|
Forward modelling
In the past, an estimate of the "likely" property distribution, typically a simple "layercake" ground-model, would be used to generate a synthetic dataset that would be compared to the measurement dataset acquired in the field. This property distribution would then be varied manually, until the synthetic and field measurement datasets agreed. This "forward modelling" approach, which can be laborious, requires knowledge of the number of layers present at a site and may be tractable for simple geology only (eg 1-D) (See Chapter 4).
Inverse modelling
Forward modelling has been largely superseded by automatic numerical inversion processing, that "inverts" measurements directly into a spatial property distribution (1-D, 2-D and 3-D) without manual intervention. Inversion is at the heart of modern geophysical data processing and interpretation. It is a significant advance on forward modelling methods, because it enables the spatial distribution of geophysical properties to be displayed "tomographically" as images (eg
...
This 250-word extract was created in the absence of an abstract.