1. Introduction
El Niño–Southern Oscillation (ENSO) is the largest mode of climate interannual variability, influencing climate disasters, ecosystems, and societies worldwide (e.g., Philander 1990; Neelin et al. 1998; McPhaden et al. 2006; Deser et al. 2010). The response of ENSO to surface warming induced by increased greenhouse gases has attracted increasing attention (Meehl et al. 2007; Yeh et al. 2009; Collins et al. 2010; Vecchi and Wittenberg 2010; Christensen et al. 2013; Cai et al. 2015b; Huang and Xie 2015). However, model projections of the changes in ENSO-related sea surface temperature (SST) variability from current state-of-the-art models show large intermodel differences (Timmermann et al. 1999; Van Oldenborgh et al. 2005; Collins et al. 2010; Vecchi and Wittenberg 2010; DiNezio et al. 2012; Stevenson 2012; Watanabe et al. 2012; Kim et al. 2014; Cai et al. 2015b).
Despite the large uncertainty of changes in ENSO-related SST, some recent studies have found that the ENSO-driven rainfall variability over the tropical Pacific will likely intensify over the central-eastern Pacific but weaken over the western Pacific in the projections of models participating in phases 3 and 5 of the Coupled Model Intercomparison Project (CMIP3 and CMIP5) (Power et al. 2013; Cai et al. 2014; Chung and Power 2014; Chung et al. 2014; Bonfils et al. 2015; Cai et al. 2015a; Huang and Xie 2015; Huang 2016). The robust changes in ENSO-induced rainfall variability correspond to the robust eastward shift of ENSO-induced teleconnection patterns (Kug et al. 2010; Zhou et al. 2014; Bonfils et al. 2015), because the local rainfall response is a crucial bridge in the response of teleconnection patterns to ENSO (Lau and Nath 1996; Klein et al. 1999; Alexander et al. 2002).
Some mechanisms have been proposed to explain the robust changes in ENSO-induced rainfall variability (Power et al. 2013; Cai et al. 2014; Chung and Power 2014; Chung et al. 2014; Huang and Xie 2015; Chung and Power 2016; Huang 2016). The nonlinear interaction between background warming and historical ENSO-related SST could increase the rainfall response to ENSO SST (Power et al. 2013; Chung et al. 2014; Huang and Xie 2015; Huang 2016). The spatial pattern of changes in background SST, an El Niño–like pattern with more warming in the tropical central-eastern Pacific (Xie et al. 2010; Huang and Ying 2015; Ying et al. 2016), could shift the response pattern of rainfall eastward to the equatorial eastern Pacific (Power et al. 2013; Cai et al. 2014; Zhou et al. 2014; Huang 2016), owing to the nonlinear relationship between tropical SST and rainfall (Graham and Barnett 1987; Johnson and Xie 2010).
A large portion of previous studies on the changes in ENSO-induced rainfall variability have mainly focused on the linear, or so-called symmetric, part of the rainfall anomalies P′ during the warm and cold phases of ENSO, that is, the El Niño (EN) and La Niña (LN) events. However, a prominent characteristic of ENSO is the asymmetry between EN and LN, which is also called the nonlinearity of ENSO (Hoerling et al. 1997; Burgers and Stephenson 1999; An and Jin 2004; Rodgers et al. 2004; McPhaden and Zhang 2009; Takahashi et al. 2011; Choi et al. 2013; Cai et al. 2015b; Takahashi and Dewitte 2016). The observed SST anomalies T′ associated with strong EN events are often centered in the eastern Pacific, whereas the maximum T′ in strong LN events are located in the central Pacific (Burgers and Stephenson 1999; Takahashi et al. 2011). Meanwhile, the tropical Pacific P′ during EN and LN also show pronounced asymmetries, even when the asymmetry of T′ is not considered (Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010). The largest positive P′ over the tropical Pacific during EN are located farther east than the largest negative P′ during LN because of the modification of zonally nonuniform mean-state SST (Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010). The asymmetric P′ over the tropical Pacific during EN and LN also leads to asymmetric responses of teleconnection patterns to ENSO (Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010).
Approximately, the changes in P′ during EN under global warming are similar to those during LN (Cai et al. 2014; Chung and Power 2014; Chung et al. 2014; Cai et al. 2015a; Chung and Power 2016), which are almost identical to the changes in ENSO-induced P′ from a linear perspective (Power et al. 2013; Huang and Xie 2015; Huang 2016). A few studies have discussed the differences in the changes in P′ between EN and LN. For example, Cai et al. (2014) suggested that the enhanced P′ over the central-eastern Pacific during EN is due to the faster surface warming in the eastern Pacific, whereas the enhanced P′ during LN is attributed to the faster warming of the Maritime Continent (Cai et al. 2015a). Chung and Power (2016) suggested that the changes in the EN- and LN-induced P′ are separately dependent on the amplitude of T′ during EN and LN. It is still unclear whether there are robust asymmetric changes in P′ between EN and LN under global warming, which could lead to asymmetric changes in teleconnection patterns responding to EN and LN (Kug et al. 2010; Zhou et al. 2014; Bonfils et al. 2015).
The present study investigates the changes in P′ induced by EN and LN under global warming, especially their asymmetric changes, based on the output of 32 CMIP5 models in the historical and RCP8.5 runs. Our analysis shows that with the increase in global-mean surface temperature (GMST), P′ during both EN and LN gradually shift eastward under global warming, but the shift during EN is farther east than that during LN. A moisture budget decomposition method is used to explain the formation mechanisms of the asymmetric changes in P′. This method was developed in Huang and Xie (2015) and applied effectively to explain the changes in the linear part of ENSO-induced P′ in Huang (2016). We reveal that the increase in mean-state moisture as a result of surface warming will enlarge the asymmetric response of rainfall to SST anomalies during EN and LN, and the role of mean-state SST on the asymmetric response of rainfall will also be enlarged when the zonal gradient of mean-state SST decreases under global warming.
2. Models and methods
The models, the definitions of climatology and interannual anomaly, and the moisture budget decomposition method used in the present study are the same as in Huang (2016) and are briefly reviewed here. The following introduction of the models and methods is derived from Huang (2016) with minor modifications.
a. Models
This study uses the output of 32 models from the CMIP5 historical runs from 1971 to 2010 and the +8.5 W m−2 representative concentration pathway (RCP8.5) runs from 2006 to 2100 (Taylor et al. 2012). The 32 models are the same as in Huang (2016): ACCESS1.0, ACCESS1.3, BCC_CSM1.1, BCC_CSM1.1(m), BNU-ESM, CanESM2, CCSM4, CESM1(BGC), CESM1(CAM5), CMCC-CESM, CMCC-CM, CMCC-CMS, CNRM-CM5, CSIRO Mk3.6.0, FGOALS-g2, GFDL CM3, GFDL-ESM2G, GFDL-ESM2M, GISS-E2-H, GISS-E2-R, HadGEM2-ES, IPSL-CM5A-LR, IPSL-CM5A-MR, IPSL-CM5B-LR, MIROC5, MIROC-ESM, MIROC-ESM-CHEM, MPI-ESM-LR, MPI-ESM-MR, MRI-CGCM3, NorESM1-M, and NorESM1-ME. (Expansions of acronyms are available online at http://www.ametsoc.org/PubsAcronymList.) These models are selected randomly. All model output is interpolated onto a 2.5° × 2.5° grid before analysis for calculating the multimodel ensemble mean (MME). The 2.5° × 2.5° grid is effective for analyzing large-scale interannual variability. The variables of SST, precipitation, air specific humidity at the surface, and vertical pressure velocity at 500 hPa are used. The MME is defined as the simple average of all models, which is presented in the present study unless stated otherwise.
b. Definition of climatology and ENSO-driven anomaly
The 1971–2000 mean of the historical runs defines the historical climatology. The period 2006–2100 in the RCP8.5 runs is divided into 30-yr segments starting from 2006 with 5-yr leaps: 2006–35, 2011–40, 2016–45, …, 2071–2100. Here, 30-yr segments are selected as in Huang (2016), which is suitable for studying ENSO-related variability and does not influence the conclusions. For the historical runs and each segment in the RCP8.5 runs, the annual cycle based on the 30-yr mean is first removed to define the interannual anomaly, and then a 13-yr running mean is removed to focus on the interannual variability (Power et al. 2013; Huang and Xie 2015).
