Assessing the accuracy of Visible Infrared Imaging Radiometer (VIIRS) and MultiUltrahigh Resolution (MUR) Sea-Surface Temperature products for inclusion of a model to forecast blooms of Alexandrium catenella in south-central Alaska

• Main Authors: Wynne, Timothy T., Kibler, Steven R., Sabo, Alexandria, Pokrzywinski, Kaytee
• Format: Conference Object
• Language: unknown
• Published: 2022
• Subjects: water temperature; sea surface temperature; Alaska; Alexandrium catenella; forecasting; Archipelago; Kodiak
• Online Access: https://zenodo.org/record/7013561; https://doi.org/10.5281/zenodo.7013561

The toxic harmful algal bloom species Alexandrium catenella has been observed to bloom in Alaskan coastal waters when sea surface temperatures (SST) exceed approximately 8oC. Consequently, acquiring, fine scale SST data for Alaskan coastal waters represents a critical first step in developing ecological models capable of predicting the occurrence of toxic Alexandrium blooms in this region. Remotely sensed satellite SST records represent the most comprehensive SST data set, but before those data can be used, the satellite data require validation which was the goal of this study. Specifically, we compare a remotely sensed monthly sea surface climatological data set produced by NASA’s Jet Propulsion Laboratory with seven meteorological buoys provided by the National Ocean and Atmosphere Administration (NOAA) National Data Buoy Center (NDBC). The comparisons were done on a point-to-pixel basis as well as an aerial estimation method. The selected study area, the Kodiak Archipelago in Alaska (U.S.A.), is characterized as possessing a varied coastline, filled with a number of coves and embayments. Three of the NDBC buoys are located in embayments and an additional one is in the ~40 km wide Shelikof Strait, with the remaining three offshore. The remotely sensed and in situ measurements were highly correlated providing the basis for the extraction fine scale SST data for the Alaska region over the past 10+ years which can be subsequently incorporated into bloom prediction models.