Spatial patterns in ocean color and temperature maps: Fronts, fractals, and ecological considerations.

摘要

Heterogeneity in ocean patterns is at once intriguing and problematic. Processes such as primary production or mixing occur on a range of scales in time and in space; variability must be captured at appropriate scales with tools flexible enough to detect changes in pattern-generating processes. Frontal and fractal analyses were applied to satellite-derived sea surface temperature and ocean color imagery to describe patterns in Northwestern Atlantic shelf and slope seas. On a seasonal scale, climatological thermal fronts were most frequent in winter and were found on the continental shelf south of the NY Bight and alongshore in the Gulf of Maine. Summer thermal fronts were least numerous and were concentrated north of the NY Bight around Georges Bank and the Scotian Shelf at the entrance to the Gulf of Maine. Ocean color fronts showed similar frontal frequencies but different distributions: fewest fronts in fall/winter, concentrated in the southern Mid-Atlantic Bight, and distinctive summer fronts along the middle Mid-Atlantic Bight shelf as well as around Georges Bank. Maps of jointly occurring thermal and ocean color fronts outlined regions where biophysical coupling may be strongest: in winter along the shelf break from Cape Hatteras to Chesapeake Bay, and in summer in the Georges Bank/Nantucket Shoals domain. These occurrences were both linked to high plant biomass. On a mesoscale, monthly maps of ocean color fronts were more spatially isolated and discrete, while temperature fronts were more diffuse. Region-wide, monthly AVHRR and CZCS gradients generally trended with frontal activity; both gradients and fronts decreased as spring progressed to summer. Fractal analysis results lead us to accept the hypothesis that fractal models describe temperature (SST) and chlorophyll (CHL) patterns in semi-enclosed regions such as a warm-core ring and the crest of Georges Bank. As warm-core ring 82b aged from April to June its CHL patterns converged to resemble its SST patterns. Three scenarios were suggested concerning the prediction of SST-CHL patterns: a strong mixing regime leads to similar CHL-SST patterns, regardless of biological activity; a low mixing/high growth regime leads to different CHL-SST patterns; and a low mixing/low growth regime leads to similar CHL-SST patterns.