Empirical orthogonal function (EOF) analysis is performed on the interannual anomalies of SST to define the ENSO-related linear SST variability for the segment in the historical runs and each segment in the RCP8.5 runs (e.g., Philander 1990; Wilks 2006). The first principal components in each segment are standardized. A month with the standardized principal component greater (less) than 1 (−1) is defined as an EN (LN) month. The composite of SST in all the EN (LN) months defines the SST anomalies T′ during EN (LN). The composite of precipitation, moisture, and 500-hPa vertical pressure velocity in all the EN (LN) months defines the ENSO-driven anomalies of these variables. The changes in the ENSO-related anomalies in each future segment of the RCP8.5 runs are defined as the differences between the anomalies in the segment and in the historical runs.
The symmetric, or linear, part of EN- and LN-related anomalies is estimated by their difference, whereas the assymetric, or nonlinear, part is estimated by their sum. The magnitudes of the symmetric and asymmetric parts are measured by the root-mean square of the patterns. A symmetry-to-asymmetry ratio (SAR), calculated by the ratio of the magnitudes of the symmetric and asymmetric parts, is further defined to show the relative importance of the symmetric and asymmetric parts.
c. Moisture budget decomposition
Decomposition of the moisture budget is often used to investigate the formation mechanism of changes in the mean state and variability of tropical rainfall (Held and Soden 2006; Chou et al. 2009; Seager et al. 2010, 2012; Huang et al. 2013; Chung et al. 2014; Huang 2014; Huang and Xie 2015). A simplified moisture budget decomposition method developed in Huang and Xie (2015) is effective for investigating changes in the interannual variability of tropical rainfall (Huang and Xie 2015; Huang 2016). The simplification of the moisture budget decomposition does not influence the conclusions and can clearly illustrate the relationship among tropical rainfall, surface moisture, and vertical mass transportation (Huang and Xie 2015).
The changes in ENSO circulation Δω′ are further decomposed into amplitude and structural changes: Δω′ = Δωa′ + Δωs′. The amplitude changes Δωa′ are obtained by projecting the total Δω′ onto the historical ω′, and the residual is defined as the structural changes Δωs′. Similarly, the changes in ENSO SST ΔT′ are decomposed as ΔT′ = ΔTa′ + ΔTs′. The structural changes in ENSO circulation Δωs′ can be attributed to the structural changes in ENSO SST ΔTs′ and the spatial pattern of the mean-state SST changes (Cai et al. 2014; Huang and Xie 2015). Single field principal component analysis is performed on the intermodel diversity of ΔTs′ and Δωs′ in the 32 models to extract the impact of the MME ΔTs′ on Δωs′, written as Δωs1′ (Bretherton et al. 1992; Huang and Xie 2015; Huang and Ying 2015). The residual of Δωs1′ is written as Δωs2′, where Δωs2 ′ = Δωs′ − Δωs1′, which is mainly dominated by the spatial pattern of the mean-state SST changes.
3. Results
a. Evaluation of modeled ENSO asymmetry
The composited SST anomalies during EN and LN in the historical runs are shown in Figs. 1a,b. The patterns of SST anomalies T′ during EN and LN both extend farther west than in the observations (Figs. S1a,b in the supplementary material), which is a long-standing bias in current global circulation models as revealed in previous studies (Guilyardi et al. 2009; Collins et al. 2010; Vecchi and Wittenberg 2010; Christensen et al. 2013; Power et al. 2013; Capotondi et al. 2015). The models successfully reproduce the observed asymmetric pattern of the SST anomalies between EN and LN, in which the SST anomalies during LN are located farther west than those during EN (Figs. 1a,b and 2b and Figs. S1a,b and S2b in the supplementary material). The sign agreement test in Fig. 2b shows that the EN–LN asymmetry of T′ is robust among the models.
The simulated tropical Pacific anomalies of (a),(b) SST and (c),(d) rainfall (shaded) composited for (a),(c) EN and (b),(d) LN in the historical runs. Contours in (c),(d) are 500-hPa vertical pressure velocity (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The simulated tropical Pacific anomalies of (a),(b) SST and (c),(d) rainfall (shaded) composited for (a),(c) EN and (b),(d) LN in the historical runs. Contours in (c),(d) are 500-hPa vertical pressure velocity (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The simulated tropical Pacific anomalies of (a),(b) SST and (c),(d) rainfall (shaded) composited for (a),(c) EN and (b),(d) LN in the historical runs. Contours in (c),(d) are 500-hPa vertical pressure velocity (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (defined as the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) (a),(b) SST and (c),(d) rainfall anomalies (shaded) associated with EN and LN in the historical runs. Contours in (c),(d) are anomalies of 500-hPa vertical pressure velocity [contour interval is 10−2 Pa s−1 in (c) and is 2 × 10−3 Pa s−1 in (d), and negative contours are dashed]. Stippling in (c),(d) indicates that more than 70% of models agree on the sign of the MME.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (defined as the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) (a),(b) SST and (c),(d) rainfall anomalies (shaded) associated with EN and LN in the historical runs. Contours in (c),(d) are anomalies of 500-hPa vertical pressure velocity [contour interval is 10−2 Pa s−1 in (c) and is 2 × 10−3 Pa s−1 in (d), and negative contours are dashed]. Stippling in (c),(d) indicates that more than 70% of models agree on the sign of the MME.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (defined as the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) (a),(b) SST and (c),(d) rainfall anomalies (shaded) associated with EN and LN in the historical runs. Contours in (c),(d) are anomalies of 500-hPa vertical pressure velocity [contour interval is 10−2 Pa s−1 in (c) and is 2 × 10−3 Pa s−1 in (d), and negative contours are dashed]. Stippling in (c),(d) indicates that more than 70% of models agree on the sign of the MME.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Figures 1c,d and 2c,d show the patterns of P′ during EN and LN and the asymmetries between them in the historical runs. The P′ exhibit apparent asymmetries between EN and LN. The P′ during EN are maximized at around the date line, whereas the P′ during LN are centered west of the date line at around 160°E. The EN–LN asymmetry of P′ (Fig. 2d) is consistent with the observations (Fig. S2d), although both of the P′ centers during EN and LN are farther west than the observations (Figs. 1c,d and Figs. S1c,d) owing to the common model bias, as in the ENSO SST anomalies (Guilyardi et al. 2009; Collins et al. 2010; Vecchi and Wittenberg 2010; Christensen et al. 2013; Capotondi et al. 2015). Also, the EN–LN asymmetry of P′ is robust among the models, illustrated by the sign agreement test (Fig. 2d). The EN–LN asymmetry of P′ in individual models shown in Fig. S3 of the supplementary material also indicates that the vast majority of models can well reproduce the observed EN–LN asymmetry of P′.
The SAR of modeled EN and LN SST anomalies (8.0) is much larger than that for the observations (4.8), as shown in Table 1, showing that the EN–LN asymmetry of T′ in the CMIP3 and CMIP5 models is underestimated relative to observations (Van Oldenborgh et al. 2005; Zhang et al. 2009; Yu and Kim 2011; Sun et al. 2013; Zhang and Sun 2014; Choi et al. 2015; Sun et al. 2016). However, these models successfully reproduce the larger EN–LN asymmetry of P′ than that of T′. The SAR of P′ is only around half of the SAR of T′, both in the MME and the observations (Table 1). This result implies that the models can reproduce the process of the larger asymmetry of P′ than T′ in the observations: the zonally nonuniform mean-state SST induces the asymmetric rainfall response to the EN and LN SST anomalies even when the asymmetry of T′ is not considered (Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010). The well-simulated asymmetric pattern of P′ and T′ and the larger asymmetry of P′ compared with T′ indicate that the models are feasible for investigating the changes in the EN–LN asymmetry of P′.
