Seasonal Simulation as a Test for Uncertainties in the Parameterizations of a Budyko-Sellers Zonal Climate Model

Abstract

The energy-transport parameterization of Budyko (1969), which was devised to parameterize mean annual net radiation as a function of zonally averaged surface temperature, is subjected to verification with seasonal transport data in order to evaluate its validity for climatic change experiments. It is found that Budyko's linear parameterization is able to describe the annual zonal heat transport divergence for all latitudes and also the seasonal cycle of heat transport divergence at high latitudes (? > 50°), but has no predictive ability for the seasonal deviation from annual average in lower latitudes. The parameterization of infrared flux at the top of the atmosphere as a linear function of zonal surface temperature is tested using seasonal data for latitude zones in which the seasonal cycle of temperature has a large amplitude. The temperature coefficients for the different zones examined are found to differ from each other by as much as a factor of 2. This uncertainty, together with the uncertainty in the strength of the ice-albedo-temperature positive feedback, propagates to an uncertainty in the sensitivity of model global climate to changes in the solar constant. The reduction in solar output required by a simple climate model to generate an ice-covered earth falls roughly in the range of 2 to 21% because of uncertainties in these two radiative coefficients alone. Uncertainty in the transport parameterization would further increase this range.