Sensitivity of regional climate change predictions to SST anomaly patterns for present-day and future scenarios
Abstract
Regional climate simulations from general circulation models (GCM) are subject to uncertainties from internal variability, external forcing and the model differences. Because changes in tropical SST patterns are a primary driver of the regional climate change over both tropical and extratropical regions, we first examine the sensitivity of the regional climate anomaly to the tropical SST anomaly from an ensemble of AGCMs as a first step identify sources of uncertainties. We use a random perturbation method to investigate the linear regression between the perturbed SST and the climate anomaly of a large set of continental regions across the world. As a metric to identify model response uncertainty, we use the global teleconnection operator (GTO), which is the linear operator relating the regional climate response to SST anomaly patterns. We use climatological SST patterns obtained from three sources: observed HadISST data and output from the CMIP3 archive for both the C20C and future A1B scenarios. Both the GTO patterns and the linear reconstruction of the regional climate anomaly for these three cases will be investigated. The linear reconstruction of the regional climate anomaly will be based on the GTO from the three cases. The reconstructions using the GTO(C20C) and GTO(Obs) will be compared with the ensemble model response to HadISST(t) and the results will be used to interpret the GTO(A1B) reconstruction. We define the predictability of the regional climate response as the covariance of the reconstructed regional response to SST anomalies based on each GTO pattern with the response estimated from ensemble runs forced by the observed SST fields. For different scenarios, we will estimate the covariance for a large set of continental regions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFMGC11A0896L
- Keywords:
-
- 1626 GLOBAL CHANGE / Global climate models;
- 1637 GLOBAL CHANGE / Regional climate change