College of Geo-Exploration Science and Technology, Jilin University, Changchun 130026, China

The traditional gravity modelling methods fail to deal with ambiguity zones. These zones are likely to be associated with uncertain boundaries. Source boundaries can be divided into "hard" and "soft" types. "Hard" boundary stands for distinguished margin of discontinuity of source distributions. Traditional modelling method is based on this type of boundary. Uncertain boundaries of ambiguity zones belong to "soft" boundary. The authors put forward the uncertainty analysis because lots of uncertainty factors need to be taken into consideration. Any powerful system designed to interpret gravity data for a real source distribution must be able to deal with uncertain information. Uncertainty analysis (UA) is a useful tool to estimate the ambiguity of the data interpretation. It is often applied in conjunction with a specific modelling or inversion method. In this study, we use an iterative inversion and its results are evaluated using uncertainty analysis and proved to be effective.