Assessing, Modifying and Synthesizing a Suitable Model for Estimation of Potential Evapotranspiration in Iran

Introduction
Iran is a dry land with very low precipitation. Annual rainfall is less than a third of the average rainfall worldwide. One of the ways of adjusting to drought in Iran, especially with agriculture is the optimal and sustainable use of water resources. Precipitation, surface water and ground water resources have to be used as efficiently as possible. This study would not be practical without first taking into consideration the exact requirements of water for agricultural fields in Iran. Knowledge on the evapotranspiration is very important in irrigation and drainage planning. Evapotranspiration is one of the key components of the hydrologic cycle and its calculation is important for a number of applications such as, the hydrologic balance of water, design and management of irrigation systems, simulation of the amount of products and design, and management of water resources.

Methodology
The key purpose of this research is to assess, modify and localise a potentially suitable evaporation and transpiration model which can be implemented for Iran. The research methodology focuses on three main parts:
1- Country clustering based on the climatological effect of evapotranspiration.
2- Testing the results of evapotranspiration using proposed relationships.
3- Modifying and localizing current evaporation and transpiration models with observed data.
This research takes into account eight components, namely; average temperature difference, minimum average maximum relative humidity, the frequency of occurrence of average wind speeds above 5 knots per sec, the amounts of monthly rainfall, days with rainfall above 10 mm and 5mm, a 25-year record (1980 to 2005) of 64 synoptical and climatological stations in Iran. The purpose of selecting these components has been clustering the studied stations based on effective parameters of evaporation and transpiration in order to after this stage exert for each cluster equal modifying coefficients based on similarity of the stations in evapotranspiration process.

Results and Discussion
In this study, the findings suggest clustering the country into six main parts. Each cluster is based on its geographical and climatic characteristics. The first cluster is the arid to semi-arid regions of the central and south-east. In the first cluster the minimum evapotranspiration is 51 mm for January and the maximum 322 mm for July. The annual average evapotranspiration for the first cluster is 2119 mm. The second cluster is wet and mountainous regions of the north and west. In the second cluster, the minimum evapotranspiration is 21 mm for January and the maximum, 272 mm for July. The annual average evapotranspiration for the second cluster is 1584 mm. The hot region of the southern coasts is the third cluster and in this cluster the minimum evapotranspiration is 118 mm for January and the maximum evapotranspiration 265 mm for June. The annual average evapotranspiration (2291 mm) for the third cluster was found to be the highest. The fourth cluster is cold and dry regions of the north-east to cold and wet regions of the north-west. The minimum evapotranspiration for the fourth cluster is 14 mm for January with a maximum of 283 mm for July. The annual average evapotranspiration for the fourth cluster is 1596 mm ranking in third place amongst the other clusters. The fifth and sixth cluster is a combination of the wet to very wet regions of the north coasts. In the fifth and sixth cluster the minimum and maximum evapotranspiration is 28mm (for January) and 161 mm (for July) respectively. The fifth and sixth cluster has the lowest annual average evapotranspiration (1012 mm) of all the clusters. In future evapotranspiration amounts for base stations can be calculated using proposed relations and the result can then be compared with the results of other formulas for convection such as the Pearson’s formula. Overall, the result that will be extracted from the total average is calculated using Blaney Cridle's method which ranks first here with an average of 0.69. The methods of Jensen-Haise, Thornth-Waite, and Hargrives-Samani rank second to fourth respectively here. After selecting Blaney Cridle s model as the most suitable, this model was then calibrated using empirical evaporation data from lysimeter.

Conclusion
Results from clustering Iran based on the evapotranspiration component shows that the annual maximum of this component is allocated to the third cluster or hot regions of the southern coasts. This maximum potential evapotranspiration is likely due to the low latitudinal location of the stations and the near vertical radiation of the sun’s rays in these regions. In contrast, cluster 5 and 6 (wet to very wet regions of the north coasts) have jointly the lowest annual average evapotranspiration amount. This low evapotranspiration amount is likely due to the lower elevation (sea level) of cluster 5 and 6. In addition to the location of cluster 5 and 6 at higher latitudes may account for the cluster having the lowest potential evapotranspiration when compared with the other clusters. The result obtained from the correlation between the output values of Blaney Cradles’ index with the values of components of this index (mean daily temperature, minimum relative humidity, mean wind speed, sunshine hours) reveal that minimum relative humidity has the most effect in comparison with the other climatic factors in the amount of potential evapotranspiration for clusters 3, 5 and 6. In clusters 1, 2 and 4, this component was temperature showing more correlation with the values of potential evapotranspiration. It is recommended utilizing normalized data in future to obtain more appropriate correlations between climatic parameters with evapotranspiration.