The SAR of the present-day T′ and P′ and their future changes. The magnitudes of the symmetric and asymmetric parts are represented by their root-mean square.
b. Changes at the end of the twenty-first century
The pattern of changes in SST anomalies ΔT′ during EN under global warming is similar to that of LN in the MME of the CMIP5 models (Figs. 3a,b); both have a weakened amplitude on the whole. Meanwhile, the weakening of SST anomalies in the central Pacific is smaller than that in other regions, implying a structural change in the EN and LN SST anomalies with more frequent central Pacific EN and LN events under global warming (Yeh et al. 2009; Collins et al. 2010). The changes in the EN and LN SST anomalies in the MME of the CMIP5 models are approximately consistent with the symmetric changes in ENSO SST anomalies revealed in previous studies (Power et al. 2013; Huang and Xie 2015). However, the sign agreement test shows that the weakened EN and LN SST anomalies in the MME are not robust among the models, consistent with the large intermodel uncertainty of the changes in the linear ENSO SST (Collins et al. 2010; Power et al. 2013). The weakened EN and LN SST anomalies lead to a weakened EN–LN asymmetry of the SST anomalies (Figs. 2c and 4c).
Changes in tropical Pacific (a),(b) SST and (c),(d) rainfall anomalies (shaded) induced by (a),(c) EN and (b),(d) LN for the period 2071–2100 in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies (contour interval is 2 × 10−5 Pa s−1 kg kg−1, and negative contours are dashed). Red stars and longitudes in (c) and (d) denote the negative–positive demarcations of ENSO rainfall changes on the equator. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Changes in tropical Pacific (a),(b) SST and (c),(d) rainfall anomalies (shaded) induced by (a),(c) EN and (b),(d) LN for the period 2071–2100 in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies (contour interval is 2 × 10−5 Pa s−1 kg kg−1, and negative contours are dashed). Red stars and longitudes in (c) and (d) denote the negative–positive demarcations of ENSO rainfall changes on the equator. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Changes in tropical Pacific (a),(b) SST and (c),(d) rainfall anomalies (shaded) induced by (a),(c) EN and (b),(d) LN for the period 2071–2100 in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies (contour interval is 2 × 10−5 Pa s−1 kg kg−1, and negative contours are dashed). Red stars and longitudes in (c) and (d) denote the negative–positive demarcations of ENSO rainfall changes on the equator. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) changes in (a),(b) SST and (c),(d) rainfall (shaded) associated with EN and LN in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies [contour interval is 4 × 10−5 Pa s−1 kg kg−1 in (c) and is 2 × 10−5 Pa s−1 kg kg−1 in (d), and negative contours are dashed].
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) changes in (a),(b) SST and (c),(d) rainfall (shaded) associated with EN and LN in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies [contour interval is 4 × 10−5 Pa s−1 kg kg−1 in (c) and is 2 × 10−5 Pa s−1 kg kg−1 in (d), and negative contours are dashed].
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a),(c) The symmetric (the difference between EN and LN) and (b),(d) asymmetric (the sum of EN and LN) changes in (a),(b) SST and (c),(d) rainfall (shaded) associated with EN and LN in the RCP8.5 runs. Contours in (c),(d) are the sum of thermodynamic and dynamic components of changes in rainfall anomalies [contour interval is 4 × 10−5 Pa s−1 kg kg−1 in (c) and is 2 × 10−5 Pa s−1 kg kg−1 in (d), and negative contours are dashed].
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Figures 3c,d show the changes in rainfall anomalies ΔP′ during EN and LN. The ΔP′ exhibits a robust zonal dipole pattern with enhancement over the central and eastern Pacific and weakening over the western Pacific during both EN and LN. These change patterns indicate that both the EN- and LN-induced P′ will shift eastward in a warmer climate, similar to the symmetric changes in ENSO P′ in Huang and Xie (2015) and Huang (2016). However, on closer inspection, the shift of P′ during EN is farther east than the shift during LN. The negative–positive demarcation of ΔP′ in EN at the equator lies at 174.2°E, whereas that in LN lies at 156.7°E, around 18° longitude west relative to that during EN. The SAR of ΔP′ is much smaller than the SAR of historical P′ (Table 1), indicating an enhanced asymmetry of P′. The ΔP′ during EN and LN and the enhanced asymmetry (Figs. 4c,d) are robust among the models, although the changes in SST anomalies ΔT′ are not robust and the asymmetry of the SST anomalies is weakened. The ΔP′ in EN and LN are not simply correlated with ΔT′, which is similar to the complicated relationship between the linear ΔP′ and ΔT′ revealed in Huang and Xie (2015) and Huang (2016).
The moisture budget decomposition developed in Huang and Xie (2015) is applied to investigate the formation of the ΔP′ during EN and LN and their asymmetry. The ΔP′ are first decomposed into the thermodynamic (Δqω′) and dynamic (qΔω′) components for EN and LN, respectively (Fig. 5). The sum of the thermodynamic and dynamic components is shown by contours overlaid on ΔP′ in Figs. 3c,d and 4c,d. Their sum can accurately describe the pattern of ΔP′ both during EN and LN, showing that the present moisture budget decomposition is also effective for analyzing ΔP′ separately for the warm and cold phases of ENSO. (Note that negative ω associated with ω′ means upward flow and corresponds to positive rainfall and hence are represented by the same color, blue.) The patterns of historical circulation anomalies (ω′) induced by EN and LN are shown as contours in Figs. 5a–c, and the changes in circulation anomalies (Δω′) are the contours in Figs. 5d–f. The patterns of the thermodynamic Δqω′ and dynamic qΔω′ components are dominated by ω′ and Δω′, respectively, because the spatial gradients of ω′ and Δω′ are much larger than those of mean-state moisture and moisture variability (Huang and Xie 2015; Huang 2016).
(a)–(c) Thermodynamic and (d)–(f) dynamic components of rainfall changes induced by (a),(d) EN, (b),(e) LN, and (c),(f) their sum. Contours in (a)–(c) are the circulation anomalies induced by ENSO in historical runs (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed). Contours in (d)–(f) are the changes in circulation anomalies induced by ENSO for the period 2071–2100 in the RCP8.5 runs (contour interval is 2 × 10−3 Pa s−1, and negative contours are dashed). Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(c) Thermodynamic and (d)–(f) dynamic components of rainfall changes induced by (a),(d) EN, (b),(e) LN, and (c),(f) their sum. Contours in (a)–(c) are the circulation anomalies induced by ENSO in historical runs (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed). Contours in (d)–(f) are the changes in circulation anomalies induced by ENSO for the period 2071–2100 in the RCP8.5 runs (contour interval is 2 × 10−3 Pa s−1, and negative contours are dashed). Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(c) Thermodynamic and (d)–(f) dynamic components of rainfall changes induced by (a),(d) EN, (b),(e) LN, and (c),(f) their sum. Contours in (a)–(c) are the circulation anomalies induced by ENSO in historical runs (contour interval is 5 × 10−3 Pa s−1, and negative contours are dashed). Contours in (d)–(f) are the changes in circulation anomalies induced by ENSO for the period 2071–2100 in the RCP8.5 runs (contour interval is 2 × 10−3 Pa s−1, and negative contours are dashed). Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
In the thermodynamic component Δqω′, the increase in mean-state moisture Δq is the same for both EN and LN. As a result, the thermodynamic component tends to enhance P′ during both EN and LN (Figs. 5a,b), with increased P′ during EN and decreased P′ during LN, and thus the EN–LN asymmetry is also enhanced (Fig. 5c). The dynamic components qΔω′ and Δω′ show an eastward shift relative to the thermodynamic component and the historical P′. It is quite clear that the eastward shift of P′ under global warming is contributed by the dynamic components. A remarkable difference in the dynamic component between EN and LN is that the dynamic component during EN shifts farther east than that during LN. The eastward shift of the dynamic components is stronger than the shift of the total ΔP′. The negative–positive demarcation of the dynamic components qΔω′ and Δω′ in EN and LN are located farther east than the demarcations of the total ΔP′ (Figs. 3c,d). The EN–LN asymmetry of the dynamic component shows a pattern of two in-phase centers west of 150°E and east of 160°W and an out-of-phase center between them (Fig. 5f). The MME results of the thermodynamic and dynamic components and their asymmetries are robust among the models.
