A forecast based on a systematic statistical
examination of data representing past observed behavior of the system to be forecast, including observations of useful predictors outside the system.
In short-term climate
forecasting, CCA (canonical correlation
analysis), as described by Barnston (1994), is a good example of a statistical forecast. Depending on method and scope, the limitations of statistical forecasts are related to shortness of record, danger of overfitting, assumptions of linearity
(often), absence (often) of physical considerations, etc. Purely statistical forecasts in weather forecasting have become rare; however, a combination of dynamical model
output and statistics
is very common in weather forecasting. Some statistical methods are guided by physical principles to such an extent that they resemble dynamical models. An example of the latter is empirical wave propagation
described by Qin and van den Dool (1996). See perfect prognosis method
Barnston, A. 1994. Linear statistical short-term climate predictive skill in the Northern Hemisphere. J. Climate. 7. 1513–1564.
Qin, J., and H. M. van den Dool 1996. Simple extensions of an NWP model. Mon. Wea. Rev.. 124. 277–287.
Copyright 2022 American Meteorological Society (AMS). For permission to reuse any portion of this work, please contact [email protected]. Any use of material in this work that is determined to be “fair use” under Section 107 of the U.S. Copyright Act (17 U.S. Code § 107) or that satisfies the conditions specified in Section 108 of the U.S.Copyright Act (17 USC § 108) does not require AMS’s permission. Republication, systematic reproduction, posting in electronic form, such as on a website or in a searchable database, or other uses of this material, except as exempted by the above statement, require written permission or a license from AMS. Additional details are provided in the AMS Copyright Policy statement.