Procedures that may be used to evaluate the operational performance of a wide spectrum of geophysical models are introduced. Primarily using a complementary set of difference measures, both model accuracy and precision can be meaningfully estimated, regardless of whether the model predictions are manifested as scalars, directions, or vectors. It is additionally suggested that the reliability of the accuracy and precision measures can be determined from bootstrap estimates of confidence and significance. Recommended procedures are illustrated with a comparative evaluation of two models that estimate wind velocity over the South Atlantic Bight.

Two coccolithophore blooms in the Gulf of Maine were studied in 1988 and 1989. Each bloom was about 50,000 km2 in area and confined to the top 20 m of the water column. Maximal cell concentrations were 2,000 cells ml−1 and coccolith densities of 3 × 105 ml−1 were observed. The coccolith : cell ratio was highest in the bloom center (region of most intense reflectance) and lowest at the bloom periphery, an indication of varying organic vs. inorganic C production. Chlorophyll concentrations were generally low within the bloom and no relation could be observed between major nutrients and coccolithophore abundance. Backscattered light was profoundly affected by coccolith density and was slightly wavelength‐dependent. We calculated total backscattering as well as backscattering (bb) caused exclusively by coccoliths and derived the algorithm relating coccolith density to backscattering. Although cells were efficient light absorbers, coccoliths showed negligible light absorption. Diffuse attenuation was lowest in the green and blue‐green part of the visible spectrum. At the center of the bloom, coccoliths contributed >75% of the backscattering signal and > 50% of the beam attenuation signal. The most accurate way to estimate coccolith concentrations via remote sensing is to measure water‐leaving radiance in the green wavebands.

Natural oil seepage in the Gulf of Mexico causes persistent surface slicks that are visible from space in predictable locations. A photograph of the sun glint pattern offshore from Louisiana taken from the space shuttle Atlantis on May 5, 1989, shows at least 124 slicks in an area of about 15,000 km2; a thematic mapper (TM) image collected by the Landsat orbiter on July 31, 1991, shows at least 66 slicks in a cloud‐free area of 8200 km2 that overlaps the area of the photograph. Samples and descriptions made from a surface ship, from aircraft, and from a submarine confirmed the presence of crude oil in floating slicks. The imagery data show surface slicks near eight locations where chemosynthetic communities dependent upon seeping hydrocarbons are known to occur on the seafloor. Additionally, a large surface slick above the location of an active mud volcano was evident in the TM image. In one location the combined set of observations confirmed the presence of a flourishing chemosynthetic community, active seafloor oil and gas seepage, crude oil on the sea surface, and slick features that were visible in both images. We derived an analytical expression for the formation of floating slicks based on a parameterization of seafloor flow rate, downstream movement on the surface, half‐life of floating oil, and threshold thickness for detection. Applying this equation to the lengths of observed slicks suggested that the slicks in the Atlantis photograph and in the TM image represent seepage rates of 2.2–30 m3 1000 km−2 d−1 and 1.4–18 m3 1000 km−2 d−1, respectively. Generalizing to an annual rate suggests that total natural seepage in this region is of the order of at least 20,000 m3 yr−1 (120,000 barrels yr−1).

No abstract

No abstract