Evaluating Evapotranspiration by Some Mathematical Models and the National Document of Irrigation Compared to Lysimeter Data in the High Plain of Shahrekord

Reference crop evapotranspiration (ETref) is measured directly (lysimeter
method) or estimated indirectly (mathematical models). In this study, daily
evapotranspiration (ET) was calculated by some mathematical models for
alfalfa and was compared with ET data gathered daily from a drainage
lysimeter during six months. To evaluate and select the best model, ET
data values were compared using statistical criteria including R2, NRMSE,
MAE, MBE, and d. Results showed that the best models in the daily timeframe
were: Hargreaves-Samani (HS), Jensen-Haise (JH) and Turc. For
the HS model, values of NRMSE and d were 0.126 and 0.930, and MAE
was 0.477 (mm.d-1), respectively. The best models for monthly time-frame
were HS, American Society of Civil Engineering-Penman-Monteith
(ASCE-PM), and Turc. Generally, HS model had the highest R2 values of
0.985, 0.998 and 0.998 for daily and monthly periods, respectively, and
ASCE-PM had the lowest MBE compared to lysimeter data. To estimate
total ET during the alfalfa growth period in the Shahrekord plain, results
showed that, models have over-estimated, except Turc and Priestley –
Taylor models. ASCE-PM had the nearest ET (1161.7 mm) to lysimeter
data (1157.6 mm). In the national document of irrigation, the alfalfa
irrigation water requirement was reported at 649 mm (in the corresponding
period), which is much less than the value measured.