whatIf? Wiki

Decision Support - Simulation models are primarily learning devices that extend our powers of perception; they do not predict what will happen nor do they prescribe what should happen.

Knowledge Sharing - Models and their results present an explicit and communicable understanding around which consensus may be built. They also provide a platfrom through which corporate memory may be transferred or preserved.

Transparency - Both the model design structure and input data are accessible so that the model user knows exactly how outputs are derived. Furthermore, the model design is comprised of processes that correspond to real world concepts and experiences so that the model logic is more easily understood.

Scenario Management - A scenario consists of the set of input variable values and the corresponding output values generated from those inputs. In models with hundreds or thousands of variables, it is important to display scenarios so that the user can easily access the set of input and/or output values in a given scenario and compare two scenarios. Saving scenarios allows output variables to be recreated because the pointers to the set of input values are maintained and retrievable.

Data Visualization - Visual interaction is essential for comprehensive interpretation of large quantities of multi-dimensional data.

Collaboration - The whatIf? modeling process relies collaboration among the model developers and the stakeholders to identify the model scope, create the model design, identify data sources, and establish the critical uncertainties to be explored. Such collaboration is supported by the high level hierarchical diagrams in Documenter, which clearly illustrate model organization and underlying processes, as well as the client server architecture of the whatIf? modeling platform.

Physical Accounting - State variables are exactly determined by some combination of initial conditions, exogenous flows, the values of the parameters reflecting the laws of motion of the system, and the values of time varying control variables.

Coherency - Scenarios must assure consistency among the input variables both over time and within time periods. For example, in an energy model, population and the services required by the population determine the level of economic activity. The level of economic activity in turn creates demands on the energy system. The way in which the energy system meets energy demands controls the emission of greenhouse gases and criteria air contaminants. The model assures coherency through the use of stock-flow accounting rules, vintaged stocks and life tables, supply/disposition balances for fuels and feedstocks, and the explicit representation of energy transformations.

Bottom-up Approach

Holistic Scope

Adaptability - new understanding and new data must be incorporated as they become available

Realism - the simulator must capture the essence the real world without oversimplification

Controllability - the user must be able to experiment with control settings that make “the airplane fly safely, or crash”