Weekly evaporations over a period of 11 years for Elk City Lake in Southeast Kansas are estimated by using the mass-transfer method. A mathematical model is developed for the weekly evaporation series. The model consists of a deterministic periodic component and a stochastic component, represented, respectively, by a Fourier approximation and an ARMA model. Monthly and annual lake evaporations are also obtained. A comparison of the distribution of the mean monthly lake evaporation with that for Class-A pan evaporation shows clearly the hysteresis effect of heat storage in the lake. The estimated annual lake evaporation at a 2% chance of exceedence is 1713.2 mm with a standard deviation of 131.0 mm. The effects of errors in the estimated model parameters and the input variables to the mass-transfer equation on evaporation estimates are determined through sensitivity analyses.
ASJC Scopus subject areas
- Water Science and Technology