In Huang and Xie (2015) and Huang (2016), the formation of the symmetric changes in ENSO-induced circulation variability Δω′ is effectively studied by performing a two-step decomposition on Δω′: Δω′ = Δωa′ + Δωs′ and Δωs′ = Δωs1′ + Δωs2′, and then the roles of the amplitude and structural changes in T′ and the spatial pattern of mean-state SST changes in the formation of Δω′ are revealed. Here, a similar analysis is performed on the Δω′ composited for EN and LN. Meanwhile, the total changes in the EN and LN SST anomalies ΔT′ are decomposed into the amplitude and structural changes: ΔT′ = ΔTa′ + ΔTs′. The amplitude and structural changes in Δω′ and T′ during EN and LN are shown in Figs. 6 and 7, respectively.
The (a)–(c) amplitude and (d)–(f) structural changes in ENSO-driven circulation anomaly for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The (a)–(c) amplitude and (d)–(f) structural changes in ENSO-driven circulation anomaly for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The (a)–(c) amplitude and (d)–(f) structural changes in ENSO-driven circulation anomaly for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The (a)–(c) amplitude and (d)–(f) structural changes in the ENSO-related SST anomalies for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The (a)–(c) amplitude and (d)–(f) structural changes in the ENSO-related SST anomalies for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The (a)–(c) amplitude and (d)–(f) structural changes in the ENSO-related SST anomalies for the period 2071–2100 in the RCP8.5 runs. Stippling indicates that more than 70% of models agree on the sign of the MME changes.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The amplitude changes in the EN and LN circulation anomalies Δωa′ are both weakened with a closer magnitude under global warming (Figs. 6a,b and contours in Figs. 1c,d), which induces a weakened EN–LN asymmetry of ω′ (Fig. 6c and contours in Fig. 2d). The weakened amplitude of ω′ and the EN–LN asymmetry correspond to the amplitude changes in T′ (Figs. 7a–c). However, the robustness of Δωa′ among the models is different from that of ΔTa′. The MME weakening of Δωa′ is robust during both EN and LN. On the other hand, the weakening of ΔTa′ during EN is robust but during LN is uncertain among the models. Huang and Xie (2015) and Huang (2016) revealed that the percentage of the amplitude changes in ENSO circulation variability relative to historical ENSO circulation (Δωa′/ω′) is dominated by the percentage of the amplitude changes in ENSO SST variability relative to historical ENSO SST (ΔTa′/T′). Figure 8 shows the relationship between Δωa′/ω′ and ΔTa′/T′ for EN and LN. The Δωa′/ω′ and ΔTa′/T′ in EN and LN both have a significant linear relationship with a ratio close to 1. As revealed in Huang and Xie (2015) and Huang (2016), the close relationship indicates that the amplitude changes in SST anomalies ΔTa′ are the dominant factor in the amplitude changes in circulation anomalies Δωa′. However, the value of Δωa′/ω′ is smaller than that of ΔTa′/T′ in general, and this difference is more pronounced compared with the results in Huang and Xie (2015) and Huang (2016), in which the regression method was used. The faster and more robust weakening in Δωa′/ω′ compared with ΔTa′/T′ could be attributed to the increased stability of the tropical atmosphere under global warming (Held and Soden 2006; Vecchi and Soden 2007; Chou and Chen 2010; Chou et al. 2013).
The relationship between the amplitude-change percentages in ENSO circulation and ENSO SST. The filled circles indicate the MME. The correlation coefficient is shown in the top-right corner.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The relationship between the amplitude-change percentages in ENSO circulation and ENSO SST. The filled circles indicate the MME. The correlation coefficient is shown in the top-right corner.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The relationship between the amplitude-change percentages in ENSO circulation and ENSO SST. The filled circles indicate the MME. The correlation coefficient is shown in the top-right corner.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The structural changes in the EN and LN circulation anomalies Δωs′ are essentially opposite to each other (Figs. 6d,e). Both are robustly enhanced over the central and eastern Pacific but weakened over the western Pacific. However, Δωs′ shows a remarkable asymmetry between EN and LN (Figs. 6d–f), with a farther eastward shift in EN than in LN. The EN–LN asymmetry of Δωs′ is similar to that of ΔP′ (Figs. 4d and 6f). Corresponding to Δωs′, the structural changes in SST anomalies ΔTs′ are enhanced over the central Pacific but weakened over the western Pacific (Figs. 7d,e). The ΔTs′ during EN and LN are consistent with the symmetric structural changes in ENSO SST based on the regression method (Huang and Xie 2015; Huang 2016). However, ΔTs′ do not show apparent EN–LN asymmetry (Fig. 7f), which cannot explain the EN–LN asymmetry of Δωs′.
The structural changes in the circulation anomalies Δωs′ in the MME are further decomposed into two parts, Δωs′ = Δωs1′ + Δωs2′, as in Huang and Xie (2015) and Huang (2016), in which Δωs1′ and Δωs2′ are respectively influenced by ΔTs′ and the spatial relative changes in mean-state SST. The two separated parts of Δωs′ during EN and LN are shown in Fig. 9. With stronger ΔTs′ during LN over the western Pacific, Δωs1′ during LN is also stronger over the western Pacific than during EN (Figs. 9a–c). However, the asymmetry of Δωs′, especially the dipole pattern centered over around 165°E and 140°W (Fig. 6f), is mainly contributed by the asymmetry of Δωs2′ (Fig. 9f). The positive Δωs2′ during LN extends to the Maritime Continent, whereas the negative–positive demarcation of Δωs2′ during EN is located at around 180° longitude (Figs. 9d,e).
(a)–(c) The estimated structural changes in ENSO circulation driven by the structural changes in ENSO SST (Δωs1′) and (d)–(f) the residual of Δωs1′ in Δωs′ (Δωs2′). Contours in (d),(e) are the relative SST changes (contour interval is 0.05°C, and negative contours are dashed) for the period 2071–2100 in the RCP8.5 runs.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(c) The estimated structural changes in ENSO circulation driven by the structural changes in ENSO SST (Δωs1′) and (d)–(f) the residual of Δωs1′ in Δωs′ (Δωs2′). Contours in (d),(e) are the relative SST changes (contour interval is 0.05°C, and negative contours are dashed) for the period 2071–2100 in the RCP8.5 runs.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(c) The estimated structural changes in ENSO circulation driven by the structural changes in ENSO SST (Δωs1′) and (d)–(f) the residual of Δωs1′ in Δωs′ (Δωs2′). Contours in (d),(e) are the relative SST changes (contour interval is 0.05°C, and negative contours are dashed) for the period 2071–2100 in the RCP8.5 runs.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Previous studies have suggested that the major driver of the eastward shift of Δωs2′ is the spatial pattern of relative changes in mean-state SST, defined as the tropical-mean-removed mean-state SST changes ΔT* (Ham and Kug 2012; Power et al. 2013; Cai et al. 2014, 2015b; Ham and Kug 2015; Huang and Xie 2015; Huang 2016). The relative changes in mean-state SST show an El Niño–like pattern with more warming in the central-eastern Pacific and less warming in other regions (Xie et al. 2010; Huang and Ying 2015; Ying et al. 2016). Owing to the nonlinear relationship between tropical SST and rainfall (Graham and Barnett 1987; Johnson and Xie 2010), the El Niño–like pattern of mean-state SST changes can enhance the sensitivity of convective rainfall to local SST anomalies over the central-eastern Pacific. As a result, the high-sensitivity regions of convective rainfall to local SST anomalies are also shifted eastward (Huang 2016), which leads to the eastward shift of the circulation pattern response to ENSO SST anomalies.
