Minimize Level-1

SLSTR Level-1 processing is divided into two processing modes corresponding to two specific branches:

  • Level-1 calibration processing (S1-L1CAL), corresponding to the calibration processing mode
  • Level-1A and Level-1B processing (S1-L1a and S1-L1b), corresponding to the observation processing mode.

Level-1 calibration processing aims to produce the VISCAL ADF of an orbit N. This file is then used to calibrate numerical counts from solar channel measured during orbit N+1.

The numerical counts measured by the VISCAL target, during the full illumination window (i.e. near the South Pole) are extracted and verified. Several parameters such as gain and offset for the calibration period are computed and gathered in one file.

The observation processing mode is divided into three main processes.

The first process, Level-1A processing, aims to compute ortho-geolocation of each instrument pixel and to compute radiometric calibration coefficients. It contains five main steps:

  1. Source Packet Processing: unpacking, validating and converting source packet data and ancillary parameters.
  2. IR Channel Calibration: calculating the calibration offset and slope that describes the linear relationship between numerical counts and radiance (focusing on thermal and fire channels).
  3. Visible/NIR/SWIR Channel Calibration: similar to Level-1 calibration processing.
  4. Time Calibration: computing the time stamp of each pixel.
  5. Geolocation: computing the longitude/latitude/altitude and the corresponding (x,y) coordinates of each instrument pixel.

The second process, Level-1B processing, aims to calibrate and re-grid the SLSTR measurements, compute the TDI averaged pixel for SWIR channels and test each pixel for cloud presence. It contains nine main steps:

  1. Signal Calibration for thermal and solar channels using the calibration coefficients derived from Level-1A processing. 1. An empirical Stray light correction model to the SLSTR LWIR brightness temperature is expected to be implemented in coming evolutions.
  2. Time Domain Averaging: computing an average of A and B stripes to generate a reduced-noise grid.
  3. Re-gridding: determining, for each instrument pixel, its indices in across and along track direction taking into account a 1 km grid for thermal IR channels or a 0.5 km grid for visible, NIR and SWIR channels.
  4. Cosmetic Filling: filling missing image pixels as well as some parameters needed for further sections.
  5. Surface Classification and Cloud flagging are applied to each image pixel. In the future it will be improved to ensure a well-behaviour in the FRP and AEROSOL processing module.
  6. Reflectance to Radiance Conversion.
  7. Tie point interpolation: delimitate rectangular SLSTR L1b image and increase the tie point resolution in along-track direction (1 km instead of 16 km)
  8. Meteorological Annotations Computation providing ECMWF variables on a sub-grid of tie-points.
  9. Product Formatting: aiming to identify the variables included in SLSTR product files and metadata.