Multi-platform Chlorophyll and Sea Surface Temperature data fusion for Potential Fishing Zone detection using MODIS/VIIRS and ancillary data
Abstract:Satellite remote sensing data are being successfully used in several countries to localize potential fishing zones (PFZ). Oceanographic conditions related to water temperature and nutrient distribution strongly influence natural concentration of fish stocks and can be identified as oceanic upwelling fronts using information about Sea Surface Temperature (SST) and Chlorophyll Concentration (CHL) obtained from satellite data.
In this context, the present work illustrates a satellite-based near-real time fishery forecasting system tuned for the Adriatic Sea in order to help fishermen in saving fuel and ship time. More specifically, we developed an innovative multi-mission, multi-sensor and multi-temporal data fusion technique for the automatic identification of the upwelling fronts from the geometry of the distribution of the SST and CHL and of their gradients, using observations from the ocean color MODIS sensors of Terra and Aqua NASA satellites in conjunction with appropriate data from the VIIRS sensor on the Suomi/NPP satellite.
To this aim, precision geo-referencing and masking procedures were used to allow automatic selection of good quality pixels. A time stack of the final maps has been obtained as three-day running average with one-day step, taking into account the average local duration of the upwelling fronts. Finally, a gradient-based edge detection algorithm using the Canny filter has been applied to each of the three-day composite maps, based on the distribution of the SST and CHL and of their gradients, for the determination of the fronts as potential fishing zones.
The procedure is being validated using the geographic information of categorized fish Catch Per Unit Effort (CPUE) of the 2013 fishing campaigns of the FEDERPESCA fleet, as part of the project "Sustainable Fishery", funded by Apulia Region, in Southern Italy (FEP 2007-2013 – Measure 3.1 ), for the development of a satellite-based near-real time fishery forecasting system to be used in the Adriatic Sea.
The results of this study will be presented and commented. Moreover, our results show that data fusion applied to information obtained from different platforms and instruments, although with not overlapping wavebands and different spectral sensitivity, can give important contribution.