All-season regressions of standardized smoothed AMO indices on SSTs for the winter 1900–fall 1999 period. Regressions for the models are calculated for each ensemble member separately and then an average is computed for each model. Red/blue shading denotes positive/negative SST anomalies; contour interval is 0.1 K. The indices are constructed by first calculating a spatial average of SST anomalies over the (5°–75°W, 0°–60°N) region and then detrended, using the least squares method. The indices are finally smoothed by applying a 1-2-1 binomial filter 50 times and normalized by using their standard deviation. Regressions are shown after 5 applications of smth9 in the GRADS plotting software. Bottom Panel Observed HadISST smoothed AMO index and other four model-derived smoothed AMO indices which have the highest correlations, R, with the observed index: GFDL-CM3, Ensemble 5 (R = 0.75), UKMO-HADCM3, Ensemble 4 (R = 0.56), ECHAM6/MPI-ESM-LR, Ensemble 3 (R = 0.01) and CCSM4 Ensemble 4 (R = 0.29). The correlation range for the different ensembles within each model is shown adjacent to the model’s name
ABSTRACT: This study aims to characterize the spatiotemporal features of the low frequency Atlantic Multidecadal Oscillation (AMO), its oceanic and atmospheric footprint and its associated hydroclimate impact. To accomplish this, we compare and evaluate the representation of AMO-related features both in observations and in historical simulations of the twentieth century climate from models participating in the IPCC’s CMIP5 project. Climate models from international leading research institutions are chosen: CCSM4, GFDL-CM3, UKMO-HadCM3 and ECHAM6/MPI-ESM-LR. Each model employed includes at least three and as many as nine ensemble members. Our analysis suggests that the four models underestimate the characteristic period of the AMO, as well as its temporal variability; this is associated with an underestimation/overestimation of spectral peaks in the 70–80 year/10–20 year range. The four models manifest the mid-latitude focus of the AMO-related SST anomalies, as well as certain features of its subsurface heat content signal. However, they are limited when it comes to simulating some of the key oceanic and atmospheric footprints of the phenomenon, such as its signature on subsurface salinity, oceanic heat content and geopotential height anomalies. Thus, it is not surprising that the models are unable to capture the majority of the associated hydroclimate impact on the neighboring continents, including underestimation of the surface warming that is linked to the positive phase of the AMO and is critical for the models to be trusted on projections of future climate and decadal predictions.
AMO’s structure and climate footprint in observations and IPCC AR5 climate simulations
Argyro Kavvada, Alfredo Ruiz-Barradas and Sumant Nigam
Climate Dynamics
Observational, Theoretical and Computational Research on the Climate System
10.1007/s00382-013-1712-1
Argyro Kavvada
Email: argyrok@atmos.umd.edu
Received: 4 June 2012
Accepted: 20 February 2013
Published online: 7 March 2013