USGS Home
Contact USGS
Search USGS

Evaluation of and Training for Assessment Methodology/Software

Task Objectives

Historically, the MARK3 and EMINER software produced by the USGS has been used to perform quantitative mineral resource assessments. However, this software requires either updating or complete re-writing to meet the needs of the Updated National Mineral Resource Assessment. This task will evaluate the most cost-effective and time-efficient method to produce the new software, and then develop that software in advance of implementation of the assessment.

The primary objectives of this task are to:

1. Evaluate whether an Excel-based or Fortran/C language software best meets the needs of the Updated National Mineral Resource Assessment.
2. Develop software requirements agreed to by Project management that will meet the needs of the Updated National Mineral Resource Assessment.
3. Code and test the software, making it available prior to the start of the Updated National Mineral Resource Assessment.
4. Make recommendations for, and if approved, design training for use of this software. To be successfully applied to the Updated National Mineral Resource Assessment, the assessment software must be employed knowledgeably and inspire confidence/a sense of reliability among by those conducting the assessment.

Statement of Work

In 2008, a new software development and application effort will begin. The existing USGS quantitative mineral resource assessment software is dated and commercial software not adequate or cost-effective for present application. The new software will require development of modules for numerical computation and graphical display on a personal computer (PC) platform. Unfortunately, the development of these modules is not easily done using existing USGS code; therefore, alternatives will be explored prior to final selection and development of software for use in the Updated National Mineral Resource Assessment

Objectives

< li>Evaluate the advantages and limitations of using the Microsoft Excel spreadsheet and commercial Crystal Ball add-on to estimate the probable number of tract deposits for a well-known strategic mineral.
1. Evaluate the advantages and limitations of using Argus or other graphical toolboxes, VBA coding, or other approaches for development of a graphical user-interface.

Methodology

1. Crystal Ball add-on to Microsoft Excel
   
   In Fiscal Year 2008, the Crystal Ball add-on to Microsoft Excel will be obtained and evaluated. To determine whether the software will perform as expected, we will evaluate the overall effect of subjective model input and correlated random variables on prediction uncertainty in the estimation of tract deposits (forecast). The general approach will involve Monte Carlo sampling of probability distributions assigned to the consensus input values, grade and tonnage cumulative distribution functions, and (if time permits) regression parameters in the economic filter. The effects of uncertainty in the consensus input variables will be compared to past assessments in which these values were considered deterministic.
   
   In addition to evaluating the effects of uncertain consensus input variables, at least three separate forecast distributions will be computed based on the assumption of correlation among grade and tonnage distributions: no correlation (correlation coefficient of 0.0), intermediate correlation (0.50), and perfect correlation (correlation coefficient of 1.0). In all cases, the actual number of Monte Carlo samples used to compute the deposit forecast will be based on a bootstrap (or similar) approach to ensure repeatability of simulations within a 1% margin of error.
2. EMINERS
   
   The most recent version of EMINERS Fortran code will be obtained, compiled, and reviewed.
3. Graphical User Interface
   
   Existing freeware and alternative coding/scripting, such as VBA, will be evaluated for application to creating a graphical user-interface and generation of random variables for use in Excel and/or the recompiled EMINERS.

Deliverables during FY08 are:

1. a decision on whether software based on Crystal Ball and Excel can meet the requirements for routine use within a National Assessment.
2. a plan for writing or modifying code in a standard programming environment (Fortrran, C, Java) as the second choice for producing software prior to the start of the Updated National Mineral Resource Assessment.

Draft requirements for the software are to:

1. Allow users to assume one of several possible frequency distributions for number of deposits.
2. Allow users to select a distribution for tonnages and grades to reproduce the distributions of contained metal in the original grade and tonnage models.
3. Allow users to include cost estimates so that an economic analysis can be performed. The user can easily modify the coefficients of the economic filters.
4. Allow users to easily add new models.
5. Have an easy to use interface.
6. Run simulations that aggregate the undiscovered deposit estimates for any combination of tracts/deposit types.
7. Aggregate the economic results for any combination of tracts/deposit types.

References to non-USGS products and services are provided for information only and do not constitute endorsement or warranty, express or implied, by the USGS, USDOI, or U.S. Government, as to their suitability, content, usefulness, functions, completeness, or accuracy.

Return to top of this page | Previous Task | Next Task | Task List