IRI Focus
Use the options Constraints > Add Constraint > IRI Focus or IRI Focus > Modify to apply IRI Focusing.
Modify IRI Focus Constraint dialog options
Constraint type |
The default is to apply IRI focus in order to sharpen the modelled property. IRI focus runs the weighted inversion in a loop, and feeds in the fitted voxel model of one inversion as the reweighting input voxel model for the next run. When this option is chosen, you are prompted for the number of iterations. You can fore-go IRI by simply choosing the None option. |
Factor |
When the option IRI focus is selected, you are prompted for the number of sharpening rounds. The process will then automatically assign the normalized fitted voxel model of the previous inversion as the Iterative Reweighting voxel model for the current inversion, Factor times. The default is 2. |
|
If checked, the interim IRI results are downloaded along with the modelled voxel. Expand the Solution tree to get access to these voxels. Each voxel is appended by its IRI factor, and can be viewed by checking the box preceding it. All interim results are also saved in the predicted database and appended by the corresponding IRI factor. By default, the interim results are not downloaded. |
Define the property distribution end to sharpen |
Positive is the default and sharpens the large positive value end of the property distribution. It is recommended for targets of interest with a strong positive property, such as VMS deposits. Selecting Negative will sharpen the large negative value end of the property distribution. This is appropriate in cases where the volume of interest has a lower property than the host rock, such as salt domes or kimberlites. Selecting Both will sharpen both ends of the property distribution while putting less emphasis around the mean. This option can be used when expecting a pronounced property variability inside the volume of interest such as with some porphyry deposits. The default for MVI models is to enhance the Positive end. For gravity and magnetics, Both ends are enhanced. |
Application Notes
IRI Focus is used to refine the inversion and emphasize a desirable model characteristic. The IRI Focus constraint normalizes and sets the outcome of the previous inversion as the weight constraint for the next inversion. You must specify how many times to run through this process.
If you suspect that the target of interest has a higher property than the surrounding rocks, like in the case of VMS deposits, you may opt to put more emphasize in the positive end of the property distribution and less emphasis on the surrounding rock. Alternatively if you are attempting to delineate a salt dome or a kimberlite both of which typically manifest a lower property than the surrounding rock, you may opt to put more emphasize in the negative end of the property distribution and less emphasis on the surrounding rock. Finally, some porphyry deposits characteristically show a pronounced variability in their property distribution, and selecting to emphasize both end of the property distribution at the expense of the mean will help better delineate the structure of interest.
It is recommended to remove the background from the observed data prior to modelling; the resulting property voxel model will consist of positive and negative values. On this basis, we assume that the mean of the data distribution lies around zero, and accordingly the IRI Focus constraint is set to emphasize the positive and/or negative ends of the property spectrum.
Effectively, each iteration is another inversion, but the looping is done internally to the inversion. As a result, subsequent inversion iterations do not carry the overhead of one full inversion.
The number of iterations has been limited to four.
Access Seequent Online Learning and select the VOXI guided paths to learn more about effective workflows and key concepts.
Got a question? Visit the Seequent forums or Seequent support
© 2023 Seequent, The Bentley Subsurface Company
Privacy | Terms of Use