Thursday, 26 January 2017: 2:15 PM
605 (Washington State Convention Center )
Yangyang Xu, Texas A&M University, College Station, TX; and A. Hu
Decadal climate variability originating from the Pacific Ocean have been defined by two EOF (Empirical Orthogonal Function)-based indices (Pacific Decadal Oscillation, PDO; and Interdecadal Pacific Oscillation, IPO). Although the two indices have different regional focus and have different filtering methods to remove shorter-term variation, they are highly correlated in time. PDO/IPO have been shown to have a major impact on regional hydroclimate over North America; and presumably as metrics for internal variability, they also play a dominant role in governing the pace of externally forced global warming. However, it is less clear how the background global warming might, in turn, impact the variability and regional impact of PDO/IPO.
One obstacle in revealing this connection is that the conventional calculation of PDO/IPO (either with no detrending or global detrending applied to the temperature dataset) failed to account for the regionality of background climate change, notably in North Pacific. Therefore, the resultant PDO/IPO are contaminated by the externally forced trend, in particular in the 21st-century climate simulations. After this local trend is removed (by subtracting the ensemble mean trend, or simply the linear fit of the local trend) before the EOF analysis, we show that revised PDO/IPO are purely an indicator of internal climate variability for a 1920-2100 large-ensemble climate simulation with Community Earth System Model. We find that PDO/IPO tend to have a lower frequency in a warmer climate, and its impact on global and regional climate are weakened, probably due to a shallower ocean mixed layer.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner