Ordinary kriging vs localised uniform conditioning

Linear Estimators, such as Ordinary Kriging, can produce smoothed assessments in the results of your resource estimation. The conventional Uniform Conditioning method estimates a tonnage and grade of mineralisation that can be recovered using the Selective Mining Unit (SMU) of size at the chosen cut–off value. A set of grade tonnage distributions is constructed for each panel under study by applying several cut-off values. A new method called Local Uniform Conditioning (LUC) enhances the Uniform Conditioning approach by localising the model results. The LUC algorithm then estimates the mean grade of the grade classes in each panel at the given SMU support level. The grade class is the portion of the panel whose grade is higher than a given cut-off, but lower than the next cut-off.

The next step is to rank the SMU blocks in each panel in their grade, distributed in increasing grade order. Finally, the mean grades of the grade class which have been deduced from the UC model are assigned to the SMU blocks whose rank matches the grade class. Thus, the key advantage of the LUC approach is the ability to calculate the mean grade of the grade class and assign these mean grades to the SMU size blocks, which have been ranked in each panel in increasing grade order (Abzalov 2006).

SRK used a Local Uniform Conditioning method to calculate an Iron deposit in the north of Chile. The results revealed there was an improvement in representing the grades pertaining to the distribution group close to the average. But, because this project had a high amount of information overall, final results are quite similar to the results obtained in the Ordinary Kriging approach.

SRK plans to continue using this technique in exploration projects with a smaller amount of information to reduce the impact of smoothing that linear estimation methods produce.

Ernesto Jaramillo: ejaramillo@srk.cl

Probability Plot, Samples VS Ordinary Kriging VS LUC

Ordnary Kriging

 


You can download a PDF of the entire
Mineral Resource Estimation newsletter

PDF
A4
PDF
Letter

 

 

 


Our newsletters focus on specific areas of interest to earth resource professionals and clients. Each is available as an Adobe Acrobat PDF file. If you don't already have Adobe's PDF reader, you can download it free.

 

 

SRK Turkey