JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 89, NO. C6, PAGES 10,599-10,614, NOVEMBER 20, 1984 MesoscaleVariability in Marine Winds at Mid-Latitude JAMES E. OVERLAND AND JUDITH G. WILSON Pacific Marine Environmental Laboratory,NOAA, Seattle, Washin•Iton Wind data were collected by the National Oceanic and Atmospheric Administration WP-3D aircraft on low-level(50 and 90 m) crosswind and along-mean-wind tracks of approximately350 km during the Storm Transfer and Response Experiment in November and December 1980. Observed mesoscale vari- ationsin the marine wind fieldsare characterized by the velocity correlation tensor for three atmospheric regimes:cloud streets, open and closedcellular convection, and prefrontal warm air advection.The dominant scale of mesoscale variation in the offshore wind field normal to the mean wind direction in the caseof cold continentalair flowing over a warmer ocean,producingcloud streets, was 27 km. For this case, the standard deviation in momentum transfer, whichwascalculated from 2-km subsets of the flight track by the bulk aerodynamic method assuming a constant drag coefficient, was 13% of the synoptic scale (330 km) mean.The dominantscale of mesoscale variationfor opencellularconvection was 62 km, and the dominant scale for closed cellular convection was 90 km. The standard deviation of mesoscale momentum transfer (scales greater than 2 km; constant drag coefficient) for a 345-km flight track containingboth cell types was 26% of the synopticscalemean. The warm air advectioncasehad no measurable mesoscale variability. For each regimea model of the horizontal velocity correlation tensor, which can be used to estimate a mesoscale variability,is fitted to the observed velocity correlation tensor with velocity componentand weather regime dependent coefficients. This general model is consistent with an interpretation of the mesoscale wind field as an ensemble of coherent structures, associated with cloud type, in which the spatial variability of the wind field in each weather regime is associated with physically determined dominant length scales (i.e., cells or rolls), ascontrasted with a continuum interpre- tation of two-dimensional turbulence. To accurately describe regionalwinds and fluxes at the seasurface, wind speed and temperature data should be averaged over the dominant mesoscale length scalewith either a suitable time averageor a spatial average, such as can be obtained by scatterometry, or an estimate of the mesoscale variability should be explicitlystated. It is also suggested that enhanced vertical flux in the oceanic mixed layer occurs at length scales of atmospheric boundary layer structures. INTRODUCTION Mesoscale variability introduces uncertaintyin determining fields of wind, temperature, and humidity in the marine boundarylayer. To performregionalscale analyses or to esti- mate the transfer of sensible and latent heat and momentum at the sea surfaceusing bulk aerodynamicmethods,routine atmospheric measurements must be made by sufficient averag- ing over time to suppress this variability, or, better, the varia- bility should be explicitly stated in estimating surfacelayer windsand fluxes. New wind measurement technology by satel- lite scatterometry involves spatial averages, and it is clear that an understanding of mesoscale variability is necessary for the interpretation of scatterometer observations. Beyond the im- portanceof mesoscale variationsin the estimationof the syn- optic wind and air-sea fluxes,ocean temperature and current fluctuationsat a depth of 20 m have been shown to correlate with the passageof mesoscale atmosphericconvection cells [Trump et al., 1982]. Indeed, modeling studies [Orlanski and Polinsky, 1983] show that atmospheric forcing on scalesof order [100 km] can be very effective in explaining mesoscale oceanic variability. Many authors have reported a region of low variance or gap in the mesoscale portion of horizontal wind spectra. The shapeof the mesoscale gap is a function of the synopticwind and atmospheric stability [Pierson, 1983]. Smedman-Hiigstrb'm and Hiigstriim [1975] found that as stratification increases, the spectral gap moves to' higher normalized frequency,and the spectracontain lessenergy. Van der Hoven [1957] and Har- rington and Heddinghaus [1974] found spectralgaps over land This paper is not subject to U.S. copyright. Published in 1984 by the AmericanGeophysical Union. Paper number 4C0895. near a period of 1 hour. However, Burt et al. [1974] found significant variance, which they attributed to roll vortices, at periods near 1 hour for a variety of wind speedsover the oceanand found no evidence of a clear mesoscale gap. Several studies [Thompson, 1973; Tennekes, 1978; Ga•te, 1979; Lilly, 1983; Lilly and Petersen, 1983; Nastrom and Gage, 1983] model the mesoscale region of the energy spectrumfor the atmosphere above the boundary layer as an en- ergy/enstrophy inertial range characteristic of two- dimensional turbulence.Boundary layer studiesoften invoke two-dimensionalturbulence conceptsin transforming time series data to the spacedomain. In contrast, we propose a stochastic model of the mesoscale wind field that emphasizes the presence of organized atmospheric motion in the meso- scale as an alternative to two-dimensional turbulence models. Estimates of mesoscalevariability were calculated from flight level data collectedby the National Oceanic and Atmo- spheric Administration (NOAA) WP-3D aircraft during the Storm Transfer and Response Experiment (STREX) in the northeast Pacific Ocean in November and December 1980 [Fieagileet al., 1982]. We analyze three crosswindand three along-wind tracks flown at 50 or 90 m altitude of approxi- mately 350 km in length obtained in three atmosphericre- gimes: postfrontal mesoscalecellular convection (MCC), strong cold air advection producing cloud streets, and pre- frontal warm air advection. These regimes are representative of the range of mid-latitude marine air masses. This paper has two objectives: (1) to estimatethe impor- tanceof mesoscale fl (15-200 km) and mesoscale • (2-15 km) variability in the atmospheric boundarylayer in the measure- ment of wind and transfer of momentum and sensible and latent heat at the sea surface, and (2) to developa stochastic model to characterize the mesoscale variability of the atmo- 10,599