An evaluation of oceanographic radiometers and deployment methodologies

The primary objective of the Sea-viewing Wide Field-of-view Sensor (SeaWiFS ) Project is to produce water-leaving radiances within an uncertainty of 5% in clear-water regions, and chlorophyll a concentrations within 35% over t he range of 0.05-50 mg m(-3). Any global mission, like SeaWiFS, requires va lidation data from a wide variety of investigators. This places a significa nt challenge on quantifying the total uncertainty associated with the in si tu measurements. because each investigator follows slightly different pract ices when it comes to implementing all of the steps associated with collect ing field data, even those with a prescribed set of protocols. This study u ses data from multiple cruises to quantify the uncertainties associated wit h implementing data collection procedures while using different in-water op tical instruments and deployment methods. A comprehensive approach is under taken and includes (a) the use of a portable light source and in water inte rcomparisons to monitor the stability of the field radiometers, (b) alterna tive methods for acquiring reference measurements, and (c) different techni ques for making in-water profiles. Three optical systems had quadrature sum uncertainties sufficiently small to ensure a combined uncertainty for the spaceborne and in situ measurements within a total 5% vicarious calibration budget. A free-fall profiler using (relatively inexpensive) modular compon ents performed best (2.7% quadrature sum uncertainty), although a more soph isticated (and comparatively expensive) profiler using integral components was very close and only 0.5% higher. A relatively inexpensive system deploy ed with a winch and crane was also close, but ship shadow contamination inc reased the quadrature sum uncertainty to approximately 3.4%.