Modeling Toolbox


The purpose of the STARS Consequence Assessment Modeling (CAM) Toolbox is to provide a collection of consequence assessment models that comply with the STARS so​ftware quality assurance (SQA) guidance document for safety-related and non-safety applications.  These STARS SQA guidelines are less rigorous than DOE requirements for safety software, but they are appropriate for modeling applications where the results are not used to formulate initial protective action recommendations but are instead used to estimate the complex dispersion patterns of pollutants, guide the deployment of field monitoring teams to optimal sampling locations, and provide other initial consequence estimates.  In these sorts of applications, decisions made based on modeling results cannot lead to an adverse impact on human health or safety (i.e., in-situ monitoring is conducted before any safety decisions are made).


The STARS SQA guidance incorporates the key elements found in the DOE guidance for safety software but does so using an appropriately graded approach that is readily implementable by DOE’s emergency management community and its software suppliers.  In particular, STARS’s SQA guidance focused on the following SQA work activities:  
  1. Software Project Management and Quality Planning 
  2. Software Risk Management 
  3. Software Configuration Management 
  4. Procurement and Supplier Management 
  5. Software Requirements Identification and Management 
  6. Software Design and Implementation 
  7. Software Safety 
  8. Verification and Validation 
  9. Problem Reporting and Corrective Action 
  10. Training of Personnel 
A graded approach SQA is intended to strike an acceptable balance between the need to model complex environmental and health processes, implement timely innovations, and achieve an appropriate level of SQA for software products that are not classified as safety software. 
Models in the STARS CAM Toolbox have been judged to be in conformance with STARS’s SQA guidance.  Any SQA gaps between STARS’s guidance and the modeling products SQA program have been identified by the model developer, reviewed by a committee from STARS’s Consequence Assessment Modeling Working Group (CAMWG), and judged to be within tolerances for the toolbox.  Any unresolved SQA gaps are judged to be comparatively minor in nature and in most cases work is underway to plan to eliminate the identified gap.   
Special allowances are made in the evaluation process for legacy models that have a long history of successful and verified use and whose SQA planning, development, and testing documentation may no longer be fully intact (owing to the disposal of old records).  However, even for legacy software, all future software modifications must conform to STARS’s SQA guidance. For legacy software, acceptance into the toolbox is heavy weighted on configuration management, software design and implementation information, verification and validation testing, problem reporting and corrective action, and training of personnel. 
For model users in the DOE complex, there are a number of benefits of using a consequence assessment model from the STARS CAM Toolbox.  Some of these advantages include: (1) the model’s presence in the toolbox indicate that certain SQA levels are met and therefore do not have to be independently assessed by the user; (2) STARS SQA gap analysis for the model identifies limitations, vulnerabilities, and strengths that may not be readily found in other code documentation; (3) the ability to communicate with model developers and share information with the broader modeling and SQA community is enhanced.   
A word of caution, just because a model is in the STARS CAM Toolbox does not mean that is the “right model” for the job.  Being in the toolbox assures you that the model meets or exceeds STARS’s SQA guidance; it does not evaluate the appropriateness of the model for your application.  Choosing a Toolbox model to tackle a problem the model is not technically designed to address is no better than using a model with significant coding errors.   First and foremost, identify the models that are designed to tackle the sorts of problems you want to address, are designed to operate in the environment (e.g., complex terrain, water-land boundaries) being studied, are compatible with the available environmental and geospatial data, operate over the spatial and temporal time scales of interest, etc.  Only then, should you determine if one or more of your candidate models are in the STARS toolbox and are available for selection. 
CURRENTLY UNDER CONSTRUCTION: The buttons in the left navigation menu of this web page provide quick links to: