Satellite-measured seasonal and inter-annual chlorophyll variability in the Northeast Pacific and Coastal Gulf of Alaska

Four years (September 1997–August 2001) of SeaWiFS ocean-color satellite data are used to present the first synoptic quantification of chlorophyll variability on seasonal and interannual timescales over the Northeast Pacific and Coastal Gulf of Alaska. Low light levels and/or cloud during November–January prevent examination of winter patterns. The climatological seasonal patterns show shelf-intensified chlorophyll (by a factor of 3 or more) around the entire basin, extending up to 300 m beyond the shelfbreak along the eastern margin and south of Kodiak Island, with peaks in May and again but weaker in August–September. Over the central basin, concentrations are minimum in February (<0.4 mg m−3), have no spring peak, and increase steadily over the season to a maximum (∼0.7 mg m−3) in September–October. The dominant mode (27% of total variance) of an empirical orthogonal function (EOF) decomposition of the monthly image time series shows interannual variability is strongest during the spring peak around the periphery of the basin with minima in 1998 and 2001. The second and third modes (8% and 7%) capture additional spring interannual variability in patterns strongly related to the shelfbreak west of Kayak Island (144°W) and in more diffuse mesoscale patterns in the eastern GOA. Cross-shelf chlorophyll variability (0–400 km offshore) at three locations contrasts strong seasonality and spring bloom development offshore, over the shelf break, in the northern GOA (off Seward) with weaker seasonality over the shelf off the Queen Charlotte Islands. The shelf between Prince William Sound and Kodiak Island consistently develops a spring along-shelf gradient in chlorophyll, maximum over the western end of the shelf. Comparisons to wind and surface temperature anomalies suggest that interannual differences in the strength of the spring bloom are more strongly related to wind forcing the previous winter than temperature anomalies associated with the 1997/98 El Nino.