Pacific Decadal Oscillation in SST Mantua et al

“Pacific Decadal Oscillation” in SST [Mantua et al. 1997] “warm” phase; AK salmon PNW salmon “cold” phase; AK salmon PNW salmon Updated time series available online at: http: //tao. atmos. washington. edu/pdo/

Updated North Pacific Index [Jim Hurrell, NCAR] Available online at: http: //www. cgd. ucar. edu/~jhurrell/np. html

Current Pacific SST anomaly map NOAA Climate Diagnostics Center Available online at http: //www. cdc. noaa. gov/map/images/sst/

ENSO/PDO/AO effects on US precipitation [Higgins et al. 2000] “ENSO” effect: Strongest along Gulf coast Add “PDO” effect: esp. important across SW? “AO” effect: Ohio River Valley

Current winter precipitation outlook NOAA Climate Prediction Center Available online at http: //www. cpc. ncep. noaa. gov/products/predictions/multi_season/

Time series of ENSO/PDO and winter precipitation in New Mexico Time series of equatorial SST (Niño 3) and PDO (Mantua et al. ) ending in late summer, 1951 -1997 [Gutzler et al. 2002] Time series of subsequent winter precipitation in Southwest NM (Climate Div 4 and town of Ft Bayard), DJFM 1951/52 – 1997/98

Quantification of seasonal predictability with a “modified” Brier Score Consider the distribution of 16 upper tercile Niño 3 years Tercile distribution of precip is 5 / 10 whereas random distribution would be 5. 33 / 5. 33 “Modified” 1 -tailed Brier Score S: Each annual “hit” = 0 1 -bin error = 1 2 -bin error = 2 (not 4) Sum annual scores, divide by N … so for 5/1/10, S=11/16=0. 69 Perfect skill: S=0. 0 No skill: S=1. 0 Informally, we find: S>0. 9 no utility 0. 75<S<0. 9 modest utility 0. 6<S<0. 75 marked skewness S<0. 6 forecaster’s dream

ENSO extrema 1951 -1997 provide significant predictability of SW US winter precip … [Gutzler et al. 2002] Predictability of a dry winter following negative JAS Niño 3 Perfect skill = 0. 00 Zero skill = 1. 0 or higher Skill darker shading Predictability of a wet winter following positive JAS Niño 3 (better for El Niño than La Niña)

… but PDO has a relatively modest effect on SW US winter precip predictability [Gutzler et al. 2002] Predictability of a dry winter following negative PDO year Perfect skill = 0. 00 Zero skill = 1. 0 or higher Skill darker shading Predictability of a wet winter following positive PDO year

Decadal modulation of ENSO predictability is the primary influence of PDO Pre-1977 [Gutzler et al. 2002] Dry winter following JAS La Niña (good predictability pre-77; none post-77) Wet winter following JAS El Niño (good predictability post-77; none pre-77) Post-1977

The “climate regime shift” of 1977 profoundly changed ENSO-based predictability … have we changed back to pre-1977 conditions? • Characterization of low-frequency variability is extremely difficult to get right the first time • At present there is neither a universally accepted index of PDO/NPO nor a satisfactory theory • Not all PDO indices show a shift circa 1998 • In mid-continent, including SW North America, the ocean is not the whole decadal climate story • Decadal climate variability bedevils monitoring of long-term change and prediction of short-term anomalies • Nevertheless decadal variability is profoundly important for ecosystems and human society