The asymmetric shifting effect of the relative changes in mean-state SST on Δωs2′ during EN and LN can be explained by expanding the previous mechanism. During EN, the positive SST anomalies over the central-eastern Pacific can arouse local rainfall anomalies more effectively when the El Niño–like ΔT* enhances the rainfall–SST sensitivity over the central-eastern Pacific under global warming. On the other hand, the negative SST anomalies during LN will suppress the enhancement of rainfall–SST sensitivity over the central-eastern Pacific. Therefore, the eastward shift of Δωs2′ during EN should be larger than that during LN when induced by the same mean-state SST changes.
The asymmetric role of ΔT* during EN and LN is similar to the asymmetric response of convective rainfall to EN and LN SST anomalies in the present-day climate (Deser and Wallace 1990; Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010). In the present-day climate, the positive SST anomalies during EN are required to arouse convective rainfall anomalies in the eastern Pacific cold tongue, whereas the negative SST anomalies during LN cannot further suppress the normally dry conditions over the cold tongue region (Deser and Wallace 1990; Hoerling et al. 1997). The zonal distance between the centers of rainfall response to the EN and LN SST anomalies is dependent on the zonal gradient of the mean-state SST. Therefore, the asymmetric effect of zonally nonuniform mean-state SST suggested in Hoerling et al. (1997) will be enlarged when the zonal gradient of mean-state SST decreases because of the El Niño–like changes in mean-state SST under global warming.
The enlargement of the relative changes in mean-state SST ΔT* on the EN–LN asymmetry is different from a similar enlargement by the increase in mean-state moisture. The role of ΔT* is to enlarge the difference of the zonal location of the convective rainfall response to EN and LN, whereas the increase of mean-state moisture enlarges the magnitude of the different rainfall responses to EN and LN.
c. Time-varying changes in the twenty-first century
Huang (2016) has shown that the changes in ENSO-induced rainfall variability vary with the increase in GMST. The linear changes in ENSO rainfall are steadily enhanced and shift eastward. The dependence of the shift of ENSO rainfall changes on the increase in GMST can also be observed by comparing the ENSO rainfall changes in the RCP4.5 and RCP8.5 runs. The ΔP′ in the RCP8.5 runs during both EN and LN (Figs. 3c,d) is shifted farther east than that in the RCP4.5 runs (Figs. S4a,b in the supplementary material). Here, a similar analysis is performed on the changes in rainfall anomalies during EN and LN at the equator in the twenty-first century to evaluate possible time-varying changes.
The ΔP′ during EN and LN both exhibit a time-varying evolution in that the changes are steadily enhanced and shifted eastward with the increase in GMST. The thermodynamic and dynamic decomposition can accurately describe ΔP′ during EN and LN for all stages. The steady enhancement of ΔP′ is contributed by the thermodynamic component Δqω′ associated with the increase in mean-state moisture (Figs. 10b and 11b) and the second part of the dynamic component qΔωs2′ (Figs. 10f and 11f). The gradual eastward shift of ΔP′ is mainly due to qΔωs2′. The analysis for the dynamic component qΔωs2′ in Figs. 12d and 13d reveals that qΔωs2′ is dominated by the spatial relative changes in mean-state SST ΔT*. The characteristics and mechanisms of the time-varying ΔP′ during EN and LN are similar to those of the evolution of the linear ΔP′ induced by ENSO, shown in Huang (2016).
(a)–(f) Evolution of the components of changes in EN-driven rainfall at the equator (2.5°S–2.5°N mean). Contours in (a) are the changes in rainfall anomalies (contour interval is 0.1 mm day−1, and negative contours are dashed). The red arrows illustrate the eastward shift of the negative–positive demarcation.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(f) Evolution of the components of changes in EN-driven rainfall at the equator (2.5°S–2.5°N mean). Contours in (a) are the changes in rainfall anomalies (contour interval is 0.1 mm day−1, and negative contours are dashed). The red arrows illustrate the eastward shift of the negative–positive demarcation.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
(a)–(f) Evolution of the components of changes in EN-driven rainfall at the equator (2.5°S–2.5°N mean). Contours in (a) are the changes in rainfall anomalies (contour interval is 0.1 mm day−1, and negative contours are dashed). The red arrows illustrate the eastward shift of the negative–positive demarcation.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Evolution of (a) ΔTs′, (b) Δωs′, (c) Δωs1′, and (d) Δωs2′ driven by EN at the equator (2.5°S–2.5°N mean). The red arrows illustrate the eastward shift of the negative–positive demarcation. Contours in (d) are the relative changes in mean-state SST (contour interval is 0.05°C, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Evolution of (a) ΔTs′, (b) Δωs′, (c) Δωs1′, and (d) Δωs2′ driven by EN at the equator (2.5°S–2.5°N mean). The red arrows illustrate the eastward shift of the negative–positive demarcation. Contours in (d) are the relative changes in mean-state SST (contour interval is 0.05°C, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Evolution of (a) ΔTs′, (b) Δωs′, (c) Δωs1′, and (d) Δωs2′ driven by EN at the equator (2.5°S–2.5°N mean). The red arrows illustrate the eastward shift of the negative–positive demarcation. Contours in (d) are the relative changes in mean-state SST (contour interval is 0.05°C, and negative contours are dashed).
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
Although the major characteristics of the time-varying ΔP′ are quite similar between EN and LN, there are still apparent asymmetries between them. The eastward shift of ΔP′ during EN (Fig. 10a) is more apparent than that during LN (Fig. 11a). The different degree of the eastward shift of ΔP′ between EN and LN along with the increase in GMST can be clearly observed in Fig. 14. The longitude of the negative–positive demarcations of ΔP′ is significantly correlated to the GMST both in EN and LN. The pattern of ΔP′ during EN (LN) shifts eastward by around 13.9° (9.4°) latitude per 1-K increase in GMST. The eastward shift of ΔP′ during EN is 50% larger than that during LN.
The relationship between the negative–positive demarcations of changes in ENSO rainfall (red for EN and green for LN) at the equator and GMST from the period 2021–50 to 2071–2100.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The relationship between the negative–positive demarcations of changes in ENSO rainfall (red for EN and green for LN) at the equator and GMST from the period 2021–50 to 2071–2100.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The relationship between the negative–positive demarcations of changes in ENSO rainfall (red for EN and green for LN) at the equator and GMST from the period 2021–50 to 2071–2100.
Citation: Journal of Climate 30, 4; 10.1175/JCLI-D-16-0427.1
The difference in the negative–positive demarcations of ΔP′ between EN and LN gradually develops along with the increase in GMST, which is mainly contributed by the second part of the dynamic components qΔωs2′ (Figs. 10f and 11f). The negative–positive demarcation of qΔωs2′ during EN extends eastward to around 175°E at the end of the twenty-first century (Fig. 10f), whereas that during LN only extends to around 155°E (Fig. 11f). The difference in qΔωs2′ is dominated by the difference in Δωs2′ between EN and LN (Figs. 12d and 13d). The Δωs2 ′ during EN over the central and eastern Pacific is much larger than that during LN, and the negative–positive demarcation of Δωs2′ during EN extends farther east than that during LN. The asymmetric Δωs2′ between EN and LN develops gradually with the increase in GMST. The asymmetric Δωs2′ can also be explained by the mechanism proposed in the last section: the positive SST anomalies during EN can effectively utilize the effect of the relative changes in mean-state SST to arouse rainfall anomalies over the central-eastern Pacific, whereas the negative SST anomalies during LN must suppress the effect of the relative changes in mean-state SST.
