Premium
Measuring the potential utility of seasonal climate predictions
Author(s) -
Tippett Michael K.,
Kleeman Richard,
Tang Youmin
Publication year - 2004
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1029/2004gl021575
Subject(s) - climatology , environmental science , precipitation , gcm transcription factors , seasonality , ensemble average , sea surface temperature , general circulation model , climate model , climate change , atmospheric sciences , meteorology , geology , statistics , mathematics , geography , oceanography
Variation of sea surface temperature (SST) on seasonal‐to‐interannual time‐scales leads to changes in seasonal weather statistics and seasonal climate anomalies. Relative entropy, an information theory measure of utility, is used to quantify the impact of SST variations on seasonal precipitation compared to natural variability. An ensemble of general circulation model (GCM) simulations is used to estimate this quantity in three regions where tropical SST has a large impact on precipitation: South Florida, the Nordeste of Brazil and Kenya. We find the yearly variation of relative entropy is strongly correlated with shifts in ensemble mean precipitation and weakly correlated with ensemble variance. Relative entropy is also found to be related to measures of the ability of the GCM to reproduce observations.