1. Introduction
Monsoon rainfall supplies more than half of the annual rainfall to global monsoon regions (Trenberth et al. 2000; Wang and Ding 2008; Lee and Wang 2014; Ni and Hsu 2018), and the variability of monsoon rainfall has a profound impact on agriculture, the economy, and natural disasters in global monsoon regions (e.g., Gadgil and Kumar 2006; Mall et al. 2006; Akinsanola and Zhou 2019a). Global monsoon rainfall in observational records has experienced substantial decadal changes (Wang and Ding 2006; Zhou et al. 2008; Zhang and Zhou 2011; Han et al. 2019), due to either natural variability or anthropogenic forcing.
Future changes in monsoon rainfall in response to global warming have attracted much attention (Hill 2019; Pascale et al. 2019; Seth et al. 2019; Wang et al. 2020a). Based on the climate models participating in phases 3 and 5 of the Coupled Model Intercomparison Project (CMIP3 and CMIP5, respectively), global monsoon rainfall may increase by 2%–3% per 1°C of warming under anthropogenic greenhouse gas (GHG) forcing (Hsu et al. 2012; Kitoh et al. 2013; Lee and Wang 2014). The overall increase in global monsoon rainfall can be attributed to an increase in atmospheric water vapor content. Atmospheric water vapor content increases by about 7% per 1°C of warming according to the Clausius–Clapeyron relation, but the upward motion over the monsoon regions generally weakens by about 4%–5% because the enhanced tropospheric static stability acts to suppress the upward motion and convection under global warming (Giannini 2010; Seth et al. 2011; Kitoh et al. 2013; Lee and Wang 2014; Li et al. 2015). This mechanism is known as the “wet-get-wetter” mechanism (Held and Soden 2006), and it explains why rainfall increases at a lower rate than the increase in atmospheric water vapor content.
Despite an overall increase in global monsoon rainfall, nonuniform changes are reported in different monsoon regions in the Northern Hemisphere, based on CMIP3 and CMIP5 models. The Asian (ASN) and North African (NAF) monsoon rainfall is projected to increase under global warming, but the change in North American (NAM) monsoon rainfall is reported to be uncertain (Kitoh et al. 2013; Maloney et al. 2014; Seth et al. 2019). The domain of the NAM monsoon can be defined following two different approaches. Following the concept of the global monsoon (Wang and Ding 2008), the NAM monsoon is located over Central America and the surrounding ocean (Fig. 1a) and is reported to decrease based on coupled model projections (Kitoh et al. 2013; Endo and Kitoh 2014; Lee and Wang 2014; Chen et al. 2020; Wang et al. 2020a,b). The term “North American monsoon” also refers to a narrow land area with complex topography over northwestern Mexico and the southwestern United States (Adams and Comrie 1997; Hoell et al. 2016; Pascale et al. 2017, 2019; Colorado-Ruiz et al. 2018), and its future change is reported to be uncertain (Seth et al. 2019). By correcting the SST bias, particularly the negative SST bias over the North Atlantic, Pascale et al. (2017) reported a reduction of rainfall in the monsoon region of northern Mexico and the southwestern United States. Because an intercomparison among different monsoon regions requires a unified definition of monsoon domain, the concept of the global monsoon is adopted in this study, in which monsoon regions are defined based on the annual cycle of local rainfall (Wang and Ding 2008; Lee and Wang 2014; Ni and Hsu 2018; Wang et al. 2020a). We use the term “NAM monsoon” to refer to the monsoon region of Central America and the adjacent ocean (Fig. 1a) in this study, based on the concept of the global monsoon.
