EVALUATION OF LATENT EVAPORATION ESTIMATES AND THEIR CONVERSION TO POTENTIAL EVAPORATION

Abstract

Daily and monthly latent evaporation (LE) estimates obtained from an earlier-described regression-type model are compared with estimates from Penman’s (PE) and Thornthwaite’s (PET) techniques. Penman’s PE was selected as a control. Daily PE estimates were more closely related to LE than to PET estimates, as seen from the coefficients of determination (100 CD) of 52 and 21%, respectively. Similarly, variations of monthly PE means were more closely associated with variations of monthly LE means (65%) than with PET means (45%). Both the LE and PET models use the same standard climatic data as input. The improvement of 31% for the daily values and 20% for the monthly means results from using maximum and minimum air temperatures separately in the LE model, instead of mean air temperature as in the PET technique. The least bias in PE as derived from converted LE estimates was obtained by a factor of 0.0094 cm/cm³ (0.0037 in./cm³). However, a review of literature on ratios of consumptive water use by irrigated crops to LE measurements suggested a factor in the order of 0.0086 cm/cm³ (0.0034 in./cm³). This factor has been successfully employed in irrigation scheduling and water budgeting experiments. The discrepancy could result from the consumptive crop water use for part of the growing season being less than Penman’s PE as computed in this study. Although eventually seasonally adjusted conversion factors based on crop development phases are preferred, the findings warrant the use of the LE model in various climates for estimating either PE, by applying 0.0094 cm/cm³ (0.0037 in./cm³), or seasonal consumptive crop water use, by applying 0.0086 cm/cm³ (0.0034 in./cm³).