Oceanic latent heat flux from satellite data

Abstract

This thesis presents a new method of estimating ocean latent heat flux (LHF) using satellite data. The surface layer equations derived from Monin-Obukhov similarity theory are closed with empirical parameterizations, and patched to a mixed layer model yielding a two-layer PBL model. This is the first proof that such a model can be applied in non-strongly convective situations. Two new retrievals of moisture parameters are derived, one for the surface to 500 meter integrated water vapor $(W\sb {B})$ and one for the mixed-layer humidity $(q\sb{m}).$ Inputs include total integrated water vapor retrieved from the Special Sensor Microwave Imager (SSM/I), sea surface temperature optimally interpolated from retrievals from the Advanced Very High Resolution Radiometer and buoy/ship measurements (OI SST), and ECMWF analyzed air-sea temperature difference (T-SST). An analytic relationship between $W\sb{B}$ and near-surface humidity is presented as a replacement for the statistical relationship of Schulz et al. (1993). LHF is then calculated using the new SSM/I-based retrieval of ML-humidity, SSM/I retrieval of wind speed, OI SST, and ECMWF T-SST. Model errors are assessed, and the $q\sb{m}$ method derived in this thesis is found to perform the best. Systematic errors are small, and random errors are 26 W/m$\sp2.$ Monthly averages of LHF have been calculated using all available SSM/I data on a 1$\sp\circ$ by 1$\sp\circ$ grid for 1992-1997. Differences with other published climatologies, both those derived from SSM/I data and from traditional data, have been discussed. Differences between the current work and previous SSM/I methods are evenly split between model parameterization differences and the new moisture retrieval. Errors due to averaging the input variables and due to errors in merchant ship measurements dominate the differences between the current work and traditional climatologies. This analysis establishes the limiting factors in LHF calculation and produces the most accurate LHF climatology to date. It is the first full SSM/I climatology which will be made available to the general scientific community.