The Surface Heat Balance at Night -- Some Sub-Saharan Data

Comparative studies of eddy correlation and Bowen ratio approaches over arid farmland in south-central Africa provide some insights into the causes of the commonly-reported discrepancies in surface energy balance calculations. In the African case, the site is largely unaffected by terrain irregularities, but is at an elevation of about 1500 m ASL. Several independent measurements are made of such properties as surface (IR) and subsurface temperatures, temperature and humidity gradients in air, and net radiation. In daytime, sensible heat fluxes measured by eddy correlation agree with those yielded by Bowen ratio methods (after application of the correction for sonic anemometer obstruction). At night, however, there is a consistent difference, with the eddy correlation values of (w'T') ̅ showing less nighttime cooling than the BREB values. The availability of replicated IR surface temperature measurements allows some alternative explanations to be explored – on the one hand involving a refinement of the conventional energy balance assumptions and on the other hand exploring the consequences of omission of potentially significant atmospheric transport mechanisms through the use of half-hour averaging periods. Dewfall and its repercussions are possible contributors; dewfall appears to have been frequent, but the data do not allow its magnitude to be determined.