Temporal dimension - IFs
Note: The documentation of IFs is 'under review' and is not yet 'published'!| Corresponding documentation | |
|---|---|
| Previous versions | |
| Model information | |
| Model link | |
| Institution | Frederick S. Pardee Center for International Futures, University of Denver (Pardee Center), Colorado, USA, https://pardee.du.edu/. |
| Solution concept | |
| Solution method | Dynamic recursive with annual time steps through 2100. |
| Anticipation | Myopic |
Future-Oriented Analysis. The base year of IFs in version 8.35 at the time of this writing is 2020 and the model runs recursively in annual time steps through alternative horizons up to 2100 with the option of running further for selected analyses. The interface also has a stop and start capability to facilitate scenario analysis. The base year is changed fairly often as new data releases allow.
Historical Analysis. IFs also includes an extensive and growing historical data base (nearly 7,000 series) starting in 1960, of which more than 10% are used to initialize model variables and the others support model development and use. The database allows initiation of model runs in 1960 and comparison of projections with empirical values.
The preprocessor of IFs and Base Year variable initializations. In the early days of world model/IAM development, rebasing the models for new first/base years or merely in the face of data updates for a fixed base year was often a very challenging enterprise. Such rebasing almost invariably involves changing some series to address unit and scaling issues, analyzing and making judgments about data for the same variables from different sources, and reconciling multiple series that flesh some part of the model (e.g., within the social accounting matrices of the economy) that are not inherently consistent from the same or multiple sources. Further, in a model like IFs with 188 countries, there are inevitably completely missing data series for some of them and only very old data for others. Complete absence requires hole-filling logic and old data requires algorithmic updating processes.
To greatly facilitate all of this, the IFs system has access within its interface to an integrated data “preprocessor” that automates these processes. In combination with the Pardee Institute’s constantly refined mechanisms and rules for putting large numbers of data series into its associated database (and regularly updating those with new data releases), the use of the preprocessor essentially automates both the updating of data in an ongoing base year and the rebasing of the model to a new base year. The interface of the IFs system (under the Main Menu option in Figure 4 labeled “Extended Features”) has a sub-option named “Rebuild Base” that does all of this in the IFs version on stand-alone computers (not possible for the web-based version). Further, a sub-option labeled “Rebuild Historical Base” allows the user to pick any base year back to l960 and rebase the model’s first year to that using the large historical database (inevitably with more holes or incompatible data in earlier years for the preprocessor to handle). Runs of the model from such earlier years allows direct comparison via the interface of historical data and values from model runs.