Variability in coastal zone color scanner (CZCS) Chlorophyll imagery of ocean margin waters off the US East Coast

Abstract

The purpose of our study was to use the 7.5-year coastal zone color scanner (CZCS) image time series (Oct. 1978 to July, 1986) to study general patterns in near-surface phytoplankton chlorophyll concentrations in ocean margin waters off the US East Coast. We defined 21 relatively large study areas (>100 km 2) within the MAB and SAB to set boundaries for averaging and subsequent analyses. Our objective was to partition the observed CZCS-derived chlorophyll concentration (CSAT, mg m −3) variability of these 21 study areas within three general categories based on time scale: daily (i.e. day–week), seasonal and interannual. An additional objective was to determine relations between the temporal patterns in the 21 study areas. All available CZCS imagery (more than 3500 scenes of Level 1 imagery, i.e. top-of-the-atmosphere radiance in satellite swath coordinates) covering some or all of our area of interest (northwest Atlantic off the US East Coast) were obtained at full resolution, processed to Level 2 (water-leaving radiance, chlorophyll concentration and other derived products in satellite swath coordinates) and mapped to two different study regions located off the southeast and northeast coasts of the US. Satellite-derived estimates of near-surface chlorophyll concentrations (CSAT) were extracted on a pixel-by-pixel basis from each of the 21 study areas (chosen based on oceanographic criteria) from each of the daily composite CSAT images. For each image and when satellite coverage permitted, CSAT values were averaged to yield a time series of daily mean values for each of the 21 study areas. We used three basic approaches to quantify temporal and spatial patterns in the 21 time series: (1) multiple linear correlation, (2) structure functions (semi-variance calculations) and (3) empirical orthogonal functions (EOF). Our results show: (1) a simple annual CSAT cycle common to all ocean margin waters along the entire US East Coast, consisting of a broad peak in CSAT concentration during winter and minimum concentrations during the summer; (2) relatively subtle across- and along-shelf changes to the timing and relative magnitude of the winter CSAT maxima and summer minima, as well as the presence of secondary seasonal peaks in some regions; (3) high variability at time scales of days to weeks superimposed on the seasonal pattern; (4) high spatial coherence of the seasonal component between all 21 study areas; (5) high coherence of the days-to-weeks component between adjacent study areas, but generally low or no coherence for study areas not adjacent or near each other; and (6) detectable, but low interannual variability.