4. Summary
This study investigates the changes in tropical Pacific rainfall anomalies ΔP′ induced by El Niño and La Niña under global warming projected by the MME of 32 CMIP5 models under the RCP8.5 scenario. The ΔP′ during EN and LN are roughly analogous to the linear results of ΔP′ induced by ENSO in previous studies: enhanced (weakened) over the central-eastern (western) Pacific (Power et al. 2013; Cai et al. 2014; Chung and Power 2014; Chung et al. 2014; Zhou et al. 2014; Huang and Xie 2015; Chung and Power 2016; Huang 2016). The patterns of ΔP′ are robust among the models, showing that the EN- and LN-induced P′ will enhance and shift eastward under global warming. However, the ΔP′ still exhibit robust asymmetries between EN and LN, which is more pronounced than the asymmetric changes in SST anomalies between EN and LN. The major EN–LN asymmetry of ΔP′ is the farther eastward shift of P′ during EN than LN. The difference in the eastward shift between EN and LN gradually enlarges with the increase in GMST.
The formation mechanism of ΔP′ during EN and LN and their asymmetries are investigated based on a simplified moisture budget decomposition method developed in Huang and Xie (2015). The results show that this simplified method is also practical for ΔP′ composited for EN and LN. The formation mechanism of ΔP′ during EN and LN is approximately consistent with the mechanism of the changes in ENSO-induced linear rainfall variability in Huang and Xie (2015) and Huang (2016), which is briefly reviewed here. The ΔP′ can be decomposed into thermodynamic and dynamic components, induced by the increase in mean-state moisture and the changes in circulation anomalies. The thermodynamic component enhances the response of rainfall to the EN and LN SST anomalies, whereas the dynamic component shifts the response pattern of rainfall eastward. The changes in circulation anomalies Δω′, as the driver of the dynamic component, can be separated into the amplitude changes and structural changes. The amplitude changes in the circulation anomalies are weakened by the weakened amplitude in EN- and LN-related SST anomalies projected by the MME of 32 CMIP5 models (Stevenson et al. 2012; Kim et al. 2014; Cai et al. 2015b). On the other hand, the structural changes in the circulation anomalies are induced by the structural changes in EN- and LN-related SST anomalies and the spatial relative changes in mean-state SST. The structural changes in SST anomalies enhance the response of circulation over the central Pacific. Meanwhile, the El Niño–like changes in mean-state SST will enhance the circulation response over the central-eastern Pacific but weaken it over the western Pacific (Cai et al. 2014; Huang and Xie 2015; Huang 2016).
The mechanism of asymmetric ΔP′ between EN and LN is also revealed based on the moisture decomposition method. The enhancement of the thermodynamic component of ΔP′ during both EN and LN enlarge the magnitude of asymmetric P′ between EN and LN in the historical runs. The decreased amplitudes of the EN and LN SST anomalies in the MME weaken the asymmetric P′ by influencing the asymmetric circulation anomalies. The structural changes in the EN and LN SST anomalies do not lead to apparent asymmetric ΔP′. The spatial relative changes in mean-state SST with an El Niño–like pattern are the most important factor causing the farther eastward shift of P′ during EN than during LN.
A mechanism is proposed to explain the asymmetric effect of the relative changes in mean-state SST on ΔP′. The EN-like changes in the mean-state SST can enhance the sensitivity of convection to positive SST anomalies associated with EN over the central-eastern Pacific, whereas the LN-related negative SST anomalies will suppress the effect of the relative changes in mean-state SST. As a result, the shift of rainfall anomalies induced by the relative changes in mean-state SST extends farther east during EN than during LN. The asymmetric role of the relative mean-state SST changes can also be interpreted as follows: the mean-state SST changes with an El Niño–like pattern will decrease the zonal gradient of mean-state SST and enlarge the zonal difference of rainfall centers in response to EN and LN SST anomalies induced by the zonally nonuniform mean-state SST (Deser and Wallace 1990; Hoerling et al. 1997; Kang and Kug 2002; Wu et al. 2010).
The increase in mean-state moisture and the relative changes in mean-state SST are quite robust among the models, whereas the decreased amplitudes of the EN and LN SST anomalies in the MME contain large intermodel uncertainties. Therefore, the enhancement and eastward shift of ΔP′ during EN and LN and their asymmetries are robust among the models. This result implies that the teleconnection patterns associated with EN and LN will both shift eastward under global warming (Kug et al. 2010; Zhou et al. 2014), and the asymmetries of the EN- and LN-induced teleconnection patterns (Hoerling et al. 1997; Wu et al. 2010) will be enlarged under global warming.
The present results are mainly based on the MME of 32 CMIP5 models, and the intermodel sign agreement is roughly tested. As we know, large intermodel differences exist in the projection of some variables associated with ENSO in these models (Collins et al. 2010; Christensen et al. 2013; Kim et al. 2014; Cai et al. 2015b). For example, the changes in ENSO-related SST variability are dependent on models, scenarios, and warming stages (Power et al. 2013; Kim et al. 2014; Huang 2017; Rashid et al. 2016), and the spatial pattern of the mean-state SST contains some uncertainties (DiNezio et al. 2009; Xie et al. 2010; Huang and Ying 2015; Zhou and Xie 2015; Ying and Huang 2016a,b; Ying et al. 2016). Huang (2017) has quantified the intermodel uncertainty of changes in ENSO-induced linear rainfall variability. The intermodel uncertainty of asymmetric changes in rainfall anomalies between EN and LN should be further studied to improve projections under global warming.
Acknowledgments
The work was supported by the National Basic Research Program of China (2014CB953904), the National Natural Science Foundation of China (Grant 41575088 and 41461164005), the Foundation of CUIT (Grant KYTZ201602), and the Youth Innovation Promotion Association CAS. We acknowledge the World Climate Research Programme’s Working Group on Coupled Modelling, which is responsible for CMIP5, and the climate modeling groups (listed in section 2a) for producing and making available their model output. We thank Dr. Jun Ying for preparing part of the CMIP5 data. We also thank the three anonymous reviewers for their constructive suggestions.
REFERENCES
Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, N.-C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 2205–2231, doi:10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.
An, S.-I., and F.-F. Jin, 2004: Nonlinearity and asymmetry of ENSO. J. Climate, 17, 2399–2412, doi:10.1175/1520-0442(2004)017<2399:NAAOE>2.0.CO;2.
Bonfils, C. J. W., B. D. Santer, T. J. Phillips, K. Marvel, L. R. Leung, C. Doutriaux, and A. Capotondi, 2015: Relative contributions of mean-state shifts and ENSO-driven variability to precipitation changes in a warming climate. J. Climate, 28, 9997–10 013, doi:10.1175/JCLI-D-15-0341.1.
Bretherton, C. S., C. Smith, and J. M. Wallace, 1992: An intercomparison of methods for finding coupled patterns in climate data. J. Climate, 5, 541–560, doi:10.1175/1520-0442(1992)005<0541:AIOMFF>2.0.CO;2.
Burgers, G., and D. B. Stephenson, 1999: The “normality” of El Niño. Geophys. Res. Lett., 26, 1027–1030, doi:10.1029/1999GL900161.
Cai, W., and Coauthors, 2014: Increasing frequency of extreme El Niño events due to greenhouse warming. Nat. Climate Change, 4, 111–116, doi:10.1038/nclimate2100.
Cai, W., and Coauthors, 2015a: Increased frequency of extreme La Niña events under greenhouse warming. Nat. Climate Change, 5, 132–137, doi:10.1038/nclimate2492.
Cai, W., and Coauthors, 2015b: ENSO and greenhouse warming. Nat. Climate Change, 5, 849–859, doi:10.1038/nclimate2743.