Mean precipitation (shading; unit: mm day−1) and wind at 850 hPa (vectors) for MJJAS during 1981 to 2010, based on (a) GPCP and NCEP–DOE data, (b) the MMM of CMIP5 models, (c) and the MMM of CMIP6 models. The regions enclosed by the yellow curves are the monsoon regions defined based on the GPCP data.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
The distinct response of NAM monsoon rainfall to global warming is attributed to a change in SST. El Niño–like SST warming over the equatorial Pacific (Collins et al. 2010; Luo et al. 2015) and the relatively weak SST warming over the subtropical North Atlantic are both claimed to be responsible for the reduction in NAM monsoon rainfall (Rauscher et al. 2011; Johnson et al. 2020; Wang et al. 2020a). Besides SST warming, CO2 direct radiative forcing also plays a nonnegligible role in redistributing rainfall and modifying atmospheric circulation under global warming (Bony et al. 2013; Merlis 2015; Chadwick 2016; Ceppi et al. 2018). In coupled GCMs (CGCMs), the effects of CO2 direct radiative forcing and SST change can be separated via the fast and slow responses of CGCMs to an abrupt increase in CO2 concentration (Chen et al. 2014; Ceppi et al. 2018). These two effects can also be investigated with stand-alone atmospheric GCMs (AGCMs) without air–sea coupling (Bala et al. 2010; Shaw and Voigt 2015; Chadwick 2016; He and Soden 2016; Chadwick et al. 2019). The ASN and NAF monsoon regions are located mostly over continents, whereas a large fraction of the NAM monsoon is located over the ocean. CO2 direct radiative forcing has distinct impacts on precipitation over continents and the ocean (Bony et al. 2013; Merlis 2015), but it is not clear how the response of the NAM monsoon rainfall to global warming is related to the direct forcing of increased CO2 concentration.
Standard experiments based on state-of-the-art climate models participating in CMIP6 have recently been released, with higher model resolution and improved model physics (Eyring et al. 2019). Future climate projection experiments under Shared Socioeconomic Pathways (SSPs) are performed by a majority of CMIP6 models, along with other useful sensitivity experiments to understand the mechanism of climate change (Eyring et al. 2016; O’Neill et al. 2016; Webb et al. 2017). With the CMIP6 output, we aim to answer the following two questions in this study: 1) Do CMIP5 and CMIP6 models consistently show a distinct response of the NAM monsoon compared with the ASN and NAF monsoons? 2) How is NAM monsoon rainfall modulated by CO2 direct forcing and SST change in model projections?
The remainder of the paper is organized as follows. The data, model, and method are described in section 2, and the response of monsoon rainfall in CMIP6 models is compared with that of CMIP5 models in section 3. The mechanism for the distinct response of NAM monsoon rainfall is investigated in section 4 by using a hierarchy of model simulations, and the major conclusion are summarized in section 5 with a discussion.
2. Data, model, and method
a. Data and model
Since a global monsoon covers not only land but also oceanic regions, global (including ocean) precipitation datasets are needed to identify monsoon regions. The Global Precipitation Climatology Project (GPCP) monthly precipitation dataset (Adler et al. 2003) with a 2.5° horizontal resolution is adopted in this study. A horizontal resolution of about 2.5° can capture global monsoon regions (Lee and Wang 2014; Ni and Hsu 2018). We use the 1981–2010 climatology based on the GPCP dataset to identify global monsoon regions, and quantitatively evaluate the monsoon rainfall change in the three monsoon regions over the Northern Hemisphere. We also use the Global Precipitation Climatology Center (GPCC) land precipitation dataset (Schneider et al. 2011) with a horizontal resolution of 0.5° to identify land monsoon regions, in order to test the sensitivity of the results to the monsoon domain definition. Wind data at 850 hPa derived from the National Centers for Environmental Prediction–Department of Energy (NCEP–DOE) reanalysis (Kanamitsu et al. 2002) are also used for a brief model evaluation.
Available models in CMIP5 and CMIP6 with a full set of monthly atmospheric variables are adopted to perform a future-projection study on the change in monsoon rainfall over the Northern Hemisphere. There are 30 models from CMIP5 and 30 models from CMIP6 with a full set of atmospheric variables in the Historical experiment and future climate projection experiments under high-emission scenarios. High-emission scenarios are adopted since forced climate change is more likely to emerge from internal variability under stronger forcing. The 30 CMIP5 models adopted in this study are exactly the same as in He et al. (2019), and the names of the CMIP6 models are listed in Table S1 in the online supplemental material. In the Historical experiment, the coupled models are driven by the observed historical external forcing (i.e., greenhouse gases, aerosols, etc.). In both the representative concentration pathway 8.5 (RCP8.5) experiment of CMIP5 and the shared socioeconomic pathway SSP5–8.5 experiment of CMIP6, anthropogenic emissions follow a high-emission pathway toward 8.5 W m−2 in AD 2100, but the temporal evolution of emissions in SSP5–8.5 is slightly different from that in the RCP8.5 scenario (van Vuuren et al. 2011; O’Neill et al. 2016). The RCP8.5 and SSP5–8.5 experiments are compared with the Historical experiment to project future change. Because both the scenario and the models are updated from CMIP5 to CMIP6, we cannot attribute the difference in projected climate change simply to the update of climate models.
