Mechanisms of the North Atlantic multidecadal internal variability in the CNRM-CM 5 model Y. Ruprich-Robert (ruprich@cerfacs. fr) and C. Cassou (cassou@cerfacs. fr) Introduction: At multidecadal timescale, the North Atlantic basin is characterized by sea surface temperature (SST) variability (Schlesinger and Ramankutty, 1994) that could be reconstructed with tree-ring back to 1600 (Gray et al. 2004). This observed variability ~ is defined by the alternation of warming and cooling phases of the entire basin. It is linked to hurricane low frequency activity (Goldenberg et al. , 2001) and precipitation anomalies over the surrounding continents, in particular over the Sahel and the Nordeste (Folland et al. , 2001). Many studies based on climate coupled models conclude that this variability is partially due to internal climate system processes (e. g. Knight 2009) and is often referred to as the Atlantic Multidecadal Variability (AMV, Kerr 2000). The mechanisms at the origin of the AMV are still unclear (model dependence, scarcity of the observation…). In the present study, the AMV is investigated using the CNRM-CM 5 model. Model: CNRM-CM 5 3) Salt advection to Labrador Sea and deep convection contours : surface salinity 0. 1 psu. std-1 2) Heat oceanic advection to GIN Sea and sea ice melting 550 750 The CNRM-CM 5 AMV is a basin scale sea surface temperature variability, comparable to the observed AMV Fig. a : CNRM-CM 5 Low-pass filtered (1/25 yr-1 cut-off frequency) standardized AMV Index (colours), and AMOC index : principal component of the low-pass filtered AMOC leading EOF (black curve) Fig. b : Sea Surface Temperature regressed on the AMV Index (only statistical significant values at the 95% level are shown, significativity from Davison et al, 1997) AMOC : Atlantic Meridionnal Overturning Circulation Barotropic streamfunction climatology : 5 Sv between iso-contours ~Lag -13 ~Lag -8 P/ SC 1 EA 6) Damping of the anomalies in GIN Sea by the NAOLag -13 Lag -8 K. std-1 EAP (15%) ~Lag -13 ~Lag -8 <T>0 -200 m regressed on AMOC 4 5 ~Lag -35 Pa. std-1 Sea level pressure regressed on AMOC (lag -8) – (lag -13) Bold black line represents statistical significant values at the 95% (lag -8) 7 1) EAP and SCAND lead AMOC by 35 years ~ b) mm. d-1 6 AN D 2 3 lag -35 lag -28 Sea Ice regressed on AMOC at lag -28 only anomalies < -1% shown GIN Sea : Greenland - Iceland - Norwegian Sea <T>0 -200 m regressed on AMOC 7) Fresher water advection from the tropics Lag -8 Lag +8 NAO EAP SCAND d) c) 950 Summer precipitation regressed on AMOC Mechanisms of the CNRM-CM 5 AMOC variability ~Lag -35 ~Lag -28 correlation 350 ~Lag -18 Cross-correlation between AMOC and SPG Intensity (black), Florida-Bahamas mass transport (blue) and North Brazil current (red) SPG : Sub. Polar Gyre WBC : Western Boundary Current contours : < T >0 -200 m 0. 1°C. std-1 SCAND (10%) 150 lag -28 lag -18 Surface salinity regressed on AMOC evaporation >0. 02 mm/j lag -28 Mix layer depth > 50 m lag -18 a) a) SPG intensity Florida-Bahamas Mass Transport North Brazil Current ~Lag -28 ~Lag -18 NAO (38%) b) AMOC leads 5) Northward ITCZ shift and NAO- excitation Lag -8 includes the ARPEGE-Climat (v 5. 2) atmospheric model (1. 4°x 1. 4°, 31 vertical levels), the NEMO (v 3. 