Reference (potential) evapotranspiration. II: Frequency distribution in humid, subhumid, arid, semiarid, and mediterranean-type climates

While evapotranspiration (ET) has been a subject of research for decades, there is insufficient data, analyses and information on frequency (probability) values and distribution of this important variable for various climatic regions. Frequency distributions of grassreference and alfalfa-reference (potential) evapotranspiration (ET_o and ET_r) were quantified and analyzed using long-term measured climate datasets for five locations in the United States that have significantly different climatic characteristics [Mediterranean-type climate, Davis, CA (ET_o); arid, Phoenix, AZ (ET_o); humid/subtropical, Gainesville, FL (ET_o); a transition zone between subhumid and semiarid climate, Clay Center, NE (ET_r); and a semiarid climate, Scottsbluff, NE (ET_r)]. Two Nebraska locations presented higher variability of ET_ref among all locations because of their highly turbulent climatic characteristics with abrupt changes in wind speed, humidity and air temperature. At Davis, the peak ET_o month is July with an average of 7.01 ± 1.33 mm=d. During this time, the probability of ET_o was between 6 and 8 mm=d with 78% of probability of occurrence. The peak ET_o month in Phoenix is June with a long-term average of 7.37 ± 1.26 mm=d. During this time, there is a 51% probability of ET_o being between 7 and 8 mm=d, with a 94% of probability if the range is expanded to 6 and 9 mm=d. For the months with higher precipitation, ET_ref exhibited more variability atributbale to the erratic rain intensity and frequency (July and August ET_o estimates present a standard deviation of 1.45 and 1.58 mm=d, respectively). This rainy season produces high ET_o variation. ET_o frequency curve in Gainesville had more gradual increase and decrease with much smother fluctuations between the months than any other location. Unlike other locations, there is not a clear peak ET_o month in Gainesville; May has the highest long-term average ET_o (4.6 ± 1.18 mm=d), but June and July averages present less than 0.4 mm=d of difference than May. For May, the probability of ET_o being between 4 and 5 mm=d is 44%. For April and June, the probability of this occurrence is 47 and 34%, respectively. The peak ET_r month at Clay Center is June with a long-term average of 7.20 ± 1.89 mm=d. The standard deviation of ET_r at Clay Center throughout most of the year is higher than those in other locations with the highest values from April to June. ET_r estimations in the peak month (June) are between 7 and 8 mm=d with a probability of occurrence of 18.7%. The peak ET_r month in Scottsbluff is July (7.78 ± 1.41 mm=d). During this month, the most likely occurrence in ET_r is between 7 and 8 mm=d with a probability of 25%; if the ET_r range is expanded from 6 to 8 mm=d, the probability of this ET_r range increases to 49%. If another 1 mm is added to the frequency interval, the probability would increase to 63%, showing the substantial variability of ET_r in this location as a function of abrupt climatic patterns. The ET_o and ET_r frequency distribution data and information presented in this study are among the first datasets for various climatic conditions and can be invaluable for water resources and planning and allocation, irrigation management as well as designing irrigation systems and other water resources-related infrastructures.