Comparative measurements of the energy fluxes over a pine forest

During the summer of 1974, simultaneous heat flux measurements were made over a pine forest in three ways: (a) Using the Bowen ratio - Energy Budget technique, with the Bowen ratio estimated from temperature and humidity profiles measured with two pairs of wet-and dry-bulb thermometers at six levels. (b) Using the Bowen ratio - Energy Budget technique, with the Bowen ratio estimated from temperature and humidity measurements using pairs of wet-and dry-bulb thermometers mechanically interchanged every ten minutes. (c) Using the Eddy Correlation - Energy Budget technique, with a direct measurement of the sensible heat flux made by a simple eddy correlation apparatus. Methods (b) and (c) are shown to give similar results apart from a systematic difference of about 25% in the measured sensible heat flux. This corresponds to an underestimate of the sensible heat flux by the eddy correlation apparatus. Although sometimes consistent with the other two methods, on occasions method (a) gave results which were significantly different from both (b) and (c). When differences occurred, they tended to be systematic and persistent over individual days; but they could change magnitude and sign if the particular sensors used at each level in the profile were rearranged. The experimental program used to collect these and previous data involved the rearrangement of sensors on a regular (two day) time scale. It is shown that, when averaged over several such rearrangements, method (a) produces median values of surface resistance which are more in keeping with those produced by the other methods. This is taken to imply that data previously gathered in this way can be used to produce physically reasonable results providing they are averaged over several days. On the basis of the results presented, recommendations are made on future experimental work in forest micrometeorology.