Satellite-derived estimates of primary production on the northwest Atlantic continental shelf

Abstract

Synoptic estimation of rates of primary production from remote observation of ocean color is essential for determination of the ocean's role in global carbon cycles. Models having a physiological and optical basis have been developed to predict water-column primary production, normalized to water-column biomass, as a linear function of sea-surface irradiance. Such a model is applied using 41 successive images of sea-surface chlorophyll concentration derived from the Coastal Zone Color Scanner (CZCS) and covering the northeast coast of the U.S. and Canada in the spring of 1979. The CZCS level III data set for the northwest Atlantic continental shelf includes estimates of surface chlorophyll concentration and percent cloud cover on a 10 km grid. Linear regression techniques are used to calibrate the CZCS chlorophyll estimates with ship measurements and to convert percent cloud cover into sea-surface irradiance; such satellite-derived irradiance estimates are significant predictors of irradiances recorded at the earth's surface. The satellite-derived primary production values have been composited as monthly images. The enhanced production of the spring bloom is evident in areas corresponding to major fisheries such as Georges Bank and offshore Yarmouth, Nova Scotia, while the lower production in the Gulf Stream and an associated warm-core ring can be seen farther to the south. Bottom topography seems to have a major influence on rates of primary production, particularly in waters shallower than 100 m. The relative error estimated for these calculations of primary production is approximately 75%.