Abstract
ABSTRACT This paper investigates the problem concerning the analysis of trend surface of spatially correlated data, we have got the model of trend surface based on an assumed covariance function. This function contains parameters called variance components, these parameters are estimated by maximum likelihood estimation that depends on Newten-Raphson algorithms to obtain the convergent solutions. The applied representation includes the analysis of real spatial data which represent the level of ground water in Al-Qaaim west of Iraq, the results is very encouraging. All algorithms are programmed by MATLAB.
