SIR 2007-5100


Prepared in cooperation with
St. Johns River Water Management District

Effects of the Temporal Variability of Evapotranspiration on Hydrologic Simulation in Central Florida

Andrew M. O'Reilly


     Select an option:


      Purpose and Scope
      Previous Studies
      Hydrologic Conditions in Central Florida
Data Collection and Analysis Methods
      Field Measurements
      Hydrologic Models
      Evapotranspiration Approximations
Comparison of the Temporal Variability in Precipitation and Evapotranspiration
Hydrologic Model Calibration
      Lake Wales Ridge
      Tiger Bay
      Blue Cypress Marsh
      Disney Wilderness Preserve
      Lyonia Preserve
Effects of Evapotranspiration Approximations on Hydrologic Simulation
      Lake Wales Ridge
      Tiger Bay
      Blue Cypress Marsh
      Disney Wilderness Preserve
      Lyonia Preserve
Factors Influencing the Adequacy of Evapotranspiration Approximations


      The transient response of a hydrologic system can be of concern to water-resource managers, because it is often extreme relatively short-lived events, such as floods or droughts, that profoundly influence the management of the resource. The water available to a hydrologic system for stream flow and aquifer recharge is determined by the difference of precipitation and evapotranspiration (ET). As such, temporal variations in precipitation and ET determine the degree of influence each has on the transient response of the hydrologic system.
      Meteorological, ET, and hydrologic data collected from 1993 to 2003 and spanning 1- to 3 2/3 -year periods were used to develop a hydrologic model for each of five sites in central Florida. The sensitivities of simulated water levels and flows to simple approximations of ET were quantified and the adequacy of each ET approximation was assessed. ET was approximated by computing potential ET, using the Hargreaves and Priestley-Taylor equations, and applying vegetation coefficients to adjust the potential ET values to actual ET. The Hargreaves and Priestley-Taylor ET approximations were used in the calibrated hydrologic models while leaving all other model characteristics and parameter values unchanged.
      Two primary factors that influence how the temporal variability of ET affects hydrologic simulation in central Florida were identified: (1) stochastic character of precipitation and ET and (2) the ability of the local hydrologic system to attenuate variability in input stresses. Differences in the stochastic character of precipitation and ET, both the central location and spread of the data, result in substantial influence of precipitation on the quantity and timing of water available to the hydrologic system and a relatively small influence of ET. The temporal variability of ET was considerably less than that of precipitation at each site over a wide range of time scales (from daily to annual). However, when precipitation and ET are of similar magnitude, small errors in ET can produce relatively large errors in available water, and accurate estimates of actual ET are more important. Local hydrologic conditions can also be an important factor influencing the hydrologic response to ET variability. Various points along a flow path in a hydrologic system respond differently to temporal variations in ET. For example, soil moisture contents in the root zone are sensitive to daily variations in ET, whereas spring flow responds to only longer term variations in ET.
      Both the Hargreaves and Priestley-Taylor equations for potential ET, when applied with an annually invariant monthly vegetation coefficient derived from comparison of measured ET with computed potential ET values, can be used with a hydrologic model to produce reasonable predictions of water levels and flows. Baseline-adjusted modified coefficients of efficiency for simulated water levels ranged from 0.0, indicating that water levels were simulated equally as well with approximated ET as with actual ET values, to -0.6, indicating that water levels were simulated better with actual ET values. Simulations using the Hargreaves approximation consistently yielded larger absolute and relative errors than the Priestley-Taylor approximation. However, the differences between the Hargreaves and Priestley-Taylor simulations generally were much smaller than differences between these simulations and the simulations using actual ET. This suggests that the simpler Hargreaves equation may be an adequate substitute for the more complex Priestley-Taylor equation, depending on the level of accuracy required to satisfy the particular modeling objectives.

Suggested Citation:

O’Reilly, A.M., 2007, Effects of the Temporal Variability of Evapotranspiration on Hydrologic Simulation in Central Florida: U.S. Geological Survey Scientific Investigations Report 2007-5100, 36 p.

For additional information, contact:

U.S. Geological Survey
Florida Integrated Science Center
12703 Research Parkway
Orlando, FL 32826

Andrew M. O'

Accessibility FOIA Privacy Policies and Notices
U.S. Department of the Interior, U.S. Geological Survey
Persistent URL:
Page Contact Information: USGS Publishing Network
Page Last Modified: Thursday, 01-Dec-2016 19:49:20 EST
FirstGov button  Take Pride in America button