Given the existence of great internal variability (Deser et al. 2012; Thompson et al. 2015), we use an average of 50 consecutive years (rather than the 20 years used by the IPCC) to construct the mean state climate in each experiment. The late twenty-first century (2050–99, referred to herein as 21C) of global warming scenarios (RCP8.5 and SSP5–8.5) is compared with the late twentieth century (1950–99, referred to as 20C) in the Historical experiment. The forced response is extracted as the multimodel ensemble median (MMM) of the difference between 21C and 20C, as the median is more robust to outlier values than the average (Gleckler et al. 2008). As the emission pathway is not exactly the same for the RCP8.5 and SSP5–8.5 scenarios, the projected change is scaled by the amplitude of tropical (30°S–30°N) mean SST warming to facilitate a comparison (i.e., scaled by 2.49 K for the RCP8.5 scenario and 2.68 K for the SSP5–8.5 scenario).
To understand the relative roles of direct CO2 radiative forcing and SST warming, the abrupt-4xCO2 experiment is compared with the preindustrial control (piControl) experiment based on 42 CGCMs from CMIP6 (see Table S1). The piControl experiment is performed under external forcing fixed at the AD 1850 level, and the abrupt-4xCO2 experiment is performed by abruptly quadrupling CO2 concentration from the piControl level and integrating the CGCM for 150 years (Eyring et al. 2016). The fast response of the CGCM after CO2 quadrupling without sufficient ocean warming is dominated by direct CO2 radiative forcing, but there is currently no exact definition of the fast response (Chen et al. 2014; Ceppi et al. 2018). We note that tropical SST warms fast in the first 10 years in the abrupt-4xCO2 experiment (Fig. S1 in the online supplemental information). To avoid mixing the effects of direct CO2 forcing and SST warming on climate change, the change in the first year of the abrupt-4xCO2 experiment relative to the average of the last 50 years of the piControl experiment is defined as the fast response, and the change in the average of the last 50 years relative to the first year of the abrupt-4xCO2 experiment is defined as the slow response. We also tried defining the fast response as the average of the first 3 years (Fig. S5), and the results are consistent.
The AMIP, AMIP-4xCO2, and AMIP-future4K experiments based on nine available CMIP6 models (Webb et al. 2017) are also adopted in this study, to further investigate the climatic effects of direct CO2 radiative forcing and the SST warming (Shaw and Voigt 2015; He and Soden 2016; Chadwick 2016; Chadwick et al. 2019). The AMIP experiment in CMIP6 is performed by forcing the AGCM with the observed monthly SST from 1979 to 2014. The AMIP-4xCO2 experiment is the same as the AMIP experiment except that the atmospheric CO2 concentration is quadrupled. The AMIP-future4K experiment is the same as the AMIP experiment except that SST warming is added to the global SST, where the amplitude of global mean SST warming is 4 K and the pattern of SST warming is derived from the response of CMIP3-CGCMs to CO2 quadrupling (Webb et al. 2017). A comparison between the AMIP-4xCO2 and AMIP experiments illustrates the climatic effect of direct radiative forcing due to increased CO2 concentration, and a comparison between the AMIP-future4K and AMIP experiments shows the climatic effect of global SST change. The CMIP5 and CMIP6 experiments adopted in this study are summarized in Table 1.
A brief overview of the CMIP5 and CMIP6 numerical experiments analyzed.
The Community Atmosphere Model version 5 (CAM5; Neale et al. 2012), at a horizontal resolution of 1.9° latitude × 2.5° longitude, is adopted to investigate the impact of SST change over a certain region on the NAM monsoon rainfall. This coarse resolution is enough to capture the NAM monsoon, as the NAM monsoon region spans from about 135° to 60°W (Fig. 1a). The Control experiment is forced by the climatological annual cycle of SST (Hurrell et al. 2008) and integrated for 110 years, and the sensitivity experiments are forced by modified SST and also integrated for 110 years. The modified SST in the sensitivity experiment is obtained by adding an SST anomaly over a specific region of interest (e.g., over the equatorial Pacific) to the SST climatology in the Control experiment, as described fully in section 4b. The average of the last 100 years of the sensitivity experiment is compared with the average of the last 100 years of the Control experiment, and a Student’s t test is adopted to test the significance.
