SEI engages in many different kinds of scenario exercises, including scenario modeling. A common problem arises when using empirical scenario models, because the base-year data do not exactly match the model. This paper shows that the gaps between observed and modeled values (the residuals, or shift factors) in the base year can be provide estimates of the residuals in the scenario that improve the performance of the empirical model. The paper proposes a self-consistent estimator that is shown to be effective by applying it to specific examples.
Link to journal article (external link – subscription required to read fulltext)
Design and development by Soapbox.