WATER RESOURCES RESEARCH, VOL. 28, NO. 6, PAGES 1605-1617, JUNE 1992 Energy Budget Evaporation From Williams Lake' A Closed Lake in North Central Minnesota A.M. STURROCK Water Resources Division, U.S. Geological Survey, Stennis Space Center, Mississippi T. C. WINTER AND D. O. ROSENBERRY Water Resources Division, U.S. Geological Survey, Denver Federal Center, Denver, Colorado Evaporation from Williams Lake, computed by the energy budget method for the five open-water seasons of 1982-1986, varied from a maximum seasonal rate of 0.282 cm/d in 1983 to a minimum seasonalrate of 0.219 cm/d in 1982. The pattern of monthly values of evaporation is not consistent from year to year. The normally expected pattern of low evaporation values in May, followed by increasing values in June to maximum values in July is true for only 3 of the 5 years. Comparisonof annual evaporation calculated by the energy budget and mass transfer methods indicates that energy budget values varied from 13% greater to 11% lessthan masstransfer values. Furthermore, there is no seasonalbias in the pattern. Large differences exist in the magnitude of energy fluxes to and from Williams Lake. By far the greatestenergy fluxes, having magnitudes of hundredsof watts per square meter, are incomingsolar radiation, incomingatmospheric radiation, and outgoing long-wave radiation emitted by the lake water. The least energy fluxes are related to advection, which generally have magnitudes less than 5 W m -2. INTRODUCTION Background Water lossby evaporationcommonlyis one of the largest components of a lake's water balance [Winter and Woo, 1990]. Despite its importance, evaporation is seldom deter- mined by one of the more accurate methods because a considerable investment in instruments and personnel is required. For example, determining evaporation by an en- ergy budget is considered to be one of the most accurate methods for long-term, continuousmonitoring of evapora- tion [Harbeck et al., 1958; Gunaji, 1968], but becauseof the money and personnel needed, these types of studies have beendone largely by government agencies responsible for managing large reservoirs in the western United States. Most research and studies of evaporation by the energy budget method have been done in these arid and semiarid regions because accurate accounting of water resourcesis particularly critical. The relatively lesser importance of knowing evaporation accurately in more humid regions has resulted in less research on methods such as the energy budget in the central and eastern United States. Therefore some assumptions about various aspects of energy fluxes have not been critically evaluated, particularly for small lakes in humid regions. Insights into energy budgets provided by the studies of Lake Mendota in Wisconsin [Dutton and Bryson, 1962], Pretty Lake in Indiana [Ficke, 1972], and Perch Lake in Ontario, Canada [Robertsonand Barry, 1985], were impor- tant contributions to evaporation research, but money and personnel limitations restricted the scope of the studies so that some questions that needed attention were not ad- dressed. Furthermore, the improvement in instrumentation This paper is not subjectto U.S. copyright. Published in 1992by the American Geophysical Union. Paper number 92WR00553. in recent years gave impetus to additional research on the energy budget method of determining evaporation. For example, the importance to a lake energy budget of heat flux to and from sediments has been addressed in only a few studies [Hughes, 1967], and the importance to the heat budget of groundwater flux to and from lakes has not been evaluated at all. In addition, unreliable Bowen ratios that are caused by rapidly changinggradients of temperature and vapor pressure over the water surface result in inaccurate energy budget evaporation values. This problem needs to be evaluated for a variety of seasonal climatic conditions. Research on lake evaporation is needed also because of the common practice of using empirical methods to estimate evaporation. The uncertainties related to various empirical methods can best be evaluatedby comparison to an accurate standard;therefore it is desirable to develop an understand- ing of the more accurate methods, such as the energy budget, for a wide range of climatic and physical settings. Purpose and Scope To address the problemsidentified above, one of the goals of the U.S. Geological Survey's Hydrology of Lakes Project is to evaluate the energy budget method of determining hydrologic fluxes to and from small lakes. Four primary field experimental lake/watershed sites for long-term studiesand several secondary sites for short-term studies have been established throughout the United States. Williams Lake, a small closed lake in north central Minnesota, was selected for long-term studies of lake hydrology because it is in an area of abundant natural lakes situated in glacial terrain, it has no surface water inlets or outlets, and it is in a climatic settingwhere annual precipitation is approximately equal to annual evaporation [Siegel and Winter, 1980]. This paper presents an evaluation of the energy budget approachfor determining evaporation from Williams Lake. The evaluation includes analysis of seasonal and annual 1605