441 



FUTURE DIRECTIONS IN OCEAN SCIENCES 103 



flux in nutrient-poor waters were inconsistent with geochemical 

 data. 



A third possible strategy is to study regions of the ocean that 

 are anomalous vis-a-vis standard paradigms. What controls pro- 

 ductivity in nutrient-rich areas of the sea, such as the subarctic 

 Pacific and the equatorial Pacific? If this question could be an- 

 swered unequivocally, it would indicate significant progress to- 

 ward a general understanding of oceanic productivity. 



In support of all three strategies is continued research on the 

 ecology, physiology, and molecular biology of representative spe- 

 cies from specific oceanographic regimes. Without an understanding 

 of the basic biology of individual organisms, one cannot hope to 

 understand how the marine food web works or to predict how the 

 ecosystem will respond to change. 



Technologies and Approaches for the 1990s 



The pace of scientific progress is often closely coupled with 

 the development and application of new technologies. Several 

 technologies and approaches will aid the study of marine ecosys- 

 tems in the 1990s and could lead to important breakthroughs. 



Satellite Remote Sensing 



By the mid-1990s, three variables will be measured simulta- 

 neously by satellite for routinely characterizing ecosystems and 

 related environmental factors. The three variables are sea surface 

 temperature,- sea surface and near-surface ocean color to deter- 

 mine chlorophyll and water clarity; and sea surface wind fields for 

 estimating rates of vertical mixing, air-sea gas exchange rates, and 

 other wind-related processes, such as the seasonal changes in the 

 depth of the surface ocean mixed layer. 



Numerical Modeling 



Two developments in modeling should make significant con- 

 tributions to ecosystem studies in the 1990s. First, models are 

 being developed that can be used to help form hypotheses regard- 

 ing the role of oceanic biota in global nutrient budgets. These 

 models ultimately will merge basic miathematical descriptions of 

 biogeochemical cycles with general circulation models and, from 

 given starting conditions, will attempt to predict the evolution of 

 fluxes over time. Global models will be particularly useful con- 



