Abstract
A technique is described to determine 24-hr stream-bed evaporation subject to the constraints that it be inexpensive, uncomplicated, and accurate to within ± 20%. Using maximum thermometers, a totalizing anemometer, and pans filled with dry sand, evaporation E is predicted from E = (b1δT + b2δU)δT, where δT is the difference between the maximum temperature of the dry sand and the maximum temperature of the stream bed, δU the 24-hr wind passage obtained from the totalizing anemometer, and b1 and b2 are constants peculiar to each soil type and are determined by a statistical curve-fitting method on experimental data. The result of statistical evaluation of the prediction equation on the experimental data gives an average percentage error of 13% and a standard error of 0.88 mm. Using this technique, it was determined that evaporation is not important in the annual hydrological balance of a stream bed.
