Vol. 18, No. 2 BALVE and PATEL 311 Journal of Agrometeorology 18 (2) : 311-314 (December 2016) Evapotranspiration (ET) is important process of hydrological cycle. It is combination of two processes like evaporation and transpiration. ET plays major role in the planning and management of waer resources system, irrigation system, design and drainage studies, etc. India is agriculture country and accurate estimation of ET is necessary because overestimation and under estimation of ET cause wastage of valuable source of water, crop yield respectively. On field, the evapotranspiration can be measured by using lysimeter and field plots. But it does not give exact value of the ET. As the field condition and soil strata are different at different places. Such field test takes more time and results are not accurate. These results depend upon field conditions so the result varies over the area. These difficulties in obtaining the reliable results of evapotranspiration have given rise to a number of methods of calculating and predicting the reference evapotranspiration (ETo) from the avaialable meteorological data. The evapotranspiration either ETo or potential evapotranspiration (PET) can be calculated by various approaches like, radiation based, temperature based, mass transfer based or pan evaporation based equations or the fusion of the available meteorological data. Penman- Monteith (FPM-56) method is standard method of determining the ETo and results of it shows better performance when compared with other existing methods. The FPM-56 equation requires the detail meteorological data for evaluation of ETo because it provides consistent ETo values in various region and climates (Allen, et at. 1998, Tabari and Hosseinzadeh, 2011). In recent years, fuzzy logic is applied to the many applications as it is an alternative and effective tool for studying complex phenomenon. FIS have been successfully used in reservoir management (Panigrahi and Mujumdar, 2000), rainfall-runoff problems (Nayak et al., 2005; Yu and Chen, 2005) and in parameters of groundwater flow. Conventional model require more parameters to estimate the ETo for same all regions and climate conditions (Tzimopoulos et al., 2008). The ETo is estimated (Patel and Balve, 2016) using FIS considering parameters of ETo as inputs. The different approaches like temperature based, radiation based, pan evaporation and mass transfer based were used to estimate the evapotranspiration (Balve and Patel, 2016) and comparision is done with FPM-56 and best suited methods of estimating evapotranspiration is found out. Kisi and Ozturk (2007) used the neurofuzzy and ANN model to estimate the FAO-56 PM ET0 using the observed climatic variables. Also investigated ability of fuzzy genetic (FG) approach in modelling of reference evapotranspiration and it is observed that FG approach is superior in modelling daily ETo than the other models. MATERIALS AND METHODS In this study the evapotranspiration is predicted for future period on daily basis scale from the available meteorological data using fuzzy inference system. Fuzzy logic base modelling of determination of evapotranspiration is a simple approach, which operates on an ‘if-then’ principle. The results obtained from fuzzy inference system were compared with evapotranspiration calculated using Penman Monteith FAO-56 method. The root-mean-square errors (RMSE), sum of square error (SSE), mean-absolute errors (MAE), and coefficient of determination statistics are used Prediction of evapotranspiration using Fuzzy logic PRANITA N. BALVE and JAYANTILAL N. PATEL Civil Engineering Department, Sardar Vallabhbhai National Institute of Technology, Surat Email: balvepranita21@rediffmail.com ABSTRACT In this paper, evapotranspiration prediction is done using Fuzzy Inference System (FIS) of Fuzzy Logic.For the prediction of evapotranspiration, mean temperature, relative humidity, wind speed and net radiation is taken as inputs to the fuzzy inference system. To check the efficiency of the FIS model, the results were compared with the FAO-56 Penman Monteith (FPM-56) method. FIS model has given the coefficient of determination (R 2 ) 0.979. Results indicated that, FIS model has better efficiency for prediction of evapotranspiration Key Words : Evapotranspiration, Penman-Monteith method, Fuzzy Inference System, Fuzzy If- Then Rules.