Comparing Different Methods of Potential Evapotranspiration and Studying Temporal and Spatial Changes in the Mahalou Watershed using GIS

Evaluating potential evapotranspiration plays an important role in planning, designing, and managing water resources, especially in arid and semi arid zones. The aim of the present research is considering spatio-temporal changes of different potential evapotranspiration methods such as Thornthwaite, Hargreaves, and Blaney-Criddle in the Maharlou Watershed, Fars province. So, yearly statistics of temperature in a course of 12 years (2002-2013) with 12 meteorology stations in Maharlou Watershed have been used. In order to zonation map of evapotranspiration and studing spatial changes, Inverse Distance Weighting (IDW) method has been used in GIS software. To choose the most suitable method for evaluating potential evapotranspiration using Blany- Criddle, Hargreaves- Samani, and Thornthwaite  in this present research, statistics indicators including Mean Absolute Error (MAE) (8.64, 12.06, 9.93), Mean Bias Error (MBE) (-0.42, -0.56, -1.36), and Root-Mean-Square error (RMSE) (2.49, 2.86, 3.48) is used. Results showed that spatio-temporal changes of potential evapotranspiration in Blaney- Criddle method indicates the most amount of evapotranspiration in Dobene Station in 2009 with the amount of 127.5 mm. Whereas the least that amount of attached is in Kaftar Station in 2006 with the amount of 85.1 mm. Also based on Hargreaves results, the most and least amount of evapotranspiration is in Jahan abad bakhtegan Stations in 2009 with the amount of 152.1 mm and Komhar station in 2003 with the amount of 96.5 mm, respectively. In contrast, based on Thornthwaite method, the most amount of evapotranspiration is included in Dobene Sstation in 2009 with the amount of 103.6 mm and the least that amount of attached is in Komhar Station in 2011 with the amount of 52.7 mm. Subsequently, the results of evaluating of different evapotranspiration methods according to RMSE and MAE showed that Blaney- Criddle method the least error amount (2.49, 8.64) and also according to the MAE and RMSE criteria, the most amount of error attached of Hargreaves- Samani method with the amount measured MAE 12.06 and RMSE 3.48. In general, the Blaney-Criddle method is the best potential evapotranspiration method. On the other hand, the low error rate is due to Hargreaves- Samani and Thornthwaite based methods.