Using overlapping VIIRS scenes to observe short term variations in particulate matter in the coastal environment

Abstract

In coastal areas, the concentrations and the optical properties of the water components have a large spatial and temporal variability, due to river discharges and meteo-marine conditions, such as wind, wave and current, and their interaction with shallow water bathymetry. This large temporal variability cannot be captured using the standard Ocean Colour Radiometry (OCR) polar orbiting satellites, the latter providing almost one image per day. On the contrary, the use of OCR geostationary sensors, like the Geostationary Ocean Colour Imager (GOCI), centred above the Korean Peninsula, enable to capture the short-term variability of the optical properties. To compensate the lack of a geostationary sensor similar to GOCI over other coastal environments, like the North Adriatic Sea (NAS), the multiple observations provided during the same day by the Visible Infrared Imaging Radiometer Suite (VIIRS) mounted on the SUOMI NPP satellite, can be exploited. Indeed, due to its large swath of 3060 km, the VIIRS orbits can overlap over the NAS during the same day within 1 h and 42 min, an important feature that can be useful in capturing the short term variability of the optical properties.A large number of VIIRS overlaps in the NAS are characterized by high sensor zenith angle (SZA) of the observation, resulting in a large portion of images masked by the high satellite zenith flag. In order to make available those observations and, in general, to reduce the dependence of the VIIRS observations from the SZA, an adjustment based on a multi linear regression scheme, which exploits radiometric in situ observations, was here applied.This study aims to prove the suitability of the adjusted overlapping VIIRS in capturing the short time scale dynamics of particulate backscattering, and this was demonstrated by the analysis of a case study for the 21st and 22nd of March 2013. In order to evaluate the advantages in using multiple observations during the same day, also the ~24 h dynamics was analysed, comparing the overlapping VIIRS results with the ones obtained from the daily product.