In traditional hydrographic surveys, chart datum is usually determined using spectral analysis of coastal tide gauge observations based on the lowest water level. Due to the variation of tidal amplitude and phase components in different locations, such chart datum is only valid in coastal areas around the tide gauge stations it is then hardly accurate when a few tens of kilometers away from the tide gauge stations in the off-shore areas. This study models the separations between the elliptical reference WGS84 and the chart datum using the data collected by the satellites Topex/Poseidon and Jason-1 for the periods of 1992-2002 and 2002-2008 in Persian Gulf and the coastal tide gauge data of the Oman Sea. The major advantages of the technique used in this study are highlighted as follows. The satellite altimetry observations on the shores and also in the shallow water is not of good accuracy and precision. An appropriate solution to this problem is to use coastal tide gauge observations near the time series generated from the satellite observations. On the other hand, chart datum determination based on the coastal tide gauge observations is suitable only for the areas close to the coastal tide gauge stations. However, due to the amplitude and phase variations of the tidal components in different parts of the sea, the accuracy of such a chart datum is not appropriate for the areas that are far away from the tide gauge stations. In this study, the satellite altimetry observations are combined with the data obtained from the coastal tide gauge stations. Due to the existence of point-wise time series in the study area and also creation of the “quasi tide gauge” points using the satellite altimetry observations, both of the problems mentioned above (i.e. weakness of the satellite altimetry observations and the chart datum determination based on the use of the coastal tide gauge observations only) can be solved. This will then lead to higher accuracy of the chart datum determination in the entire area. To achieve higher accuracy, observations of the tide gauge stations are also used. It is because the distances between the tide gauge stations and the time series obtained from the satellite observations are considerably large (sometimes more than 20 km). Therefore, this combination will lead to the establishment of a regular and continuous network of tide gauge observations in the entire area having acceptable accuracy. To determine this model, sea level variations due to the tide, polar motion variations, plates tectonic movements and all the factors affecting the potential tide with significant components (M2, S2, K1, O1) and even less important components such as signals with the period of 14 days, monthly, semi-annual, annual, 8.5 year and 18.6 year are considered in the case study area. It is also highlighted that because the satellite altimetry observations are only available on periods of 9.915 days, the high-frequency tidal signals cannot be detected using these data. Therefore, these frequencies are also included into the functional part of the model. Based on the above strategy, the separation between the elliptical reference and the chart datum has been computed by comparing the tide gauge data and the satellite altimetry data for the period of 2002-2005.