Total Electron Content (TEC) is an important parameter in monitoring ionospheric variability. In a given region, TEC can be obtained only by interpolation of measurements due to sparsity of the useful data. The lack of a complete ionospheric model hinders the choice of the optimum interpolation algorithm. A plausible solution to this problem can be found by investigating the performance of alternative interpolation algorithms on synthetically generated TEC surfaces for various sampling scenarios. The synthetic TEC data should represent the possible trends and variations of ionosphere. In this study, the performance of Random Field Priors (RFP) and Kriging interpolation algorithms are investigated over the parameter set of spatially correlated synthetic TEC data for various variance, range and trend options. Synthetic TEC data are sampled with regular and random sampling patterns, for number of samples from sparse to dense samplings. Interpolation scenarios are generated to investigate the improvement of the interpolation accuracy of the methods for each parameter. It is observed that for the random sampling patterns, when the trend is not modeled correctly, the errors of the algorithms increase and when the trend is modeled correctly, the reconstruction errors decrease. For the regular sampling patterns, the trend model does not affect the accuracy of the methods, and the reconstruction errors are close to lower bound error values. An example reconstruction is also provided over GPS-derived TEC, and error variances are compared over Kriging and Random Field Prior algorithms.