Stochastic Analysis of Air and Water Temperatures

Water temperature data in streams and lakes are relatively difficult to obtain compared with air temperature data and other related hydrologic and meteorologic data. A procedure is proposed whereby synthetic daily water temperatures in a natural stream can be generated using available real or synthetic daily air temperature data and other physical parameters of the watershed. In addition to the seasonal variations, the short-term variations of the water temperature due to the atmospheric temperature fluctuations are included in this model as well as the purely random fluctuations. The results of the regression analysis of some available data in Minnesota suggest that the response coefficient relating the water temperature fluctuations and the air temperature fluctuations can be estimated using physical parameters of the watershed alone.