Abstract

Inverse problems are solved in geophysics and environmental sciences, basically in two different situations: either imaging or calibration of predictive models. Roughly speaking, “imaging” includes techniques to map the heterogeneity of the physical fields of interest and in essence it is a “diagnostic” tool. However, in many fields of geophysics and of environmental sciences, it is also necessary to forecast the future behavior of natural systems with predictive models, i.e., with “prognostic” tools. In that case, inverse problems are aimed to calibrate models. In this presentation, a general framework for discrete inverse problems is reviewed and discussed in order to emphasize the similarities and the differences between the two situations for which inverse problems are solved, namely imaging and model calibration. In particular, specific attention is drawn to the relevance of the ultimate goals of the solution of inverse problems for imaging and predictive modeling and to the role of field data and their processing.