The Linear Baroclinic Model (LBM), developed by Watanabe and Kimoto (2000), is a simple model with only a dynamic core, and it is adopted in this study to investigate the atmospheric circulation anomaly stimulated by prescribed diabatic heating over a specific region. The LBM is configured at T42 resolution with 20 vertical levels and is forced by a prescribed steady diabatic heating anomaly, under a mean state averaged for May–September (MJJAS). The projected change in global diabatic heating in the CMIP6 models under the SSP5–8.5 scenario is calculated following Yanai and Tomita (1998), and the projected changes in three-dimensional diabatic heating over two regions are adopted to force the LBM as described in section 4c. Each LBM experiment is integrated for 70 days, and the average over the last 40 days is taken as the steady response to the prescribed heating.
b. Method
To facilitate a comparison among different monsoon regions, a unified definition of a monsoon is required, and we follow the concept of the global monsoon (GM) in this study (Wang and Ding 2008; Lee and Wang 2014; Ni and Hsu 2018; Wang et al. 2020a). The monsoon region in the Northern Hemisphere is defined where the local summer (MJJAS) precipitation minus the winter [November–March (NDJFM)] precipitation exceeds 2.5 mm day−1 and local summer precipitation exceeds 55% of total annual precipitation (Lee and Wang 2014), based on GPCP data for the 30-yr climatology between 1981 and 2010.
A brief model evaluation of the precipitation and 850-hPa wind in MJJAS simulated by the Historical experiment is carried out. Given the available temporal coverage of the GPCP dataset, the model evaluation is performed for the period from 1981 to 2010. As the Historical experiment in CMIP5 models terminates in 2005, the 2006–10 period of the RCP4.5 experiment is added at the end of the Historical experiment to facilitate the model evaluation. The significance in the MMM-projected change is evaluated as the intermodel consistency, and the MMM-projected change is considered significant if more than 70% of the individual models agree on the sign of the change.
Moisture budget analysis is adopted to track the source of the change in precipitation (Chou et al. 2009; Endo and Kitoh 2014; Akinsanola and Zhou 2019b; Chen et al. 2020). Following Chou et al. (2009), the moisture budget equation is written as
where P, E, q, ω, and V stand for precipitation, evaporation, specific humidity, vertical velocity, and horizontal wind vector, respectively; R is the residual. The angle brackets ⟨·⟩ stand for vertical integration from the surface to the top of the troposphere (200 hPa), as almost all the moisture that produces precipitation concentrates in the troposphere, particularly in the boundary layer (Huang et al. 2013). Variables with an overbar stand for the climatology in 20C, and variables with a prime stand for the projected change. Equation (1) shows that the change in precipitation is contributed by the following factors: the change in local evaporation (E′), the thermodynamic and dynamic changes in vertical moisture advection (
3. Monsoon rainfall response in CMIP5 and CMIP6 models
The 1981–2010 climatology of summer precipitation and 850-hPa wind is shown in Fig. 1 for observations and the MMMs of CMIP5 and CMIP6 models, respectively. Based on the 1981–2010 precipitation climatology of GPCP data, the global monsoon region is identified and shown as yellow contours in Fig. 1a, following the definition of Lee and Wang (2014). The ASN monsoon covers a large area from South Asia to East Asia and the western Pacific, the NAF monsoon covers the region between equatorial Africa and the Sahel, and the NAM monsoon covers the region around Central America. The main bodies of the ASN and NAF monsoons are located over continents, whereas a large part of the NAM monsoon is over the ocean. The land monsoon regions derived from the GPCC dataset are generally consistent with the land portion of monsoon regions derived from the GPCP dataset (Figs. S2a,b). The MMMs of the CMIP5 and CMIP6 models capture the intense rainfall in summer over the three monsoon regions and along the ITCZ (Figs. 1b,c; cf. Fig. 1a), and also the key atmospheric circulation transporting moisture into the monsoon regions, such as the southwesterly Somali jet over the northwestern Indian Ocean and the easterly Caribbean low-level jet over the Caribbean Sea (vectors in Fig. 1).
