Scaling reasoning and field data on the sea surface roughness lengths for scalars

The heat and mass transfer over the sea is considered in terms of the sea s urface roughness lengths for scalars, z(OT) for potential temperature theta , and z(Oq) for specific humidity q, or alternatively, in terms of the roug hness-layer scalar increments, delta theta and deltaq. A new scaling reason ing is proposed in support of the familiar square root dependence of the ab ove increments on the roughness Reynolds number, Re-Ou = z(Ou) u(*)/nu, whe re z(Ou) is the sea surface aerodynamic roughness length, u(*) is the frict ion velocity, and nu is the molecular viscosity of the air. Scaling predict ions are validated using data from measurements made by the National Oceani c and Atmospheric Administration's Environmental Technology Laboratory aboa rd the R/V Moana Wave in the Tropical Ocean Global Atmosphere Coupled Ocean -Atmosphere Response Experiment in 1992-93 and the R/P FLIP in the San Clem ente Ocean Probing Experiment in September 1993. Data presented as the dime nsionless scalar increments (or the ratios z(Ou)/z(OT) and z(Ou)/z(Oq)) ver sus Re-Ou show a good agreement with theoretical predictions, especially at Re-Ou. > 2 (over stormy sea). The resulting roughness-length formulations are recommended for practical use in climate and mesoscale air-sea interact ion models.