A New Improved Na-K Geothermometer By Artificial Neural Networks

Jump to: navigation, search


OpenEI Reference LibraryAdd to library

Journal Article: A New Improved Na-K Geothermometer By Artificial Neural Networks

Abstract
A new Na/K geothermometer equation has been developed. The temperature function is:Concentrations are in mg/kg. The new improved geothermometer equation was developed by artificial neural networks. The normalized mean square error (NMSE) used in the new improved Na/K equation for temperatures ranging from 94 to 345°C is 0.179, which is lower than the corresponding NMSE 0.226, 0.598, 0.656, 0.268, 0.328 and 0.225 for the equations of Arnorsson et al. (1983; Geochim. Cosmochim. Acta 47, 567-577), Truesdell (1975; Proc. 2nd UN Symposium), Tonani (1980; Proc. Adv. Eur. Geoth. Research, 2nd Symposium), Fournier (1979a; J. Volcanol. Geotherm. Res. 5, 1-16), Nieva and Nieva (1987; Heat Recovery Systems and CHP 7, 243-258) and Giggenbach (1988; Geochim. Cosmochim. Acta 52, 2749-2765), respectively. It is therefore strongly recommended to use this new improved Na/K geothermometer equation for geothermal exploration. Since the proposed Na/K geothermometer was developed on a purely empirical basis, it should not be applied outside the T-P-X range of calibration.

Author 
Ibrahim Can








Published Journal 
Geothermics, 2002





DOI 
10.1016/S0375-6505(02)00044-5


 

Citation

Ibrahim Can. 2002. A New Improved Na-K Geothermometer By Artificial Neural Networks. Geothermics. (!) .