The intermodel spread of the regional averaged precipitation over monsoon regions in summer generally covers the rainfall amount derived from the GPCP dataset (Fig. S3). The fractional biases for the MJJAS precipitation over the ASN, NAF, and NAM monsoon regions, defined as the difference between the model and observations scaled by observations, are −4.3%, −6.5%, and −21.0% for the MMM of the CMIP5 models, and 1.3%, −13.2%, and −0.9% for the MMM of the CMIP6 models. Generally, the CGCMs underestimate the monsoon rainfall, and the bias is smaller in the CMIP6 models compared with the CMIP5 models, especially over the NAM region.
The projected change in rainfall in MJJAS under the high-emission scenarios is shown in Fig. 2, based on the MMM of the 30 CMIP5 models under the RCP8.5 scenario (Fig. 2a) and the MMM of the 30 CMIP6 models under the SSP5–8.5 scenario (Fig. 2b). They share a similar pattern of substantially increased rainfall over the equatorial Pacific and the ASN–NAF monsoon region. The in-phase change between ASN and NAF monsoon rainfall was also identified by previous studies in terms of decadal variability (Li et al. 2017) and paleoclimate evidence (Stager et al. 2011). Decreased rainfall is seen over the NAM monsoon region, and it extends to the subtropical North Atlantic Ocean. Based on the MMM, the fractional precipitation changes are 3.8%, 1.6%, and −2.3% K−1 over the ASN, NAF, and NAM regions under the RCP8.5 scenario of the CMIP5 models, and 4.1%, 3.8%, and −3.0% K−1 over the ASN, NAF, and NAM regions under the SSP5–8.5 scenario of the CMIP6 models (purple circles in Fig. 3). More than 70% of the individual models agree on these MMM-projected changes, as seen from the 30th and 70th percentiles among the individual models, shown as the error bar in Fig. 3.
(a) The MMM-projected change in MJJAS precipitation based on the RCP8.5 scenario of CMIP5 models. (b) As in (a), but based on CMIP6 models under the SSP5–8.5 scenario. All changes are scaled by the amplitude of tropical mean SST warming, and significant changes agreed by more than 70% of the individual models are stippled. The regions enclosed by the purple curves are the monsoon regions defined based on the GPCP data.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
Regional mean changes of ASN, NAF, and NAM monsoon rainfall and their moisture budget by (a) CMIP5 and (b) CMIP6 models. The MMMs of changes in precipitation and moisture budget terms are shown as colored bars using the left y axis, and the MMMs of fractional precipitation changes are shown as purple bullets using the right y axis. The thin error bar indicates the range between the 30th and 70th percentiles among the individual models, and the diamonds indicate the 15th and 85th percentiles among the individual models for the changes in precipitation.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
Among the 30 individual models from CMIP5 and the 30 models from CMIP6, all models project an increased ASN monsoon rainfall; 22 models from CMIP5 and 23 from CMIP6 project increased NAF monsoon rainfall; and 22 models from CMIP5 and 27 from CMIP6 project decreased NAM monsoon rainfall (Fig. S3). Averaging over the global land monsoon regions based on the GPCC dataset (Fig. S2b), the increased ASN–NAF monsoon rainfall and decreased NAM monsoon rainfall are still evident, and 70% of the individual models from both CMIP5 and CMIP6 agree on all these changes (Fig. S4b), consistent with recent studies (Chen et al. 2020; Wang et al. 2020a,b). To disclose the large-scale dynamics responsible for the distinct change in the NAM monsoon rainfall, we focus on the entire NAM monsoon region (land and ocean) in this study.
A moisture budget analysis is performed, and the regional averaged values for the terms in Eq. (1) are shown in Fig. 3 to track the source of the distinct rainfall changes over the three regions. Similar results are obtained based on the CMIP5 and CMIP6 models (Figs. 3a,b). Overall, the residual term is not the largest term in the MMM (green bars in Fig. 3) or in any of the individual models (figure not shown). The changes in monsoon rainfall are dominated by changes in vertical moisture advection, whereas changes in evaporation and horizontal moisture advection make little contribution. The thermodynamic component of vertical moisture advection (blue bars in Fig. 3) acts to increase the precipitation, whereas the dynamic change in vertical moisture advection (red bars in Fig. 3) acts to reduce the precipitation over all three monsoon regions. Over the ASN and NAF monsoon regions, the dynamic component is overwhelmed by the thermodynamic component of vertical moisture advection, and the precipitation increases. Over the NAM monsoon region, the dynamic component of vertical moisture advection is particularly strong; thus it overwhelms the thermodynamic component and results in decreased precipitation. With the aid of moisture budget analysis, previous studies have also confirmed that the dynamic component is responsible for the reduced rainfall over the entire NAM monsoon region (Endo and Kitoh 2014) and over the NAM land monsoon region (Chen et al. 2020).
