Abstract: A method of computer generating a subsurface characteristic model which is substantially consistent with readings taken in a well with a well logging device. A set of actual measurement values which represent readings taken over a given depth interval of a well is derived. An initial trial model of the subsurface characteristic under investigation is then selected. Next, a set of simulated measurement values is generated by applying the approximate response characteristic of the logging device to the discrete trial model. The simulated measurement values are compared with the actual measurement values and the initial trial model is then modified in accordance with the comparision. The modified trial model is utilized to form a new set of simulated measurements which are, in turn, compared with the true measurement values. The model can then be remodified and the procedure continued in iterative fashion until there is obtained a model which yields acceptable simulated measurement values; i.e.