outputs of Sen2Cor: Surface reflectance images, Aerosol Optical Thickness (AOT) and Water Vapour (WV) maps.
  • Cloud Screening and Classification (CSC) outputs of Sen2Cor: Scene classification (SCL) map which assigns to each pixel a class (vegetation, soil, water, clouds, thin cirrus, snow, etc…), Cloud probabilistic mask (CLDPRB) and Snow probabilistic (SNWPRB) mask.

Moreover, a proper characterization and update (if necessary) of the auxiliary data used in the L2A processing is also performed. These auxiliary data are: Digital Elevation Model, Meteorological data (CAMS & ECMWF), ESA CCI Land Cover layers, MSI Spectral responses, Solar Spectrum, Sentinel- 2 viewing angles and solar illumination angles.

The validation activities of the Level-2A products consist in:

  • Validation of the Aerosol Optical Thickness (AOT) (550 nm) and Water Vapour products processed with Sen2Cor with AERONET measurements as reference,
  • Quantitative validation of Sen2cor surface reflectance outputs with reference measurements,
  • Validation of the surface reflectance product using intercomparisons and computed references,
  • Cloud Screening and Classification performance assessment.

The status of the geometric and radiometric validation of the Level-2 is reported in the Monthly Level-2 Data Quality Report.


Menu Display