Because the dynamic component of vertical moisture advection is controlled by the change in vertical velocity, the change in the midtropospheric vertical velocity is examined in Fig. 4a based on the CMIP6 models. The changes in upward motion over the ASN and NAF monsoon regions are generally weak, consistent with the weaker dynamic component of vertical moisture advection. In contrast, the upward motion over the NAM monsoon region is substantially weakened, and the anomalous subsidence over the NAM monsoon region is the strongest in the subtropical Northern Hemisphere (Fig. 4a). The NAM monsoon rainfall decreases as a direct result of the strong anomalous subsidence, although the Caribbean low-level jet transporting moisture into the NAM monsoon region (Wang 2007; Martinez et al. 2019) seems to be enhanced (vectors in Fig. 4a). The anomalous subsidence is not limited within the NAM monsoon region but extends into the subtropical North Atlantic, explaining the decreased precipitation over this region. In addition, the upward motion is substantially enhanced over the equatorial Pacific (Fig. 4a).
MMM-projected change in (a) 500-hPa vertical velocity (shading; unit: 10−2 Pa s−1) and 850-hPa wind (vectors), and (b) SST* (unit: K) based on CMIP6 models. All changes are scaled by the amplitude of tropical mean SST warming, and significant changes agreed by more than 70% of the individual models are stippled.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
As tropical precipitation follows relatively high SST rather than a fixed SST threshold (Johnson and Xie 2010; Xie et al. 2010; Seth et al. 2019), we show the projected change in SST* in Fig. 4b, where SST* is defined as the difference between SST and tropical (30°S–30°N) mean SST. The projected change in SST* is characterized by an obvious El Niño–like pattern in the CMIP6 models, with enhanced warming in the equatorial Pacific and relative cooling over the subtropical North Atlantic. This pattern is consistent with the SST warming pattern in the CMIP3 and CMIP5 models (Vecchi and Soden 2007; Collins et al. 2010; Luo et al. 2015; Li et al. 2016) and explains the substantially enhanced upward motion and precipitation over the equatorial Pacific. One thing worth noting is that the observed tropical Pacific SST change in past decades shows a La Niña–like pattern rather than an El Niño–like pattern, and it is still under debate whether this inconsistency results from internal climate variability, such as the interdecadal Pacific oscillation (Kosaka and Xie 2013), or a common bias of the models (Seager et al. 2019).
Weakened upward motion is a direct cause for the robust decrease in NAM monsoon rainfall under global warming. Although increased tropospheric static stability acts to weaken the upward motion in the global monsoon regions (Kitoh et al. 2013; Lee and Wang 2014; Wang et al. 2020b), the distinct response of NAM monsoon in contrast to ASN and NAF monsoon cannot be explained by enhanced tropospheric static stability, as the increase of tropospheric stabilization is rather spatially uniform (He and Li 2019). The CMIP6 models show an El Niño–like SST warming and relatively weak SST warming over the subtropical North Atlantic, and previous studies attributed the reduced NAM monsoon rainfall to the SST pattern change in these two regions (Rauscher et al. 2011; Wang et al. 2020a,b). However, up to now it has been unclear whether the change in SST is the primary cause for the decrease in NAM monsoon rainfall, and this will be investigated in the next section.
4. Mechanism for reduced NAM rainfall
a. Fast and slow responses of CGCMs to quadrupling CO2
Based on the MMM of the 33 CGCMs from CMIP6, the fast and slow responses to abrupt quadrupling of CO2 are shown in Figs. 5 and 6, to investigate the effects of direct CO2 forcing and SST change. Decreased precipitation over the NAM monsoon region and subtropical North Atlantic is seen in the fast response (Fig. 5a), in contrast to increased precipitation over the ASN and NAF monsoon regions. This is not in contrast with Pascale et al. (2017), who found a negligible impact of direct CO2 forcing on the land monsoon rainfall over North America at a higher latitude. In the fast response to CO2 quadrupling, anomalous upward motion is seen over the land area in North Africa, whereas anomalous downward motion is seen over a vast area of tropical ocean (Fig. 5b). Indeed, Bony et al. (2013) suggested that the dynamic component is more important than the thermodynamic component for the precipitation changes in the fast response. The response of 850-hPa wind is characterized by a large-scale cyclone anomaly from the Eurasian-African continent to the North Atlantic, with a westerly wind anomaly from the equatorial Atlantic to the NAF region, and a northeasterly wind anomaly over the subtropical North Atlantic (Fig. 5b). The westerly wind anomaly over the NAF region favors enhanced NAF monsoon rainfall by transporting more water vapor from the equatorial Atlantic. The low-level northeasterly wind anomaly over the subtropical North Atlantic may stimulate anomalous subsidence via negative moist enthalpy advection (Wu et al. 2017).