2) ocean model (ORCA 1°, 42 vertical levels), the ISBA land surface scheme and the GELATO (v 5) sea ice model coupled through the OASIS (v 3) system (see Voldoire et al. , 2012). The AMV mechanisms are investigated using the preindustrial control run of CNRMCM 5 produced within the 5 th Coupled Model Intercomparison Project (CMIP 5) framework. This is a 1000 -yr long simulation where all external forcings (solar, volcanoes and anthropogenic Green House Gases and aerosols) are kept constant to their observed values of 1850. North Atlantic Sea Surface Temperature Variability and its link with thermohaline circulation in CNRM-CM 5 4) SPG and WBC intensification AMOC lags yr yr psu. std-1 yr Pa. std-1 Fig. a-c : Sea Level Pressure EOF over the North Atlantic – Europe region (colours), and associated wind (arrows). Fig. d : cross-correlations between the PCs and AMOC, statistical significant correlation at the 95% level are dotted NAO : North Atlantic Oscillation EAP : East Atlantic Pattern SCAND : Scandinavian mode ( : « lag X » is a mean between lag <X-2; -X+2> ) ! <salinity>0 -500 m regressed on AMOC ~Lag -8 ~Lag 8 Low-pass filtered AMOC Leading EOF (80%) b) correlation c) Profondeur (m) a) 2 Sv between Iso-contours S Eq N N N AMOC leads N Sv Fig. a : Standardized AMV Index wavelet. Fig. b : AMOC climatology (contours) and low-pass filtered (>25 yr) leading EOF (colours). Fig. c : AMV Index auto-correlation (black), and AMV/AMOC Index cross-correlation (blue). Statistical significant correlation at the 95% level are bold In CNRM-CM 5 the AMV is a multidecadal damped mode, primarily linked to the AMOC variability CNRM-CM 5 produces an internal variability mode in the North-Atlantic region comparable to the AMV documented in many model studies. The modelled AMV is a damped mode mainly linked to the AMOC variability. The mechanisms for this variability are the following: EAP or/and Scandinavian atmospheric circulation modes (1) force a northward oceanic heat transport between the eastern branch of the SPG and the GIN Sea (2). After advection along the Norwegian current, this heat transport anomaly precludes sea ice formation along the eastern Greenland coast (2), leading to positive surface salinity anomalies due to enhanced local evaporation (3). The latter is advected to the Labrador Sea by the Eastern Greenland Current where it drives deeper convection (3). By geostrophy, the SPG intensifies and current anomalies gradually propagate backward up to the Equator (4). The Atlantic ITCZ northward shift favours negative NAO (5). These all together product an increase of northward heat transport into the North Atlantic ocean, leading in fine to AMV maximum. This cycle takes about 30 -40 yr to build and is damped about 20 yr later by negative salinity anomalies advected into the SPG by the mean circulation from western tropical Atlantic (6). References : Davison, A. C. , and D. V. Hinkley, 1997: Bootstrap Methods and their Application, Camb. Univ. Press. Folland, C. K. , A. W. Colman, D. P. Rowell, and M. K. Davey, 2001: Predictability of northeast Brazil rainfall and real-time forecast skill, 1987– 98, J. Clim. , 14, 1937– 1958. Goldenberg, S. B. , C. W. Landsea, A. M. Mestas-Nuñez, and W. M. Gray, 2001: The recent increase in Atlantic hurricane activity: Causes and implications, Science, 293, 474– 479. Gray, S. T. , L. J. Graumlich, J. L. Betancourt, and G. T. Pederson, 2004: A tree-ring based reconstruction of the Atlantic Multidecadal Oscillation since 1567 A. D. , Geophys. Res. Lett. , 31, L 12205, doi: 10. 1029/2004 GL 019932. Kerr, R. A. , 2000: A North Atlantic climate pacemaker for the centuries, Science, 288, 1984– 1985. Knight, J. R. , 2009: The Atlantic Multidecadal Oscillation Inferred from the Forced Climate Response in Coupled General Circulation Models, J. Clim. , 22, 1610 -1625 Schlesinger, M. E. , and N. Ramankutty, 1994: Have Solar-Irradiance Variations Influenced Climate? In The Solar Engine and its Influence on the Terrestrial Atmosphere and Climate, E. Nemes-Ribes (ed. ), Springer-Verlag, Heidelberg, pp. 493 -506. Voldoire, A. and coauthors, 2012 : The CNRM-CM 5. 1 global model: description and basic evaluation, Clim. Dyn. , DOI 10. 1007/s 00382 -011 -1259 -y
Скачать презентацию Mechanisms of the North Atlantic multidecadal internal variability
8f5960372fc8255aebe7ab3bc2e9b4bc.ppt