The Proton Magnetic Resonance (PMR) method is one of the newest technologies used in the field of geophysics for assessing the hydraulic features of undersurface aquifers and reservoirs in recent decades. The advantage of this method is direct identification of the free molecules inside holes. In addition to the percentage of water content of formations, two important hydrodynamic parameters of undersurface formations, i.e. porosity and permeability can be estimated by this method. The shortage of tap water in the country increases the significance of this method. Especially because one of the main resources of tap water supply are aquifers. In spite of high cost of excavation and accessing the aquifers, assessing the water content and hydraulic features of the land are very important. Generally, in order to obtain the optimized undersurface model from PMR data, knowing geometrical and physical structure of distribution of Vertical Electrical Sounding (VES), substantially helps the more detailed inversion of PMR data. The area under study in this research is Hessa site located in the northeast of Shahin-Shahr town and Esfahan-Borkhar hydrologic unit. This unit is a part of Zayanderud drainage basin. Purpose of this research is using the Simplex downhill optimization algorithm for inversion of PMR data, in order to assess the water content of formations in the area under study. In this study we first calculate the kernel matrix of forward modeling of PMR data using RES (Resistivity Electrical Sounding) in the area, and the data obtained from magnetic features of the land, and for validation of the resulting kernel matrix, we compared it several artificial models. After we ensure the validity of the matrix, second phase of the research, containing target function definition for assessing water content of the land, was carried out. The target function which is measured based on RMS Error of PMR data is designed with corresponding theoretical response from the forwarding model. This function was used for entering the simplex downhill algorithm to find water content of each layer. Simplex downhill is a multivariable optimization method which doesn’t need the target function derivatives calculations. It has a relatively simple structure and speed for obtaining the optimal point. Simplex downhill is a global optimizer that reduces the risk of finding local optimums.
