Simulation of Marine Latent Heat Fluxes in the Unified Model

The marine latent heat flux plays a key role in the atmospheric energy and hydrological cycles and its accurate representation is important in both numerical weather forecasting and climate models. Relative to both Da Silva's and the NOC1.1 climatology, latent heat fluxes in the Met Office's Unified forecasting/climate model (UM) appear significantly high. This discrepancy is examined both with respect to differences between climatologies and from the point of view of the parametrization scheme.

Both the NOC1.1 and Da Silva's climatology were derived from ship data. A known problem with the NOC1.1 climatology of surface fluxes is that the net flux into the ocean is implausibly high. Grist and Josey (2003) showed that consistency with measurements of ocean heat transport could be improved if the latent heat fluxes in the NOC1.1 climatology were increased and released these revised fluxes as the NOC1.1a climatology. Their adjusted global mean latent heat flux is actually in close agreement with the new satellite-derived GSSTF2.0 climatology, supporting higher values of the latent heat flux, although noticeable regional differences remain.

The atmospheric energy and hydrological cycles impose constraints on the surface fluxes, which are necessarily obeyed by models. In the case of climatologies, the extent to which these cycles are not balanced provides a provides an indication of the accuracy of our knowledge of the various components of these cycles. Using currently vailable climatologies of latent, sensible and radiative fluxes at the surface and the top of the atmosphere, together with climatologies of precipitation and estimates of atmospheric transports from reanalyses, the simulation of the hydrological and energy cycles is considered. This extends the zonal annual mean analysis of the consistency of climatologies performed by Yu et al. (2000) to the regional and seasonal scales. The greatest consistency in the hydrological cycle is achieved using the GPCP2 precipitation climatology and the adjusted NOC1.1a latent heat fluxes, although a global imbalance remains. The position with the energy cycle is more complicated because of uncertainties in the radiative fluxes, but a seasonal analysis, including tendency terms, shows that imbalances in the winter hemisphere are at least as significant as those in the summer hemisphere, indicating that imbalances in this cycle cannot be explained mainly by errors in climatological shortwave fluxes at the surface, as has been suggested.

Comparing the parametrization scheme for latent heat fluxes used in the Unified Model with field observations, it was found that the use of a constant roughness length for moisture and heat over the oceans led to excessive evaporation at high wind speeds. The introduction of revised scheme, agreeing more closely with field observations and also incorporating an allowance for the salinity of sea water, slightly reduces the latent heat flux and dries the boundary layer. In climate simulations there is a slightly improved pattern of mean sea level pressure.

It is concluded that the original discrepancy of about 30 Wm-2 between the modelled and climatological latent heat fluxes is due largely to an underestimate of latent heat fluxes in traditional climatologies, but also to a deficiency in the model's parametrization of the surface transfer and the discrepancy is therefore reduced to about 10 Wm-2.