Capotondi, A., Y.-G. Ham, A. Wittenberg, and J.-S. Kug, 2015: Climate model biases and El Niño Southern Oscillation (ENSO) simulation. U.S. CLIVAR Variations, Vol. 13 (1), U.S. CLIVAR Project Office, Washington, DC, 21–25. [Available online at https://usclivar.org/sites/default/files/documents/2015/Variations2015Winter_0.pdf.]
Choi, K.-Y., G. A. Vecchi, and A. T. Wittenberg, 2013: ENSO transition, duration, and amplitude asymmetries: Role of the nonlinear wind stress coupling in a conceptual model. J. Climate, 26, 9462–9476, doi:10.1175/JCLI-D-13-00045.1.
Choi, K.-Y., G. A. Vecchi, and A. T. Wittenberg, 2015: Nonlinear zonal wind response to ENSO in the CMIP5 models: Roles of the zonal and meridional shift of the ITCZ/SPCZ and the simulated climatological precipitation. J. Climate, 28, 8556–8573, doi:10.1175/JCLI-D-15-0211.1.
Chou, C., and C.-A. Chen, 2010: Depth of convection and the weakening of tropical circulation in global warming. J. Climate, 23, 3019–3030, doi:10.1175/2010JCLI3383.1.
Chou, C., J. D. Neelin, C.-A. Chen, and J.-Y. Tu, 2009: Evaluating the “rich-get-richer” mechanism in tropical precipitation change under global warming. J. Climate, 22, 1982–2005, doi:10.1175/2008JCLI2471.1.
Chou, C., T.-C. Wu, and P.-H. Tan, 2013: Changes in gross moist stability in the tropics under global warming. Climate Dyn., 41, 2481–2496, doi:10.1007/s00382-013-1703-2.
Christensen, J. H., and Coauthors, 2013: Climate phenomena and their relevance for future regional climate change. Climate Change 2013: The Physical Science Basis, T. F. Stocker et al., Eds., Cambridge University Press, 1217–1308. [Available online at http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_Chapter14_FINAL.pdf.]
Chung, C. T. Y., and S. B. Power, 2014: Precipitation response to La Niña and global warming in the Indo-Pacific. Climate Dyn., 43, 3293–3307, doi:10.1007/s00382-014-2105-9.
Chung, C. T. Y., and S. B. Power, 2016: Modelled impact of global warming on ENSO-driven precipitation changes in the tropical Pacific. Climate Dyn., 47, 1303–1323, doi:10.1007/s00382-015-2902-9.
Chung, C. T. Y., S. B. Power, J. M. Arblaster, H. A. Rashid, and G. L. Roff, 2014: Nonlinear precipitation response to El Niño and global warming in the Indo-Pacific. Climate Dyn., 42, 1837–1856, doi:10.1007/s00382-013-1892-8.
Collins, M., and Coauthors, 2010: The impact of global warming on the tropical Pacific Ocean and El Niño. Nat. Geosci., 3, 391–397, doi:10.1038/ngeo868.
Deser, C., and J. M. Wallace, 1990: Large-scale atmospheric circulation features of warm and cold episodes in the tropical Pacific. J. Climate, 3, 1254–1281, doi:10.1175/1520-0442(1990)003<1254:LSACFO>2.0.CO;2.
Deser, C., M. A. Alexander, S.-P. Xie, and A. S. Phillips, 2010: Sea surface temperature variability: Patterns and mechanisms. Annu. Rev. Mar. Sci., 2, 115–143, doi:10.1146/annurev-marine-120408-151453.
DiNezio, P. N., A. C. Clement, G. A. Vecchi, B. J. Soden, B. P. Kirtman, and S.-K. Lee, 2009: Climate response of the equatorial Pacific to global warming. J. Climate, 22, 4873–4892, doi:10.1175/2009JCLI2982.1.
DiNezio, P. N., B. P. Kirtman, A. C. Clement, S.-K. Lee, G. A. Vecchi, and A. Wittenberg, 2012: Mean climate controls on the simulated response of ENSO to increasing greenhouse gases. J. Climate, 25, 7399–7420, doi:10.1175/JCLI-D-11-00494.1.
Graham, N. E., and T. P. Barnett, 1987: Sea surface temperature, surface wind divergence, and convection over tropical oceans. Science, 238, 657–659, doi:10.1126/science.238.4827.657.
Guilyardi, E., and Coauthors, 2009: Understanding El Niño in ocean–atmosphere general circulation models: Progress and challenges. Bull. Amer. Meteor. Soc., 90, 325–340, doi:10.1175/2008BAMS2387.1.
Ham, Y.-G., and J.-S. Kug, 2012: How well do current climate models simulate two types of El Niño? Climate Dyn., 39, 383–398, doi:10.1007/s00382-011-1157-3.
Ham, Y.-G., and J.-S. Kug, 2015: Improvement of ENSO simulation based on intermodel diversity. J. Climate, 28, 998–1015, doi:10.1175/JCLI-D-14-00376.1.
Held, I. M., and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19, 5686–5699, doi:10.1175/JCLI3990.1.
Hoerling, M. P., A. Kumar, and M. Zhong, 1997: El Niño, La Niña, and the nonlinearity of their teleconnections. J. Climate, 10, 1769–1786, doi:10.1175/1520-0442(1997)010<1769:ENOLNA>2.0.CO;2.
Huang, P., 2014: Regional response of annual-mean tropical rainfall to global warming. Atmos. Sci. Lett., 15, 103–109, doi:10.1002/asl2.475.
Huang, P., 2016: Time-varying response of ENSO-induced tropical Pacific rainfall to global warming in CMIP5 models. Part I: Multimodel ensemble results. J. Climate, 29, 5763–5778, doi:10.1175/JCLI-D-16-0058.1.
Huang, P., 2017: Time-varying response of ENSO-induced tropical Pacific rainfall to global warming in CMIP5 models. Part II: Intermodel uncertainty. J. Climate, 30, 595–608, doi:10.1175/JCLI-D-16-0373.1.
Huang, P., and S.-P. Xie, 2015: Mechanisms of change in ENSO-induced tropical Pacific rainfall variability in a warming climate. Nat. Geosci., 8, 922–926, doi:10.1038/ngeo2571.
Huang, P., and J. Ying, 2015: A multimodel ensemble pattern regression method to correct the tropical Pacific SST change patterns under global warming. J. Climate, 28, 4706–4723, doi:10.1175/JCLI-D-14-00833.1.
Huang, P., S.-P. Xie, K. Hu, G. Huang, and R. Huang, 2013: Patterns of the seasonal response of tropical rainfall to global warming. Nat. Geosci., 6, 357–361, doi:10.1038/ngeo1792.
Johnson, N. C., and S.-P. Xie, 2010: Changes in the sea surface temperature threshold for tropical convection. Nat. Geosci., 3, 842–845, doi:10.1038/ngeo1008.
Kang, I.-S., and J.-S. Kug, 2002: El Niño and La Niña sea surface temperature anomalies: Asymmetry characteristics associated with their wind stress anomalies. J. Geophys. Res., 107, 4372, doi:10.1029/2001JD000393.
Kim, S. T., W. Cai, F.-F. Jin, A. Santoso, L. Wu, E. Guilyardi, and S.-I. An, 2014: Response of El Niño sea surface temperature variability to greenhouse warming. Nat. Climate Change, 4, 786–790, doi:10.1038/nclimate2326.
Klein, S. A., B. J. Soden, and N.-C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917–932, doi:10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.
Kug, J.-S., S.-I. An, Y.-G. Ham, and I.-S. Kang, 2010: Changes in El Niño and La Niña teleconnections over North Pacific–America in the global warming simulations. Theor. Appl. Climatol., 100, 275–282, doi:10.1007/s00704-009-0183-0.
Lau, N.-C., and M. J. Nath, 1996: The role of the “atmospheric bridge” in linking tropical Pacific ENSO events to extratropical SST anomalies. J. Climate, 9, 2036–2057, doi:10.1175/1520-0442(1996)009<2036:TROTBI>2.0.CO;2.
