In model–based inversion, a geological model is compared to seismic data. Then, the result of comparison between real and modeled data is used to iteratively update the model to have a better match of seismic data. This method is intuitively very appealing since it avoids the direct inversion of the seismic data itself. On the other hand, it may be possible to come up with a model that matches the data very well, but it is incorrect. There are two ways in which constraints may be used.
One way is to consider the additional information as a soft constraint, meaning that the initial impedance is a separate piece of information which is added to the seismic trace with some weighting of two. This approach is called “stochastic”. The second method is to consider additional information as a hard constraint that sets absolute boundaries on how far the final answer may deviate from the initial model. This approach is called “hard constraint”. In the model-based inversion, average block size and the number of iterations are crucial. Using of average block size greater than seismic sample interval is recommended. The main cause of this idea is the fact that there is some noise in the data. If the number of iterations sets to large numbers, the results will be better. The problem here is computational time.
By investigation on Ab-Teymur 3-D seismic data, the following results were obtained: proper values for constraint limit, average block size and number of iterations are 35%, 8 ms and at least 8, respectively.  Two low impedance regions were detected that could possibly have the potential of hydrocarbon by using model-based inversion with the mentioned parameters for Sarvak formation.