| Abstract: | A model to predict the surface temperature of a variety of surfaces is described. The model solves the surface energy balance equation iteratively, using only standard meteorological data. Since surface and soil temperature information is not required for initialisation, the model is portable and, in theory, could be used for any surface and location. It is shown that, in order to obtain the correct cooling rates for vegetation during the night, the direct influence of the ground flux must be removed from the energy balance equation for the layer of vegetation. A scheme that couples a vegetation canopy to the ground solely by radiation is described, giving satisfactory cooling rates when compared with observations. Observations from a field site at Cardington, near Bedford, UK, are used to test the accuracy of the model for road and grass surfaces. When compared against these data, the model predicts surface temperatures with a root mean square error of about 1 °C for the road and 2 °C for the grass. Data from other sources not only give similar results to the Cardington data, but also demonstrate that the model can reproduce the characteristics of wet and partially dry soils and also dry desert sand. A study of the sensitivity of the model to errors in the forcing data indicates that inaccuracies in the air temperature data lead to similar sized errors in the predicted surface temperatures. Fluctuations in the forcing data that are not resolved by the model will affect a grass surface much more than a road surface, due to the relatively small thermal inertia of the grass. |