McPhaden, M. J., and X. Zhang, 2009: Asymmetry in zonal phase propagation of ENSO sea surface temperature anomalies. Geophys. Res. Lett., 36, L13703, doi:10.1029/2009GL038774.
McPhaden, M. J., S. E. Zebiak, and M. H. Glantz, 2006: ENSO as an integrating concept in Earth science. Science, 314, 1740–1745, doi:10.1126/science.1132588.
Meehl, G. A., and Coauthors, 2007: Global climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 747–845. [Available online at http://www.ipcc.ch/pdf/assessment-report/ar4/wg1/ar4-wg1-chapter10.pdf.]
Neelin, J. D., D. S. Battisti, A. C. Hirst, F.-F. Jin, Y. Wakata, T. Yamagata, and S. E. Zebiak, 1998: ENSO theory. J. Geophys. Res., 103, 14 261–14 290, doi:10.1029/97JC03424.
Philander, S. G., 1990: El Niño, La Niña, and the Southern Oscillation. Academic Press, 293 pp.
Power, S., F. Delage, C. Chung, G. Kociuba, and K. Keay, 2013: Robust twenty-first-century projections of El Niño and related precipitation variability. Nature, 502, 541–545, doi:10.1038/nature12580.
Rashid, H. A., A. C. Hirst, and S. J. Marsland, 2016: An atmospheric mechanism for ENSO amplitude changes under an abrupt quadrupling of CO2 concentration in CMIP5 models. Geophys. Res. Lett., 43, 1687–1694, doi:10.1002/2015GL066768.
Rodgers, K. B., P. Friederichs, and M. Latif, 2004: Tropical Pacific decadal variability and its relation to decadal modulations of ENSO. J. Climate, 17, 3761–3774, doi:10.1175/1520-0442(2004)017<3761:TPDVAI>2.0.CO;2.
Seager, R., N. Naik, and G. A. Vecchi, 2010: Thermodynamic and dynamic mechanisms for large-scale changes in the hydrological cycle in response to global warming. J. Climate, 23, 4651–4668, doi:10.1175/2010JCLI3655.1.
Seager, R., N. Naik, and L. Vogel, 2012: Does global warming cause intensified interannual hydroclimate variability? J. Climate, 25, 3355–3372, doi:10.1175/JCLI-D-11-00363.1.
Stevenson, S. L., 2012: Significant changes to ENSO strength and impacts in the twenty-first century: Results from CMIP5. Geophys. Res. Lett., 39, L17703, doi:10.1029/2012GL052759.
Stevenson, S. L., B. Fox-Kemper, M. Jochum, R. Neale, C. Deser, and G. Meehl, 2012: Will there be a significant change to El Niño in the twenty-first century? J. Climate, 25, 2129–2145, doi:10.1175/JCLI-D-11-00252.1.
Sun, Y., D.-Z. Sun, L. Wu, and F. Wang, 2013: Western Pacific warm pool and ENSO asymmetry in CMIP3 models. Adv. Atmos. Sci., 30, 940–953, doi:10.1007/s00376-012-2161-1.
Sun, Y., F. Wang, and D.-Z. Sun, 2016: Weak ENSO asymmetry due to weak nonlinear air–sea interaction in CMIP5 climate models. Adv. Atmos. Sci., 33, 352–364, doi:10.1007/s00376-015-5018-6.
Takahashi, K., and B. Dewitte, 2016: Strong and moderate nonlinear El Niño regimes. Climate Dyn., 46, 1627–1645, doi:10.1007/s00382-015-2665-3.
Takahashi, K., A. Montecinos, K. Goubanova, and B. Dewitte, 2011: ENSO regimes: Reinterpreting the canonical and Modoki El Niño. Geophys. Res. Lett., 38, L10704, doi:10.1029/2011GL047364.
Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485–498, doi:10.1175/BAMS-D-11-00094.1.
Timmermann, A., J. Oberhuber, A. Bacher, M. Esch, M. Latif, and E. Roeckner, 1999: Increased El Niño frequency in a climate model forced by future greenhouse warming. Nature, 398, 694–697, doi:10.1038/19505.
Van Oldenborgh, G. J., S. Y. Philip, and M. Collins, 2005: El Niño in a changing climate: A multi-model study. Ocean Sci., 1, 81–95, doi:10.5194/os-1-81-2005.
Vecchi, G. A., and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20, 4316–4340, doi:10.1175/JCLI4258.1.
Vecchi, G. A., and A. T. Wittenberg, 2010: El Niño and our future climate: Where do we stand? Wiley Interdiscip. Rev.: Climate Change, 1, 260–270.
Watanabe, M., J.-S. Kug, F.-F. Jin, M. Collins, M. Ohba, and A. T. Wittenberg, 2012: Uncertainty in the ENSO amplitude change from the past to the future. Geophys. Res. Lett., 39, L20703, doi:10.1029/2012GL053305.
Wilks, D. S., 2006: Statistical Methods in the Atmospheric Sciences. 2nd ed. International Geophysics Series, Vol. 91, Elsevier, 627 pp.
Wu, B., T. Li, and T. Zhou, 2010: Asymmetry of atmospheric circulation anomalies over the western North Pacific between El Niño and La Niña. J. Climate, 23, 4807–4822, doi:10.1175/2010JCLI3222.1.
Xie, S.-P., C. Deser, G. A. Vecchi, J. Ma, H. Teng, and A. T. Wittenberg, 2010: Global warming pattern formation: Sea surface temperature and rainfall. J. Climate, 23, 966–986, doi:10.1175/2009JCLI3329.1.
Yeh, S.-W., J.-S. Kug, B. Dewitte, M.-H. Kwon, B. P. Kirtman, and F.-F. Jin, 2009: El Niño in a changing climate. Nature, 461, 511–514, doi:10.1038/nature08316.
Ying, J., and P. Huang, 2016a: Cloud–radiation feedback as a leading source of uncertainty in the tropical Pacific SST warming pattern in CMIP5 models. J. Climate, 29, 3867–3881, doi:10.1175/JCLI-D-15-0796.1.
Ying, J., and P. Huang, 2016b: The large-scale ocean dynamical effect on uncertainty in the tropical Pacific SST warming pattern in CMIP5 models. J. Climate, 29, 8051–8065, doi:10.1175/JCLI-D-16-0318.1.
Ying, J., P. Huang, and R. Huang, 2016: Evaluating the formation mechanisms of the equatorial Pacific SST warming pattern in CMIP5 models. Adv. Atmos. Sci., 33, 433–441, doi:10.1007/s00376-015-5184-6.
Yu, J.-Y., and S. T. Kim, 2011: Reversed spatial asymmetries between El Niño and La Niña and their linkage to decadal ENSO modulation in CMIP3 models. J. Climate, 24, 5423–5434, doi:10.1175/JCLI-D-11-00024.1.
Zhang, T., and D.-Z. Sun, 2014: ENSO asymmetry in CMIP5 models. J. Climate, 27, 4070–4093, doi:10.1175/JCLI-D-13-00454.1.
Zhang, T., D.-Z. Sun, R. Neale, and P. J. Rasch, 2009: An evaluation of ENSO asymmetry in the community climate system models: A view from the subsurface. J. Climate, 22, 5933–5961, doi:10.1175/2009JCLI2933.1.
Zhou, Z.-Q., and S.-P. Xie, 2015: Effects of climatological model biases on the projection of tropical climate change. J. Climate, 28, 9909–9917, doi:10.1175/JCLI-D-15-0243.1.
Zhou, Z.-Q., S.-P. Xie, X.-T. Zheng, Q. Liu, and H. Wang, 2014: Global warming–induced changes in El Niño teleconnections over the North Pacific and North America. J. Climate, 27, 9050–9064, doi:10.1175/JCLI-D-14-00254.1.