ABSTRACT The suitability of three evapotranspiration models (Penman-Monteith, Penman, and a modified Priestley-Taylor) was evaluated at a site of successional vegetation in a deforested area of the Lake Wales Ridge, Florida. Eddy correlation measurements of evapotranspiration made during 22 about 1-day periods at a temporal resolution of 20 minutes from September 1993 to August 1994 were used to calibrate the evapotranspiration models. Three variants of the eddy correlation method that ascribe measurement error to three different sources were considered in the analysis. The Penman-Monteith and modified Priestley-Taylor models were successful in approximating measured 20-minute values of evapotranspiration (r2 = 0.918). The most successful approaches were the modified Priestley-Taylor model (r2 = 0.972) and a nontraditional and simplified form of the Penman-Monteith model (r2 = 0.967). The Penman approach was unsuccessful as a predictor of evapotranspiration.
The evapotranspiration models were used to estimate evapotranspiration between measurements. When evapotranspiration values measured with a Bowen ratio variant of the eddy correlation method were used for model calibration, estimated daily evapotranspiration rates varied seasonally ranging from 0.2 millimeters per day (0.008 inch per day) in late December 1993 to 5 millimeter per day (0.2 inch per day) in mid-July 1994. Annual evapotranspiration (September 15, 1993, to September 15, 1994) was estimated to be about 680 millimeters (27 inches). Evapotranspiration models calibrated to the standard eddy correlation method and to an energy-balance residual variant provided estimates of annual evapotranspiration that were about 10 percent lower and higher, respectively. These data indicate that of the 1,320 millimeters (52 inches) of precipitation during the 1-year period, about 570 to 700 millimeters (22 to 28 inches) recharged the surficial aquifer. Evapotranspiration at this study site probably defines the lower limit of evapotranspiration from vegetated surfaces in central Florida because of the shallow-rooted plants, rapidly-drained soils, and relatively deep water table.