High spatial and temporal variability of the tropospheric wet refractivity index, makes it difficult to present an accurate model for this variable. Up to now, Radiosonde stations data have been used for monitoring atmosphere parameters. Furthermore, because of the sparse distribution of radiosonde stations to monitor the lower levels of the atmosphere, the numerical weather models do not have enough accuracies for atmospheric parameters. Using the GPS tropospheric delay measurements and tomography approaches, the wet refractivity index can be estimated. In this research, three-dimensional and four-dimensional basis-function tomography is used to demonstrate the distribution of wet refractivity index of the troposphere. In this model, spherical cap harmonics are used for the horizontal distribution of the wet refractivity index, and empirical orthogonal functions are used for the vertical distribution of the index. In addition, temporal changes are considered by correlating the unknown coefficients using fourier series. The region of study is in the west California State, and the wet refractivity index is retrieved from the wet tropospheric delay measurements. To validate the results, radiosonde profiles were compared to the tomographically retrieved profiles. The results show that wet refractivity indices can be retrieved using functional models with RMSE about 2. 4 ppm till 3. 9 in the four-dimensional method. The comparisons show that the four-dimensional retrieved profiles show improvement up to 34 and 42 percentages in mid-day tomography epochs compared to the three-dimensional tomography results. Also it can be seen that in mid-night epochs, the three-dimensional tomography has higher accuracy compared to four-dimensional method because of low variation of wet refractivity indices.