Date: April 13, 2008 - April 18, 2008
Venue: EGU General Assembly, 2008
Hoffman, R. and J. Henderson, 2008: Exigent forecasting of extreme weather. Preprints, 2008 EGU General Assembly, Vienna, Austria. 13-18 April 2008
Resource File: EGU2008-A-09760.pdf
A forecast of a very large specific impact on people or property or nature might result from a particular but plausible perturbation, i.e., one that is consistent with known uncertainty or errors in observations and the model background field. Examples of interest include forecasts of aircraft icing at a particular airport, of freezing conditions in an area devoted to citrus growing, of precipitation for a Formula 1 race track, etc. Henderson et al. (2005) call these “exigent” forecasts because of the requirement of precision in the forecast and urgency in the response. Exigent forecasts can be determined with modified forms of 4d-VAR similar to our damage cost function experiments. Exigent 4d-VAR may be considered to be a nonlinear form of singular vector analysis. In general the damage cost function for this purpose measures the costs minus benefits for a particular interest (i.e., customer) resulting from the forecast. It is critical that accurate forecast error statistics are used in the definition of the background term so that the increments determined would be consistent with our knowledge of the atmosphere and associated uncertainty.