High resolution ocean fronts product from JPSS VIIRS for improved composite mapping

High-resolution composite maps of oceanic thermal fronts reveal their spatial distribution and variability, along with seasonal variability and climatology, which are used to study a variety of marine environmental phenomena. Customarily, fronts are detected in a clear sky domain of instantaneous Level 2 (swath) or 3 (gridded) SST images (e.g., Miller 2009, 2014). However, dynamic regions of the ocean (e.g., ocean currents, eddies and upwellings) are often misidentified as cloud by many current cloud masks, thus significantly limiting the front detection, exactly in the areas where it is needed the most.

Using a gap-free, high-resolution Level 4 SST analysis (such as the 1km global NASA MUR), in which various high-resolution satellite data have been assimilated and blended with in situ data, appears attractive. Application of SIED algorithm (Cayula 1995) to MUR indeed provides daily frontal product at a 1km resolution. However, high-resolution thermal structure is degraded in the L4 products. In particular, it is less reliable in the coastal zones, and may over-smoothed (especially, when over-screened L2/3 products are used as input).

The new NOAA operational instrument, VIIRS onboard S-NPP (launched in 2011) and two future satellites, J1 (2017) and J2 (2021), provides high quality and resolution SST imagery, superior to the current operational AVHRR and experimental MODIS sensors. The cloud mask employed in the NOAA Advanced Clear-Sky Processor for Oceans (ACSPO) SST system, is most liberal among the community SST products. Yet, false alarms do occasionally occur in ACSPO, especially in the dynamic oceanic regions. Performing front detection at the stage of cloud masking improves both fronts and mask. We plan to output the frontal product as an extra layer of the ACSPO SST product, which can be directly used in the composition process. We are very interested in discussing the new ACSPO fronts product, its utility and improvements, with our potential users.