Analysis of ocean surface heat fluxes in a NOGAPS climate simulation: Influences from convection, clouds and dynamical processes

摘要

This study examines the simulation quality of the surface heat flux fields produced during a climate simulation of the Navy Operational Global Atmospheric Prediction System, version 3.4, with a reduced spectral truncation of T63 and Is levels (herineafter referred to as NOGAPS-CL). Comparisons are made between a 17-year NOGAPS-CL simulation using monthly sea surface temperatures as surface boundary conditions and a number of validating data sets consisting of ship, satellite, and/or reanalysis-based surface heat fluxes, precipitation, top of the atmosphere radiation budget, water vapor, cloud frequency, surface wind stress, and tropospheric winds. In this extended, long-range integration, NOGAPS-CL underpredicts the net surface shortwave flux in much of the subtropical oceans and overpredicts the net shortwave flux in the western Pacific warm pool and the midlatitude oceans, when compared to several satellite-derived climatological data sets. In addition, NOGAPS-CL over predicts the latent heat flux in much of the subtropics and under predicts the latent heat flux over the northern ocean western boundary currents and under the storm track regions that extend eastward from them. These shortwave and evaporation biases combine to produce errors in the surface net heat flux, with too little heat entering the subtropical/tropical oceans and too much heat loss in the midlatitudes oceans. Examination of related quantities indicates that the tropical climate biases are coupled to shortcomings in the convective cloud and/or boundary layer parameterizations which leads to the premature release of moist instability from the boundary layer in regions just outside the deep convective zones. This leads to enhanced climatological cloudiness, rainfall, and surface evaporation, as well as to a reduction in the surface shortwave flux and outgoing longwave radiation (OLR), in the subtropical regions. Furthermore, because of this early release of the moist static energy, there is a reduction in clouds, rainfall and water vapor content, as well as enhanced surface shortwave flux and outgoing longwave radiation, in the deep convective zones. The reduction in rainfall and enhanced OLR reduces the strength of the tropical large-scale circulation, which in turn reduces the strength of the subsidence in the subtropical regions which normally acts to suppress the convection processes in these regions. The implications of these results are discussed in terms of the relationship among the forecast model climatological surface fluxes, convection, clouds, and the dynamical processes, as well as their similarities to other climate models and their possible impact on the simulation of transient systems. [References: 59]