MMM of the (a),(b) fast and (c),(d) slow responses of 33 CGCMs to abrupt quadrupling of CO2, and (e),(f) the sum of the fast and slow responses. Columns show (left) the changes in precipitation (shading; unit: mm day−1) and (right) the change in 500-hPa vertical velocity (shading; unit: 10−2 Pa s−1) and 850-hPa wind (vectors). Significant changes on which more than 70% of the individual models agree are stippled. The ASN, NAF, and NAM monsoon regions are indicated by purple curves.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
MMM of the (a) fast and (b) slow responses of 33 CGCMs to abrupt quadrupling of CO2. The shading shows the change in land surface temperature and SST after removing tropical mean SST warming (LST* and SST*; unit: K), and the contours show the change in sea level pressure. The purple contours stand for 0.5, 1.5, and 2.5 hPa, and the green contours stand for −0.5, −1.5, and −2.5 hPa. Significant changes on which more than 70% of the individual models agree are stippled.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
In the fast response of the CGCMs, the amplitude of tropical SST warming is only 0.62 K, and the SST pattern change is weak (Fig. 6a). As expected, the amplitude of land warming is much greater than that of the ocean (Fig. 6a). The widespread warming over the subtropical Eurasian-African continent, the largest continental area in the subtropical Northern Hemisphere, results in a large-scale decrease in sea level pressure (SLP) over the subtropical continent, in comparison with an increase in SLP over the other regions in the subtropical Northern Hemisphere (contours in Fig. 6a) This is consistent with the zonal wavenumber-1 response suggested by Shaw and Voigt (2015). The strong warming over the largest subtropical continent and the associated change in SLP pattern explain the low-level cyclone anomaly seen in Fig. 5b.
The slow response of precipitation to CO2 quadrupling is characterized by a further reduction in NAM monsoon rainfall and a sharp increase in rainfall over the equatorial Pacific (Fig. 5c). Although the NAM monsoon rainfall is reduced in both the fast and slow responses, the negative rainfall anomaly in the slow response is confined to a narrow band north of the equator (Fig. 5c). In contrast to the vast subsidence anomaly over almost the entire subtropical Atlantic in the fast response, the anomalous subsidence in the slow response is confined within the NAM monsoon region (Fig. 5d), which explains the narrow but strong reduction in rainfall over the NAM monsoon region (Fig. 5c). The change in atmospheric circulation is weak over the subtropical Atlantic, but a strong ascending anomaly is seen over the equatorial Pacific, with a low-level northerly wind anomaly across the equator over the equatorial eastern Pacific (Fig. 5d).
The SST pattern changes substantially in the slow response, and it is characterized by enhanced SST warming over the equatorial Pacific and relative cooling over the subtropical North Atlantic (Fig. 6b), consistent with the SSP5–8.5 scenario and previous studies (Vecchi and Soden 2007; Collins et al. 2010; Luo et al. 2015). This El Niño–like SST warming pattern is also a robust phenomenon in coupled models after model-bias correction (Li et al. 2016). Because the enhanced SST warming over the equatorial Pacific is located on the southern flank of the NAM monsoon, the El Niño–like SST warming is responsible for enhanced rainfall over the equatorial Pacific and suppressed rainfall over the NAM monsoon region (Fig. 5c) via a southward shift of ITCZ (Rauscher et al. 2011; Wang et al. 2020a,b), since tropical convection follows the relatively higher SST (Johnson and Xie 2010; Long et al. 2014; Seth et al. 2019).
The sum of the fast and slow responses (i.e., the total response of the CGCMs; Figs. 5e,f) shares a similar pattern to the projected changes under the SSP5–8.5 scenario. The change in NAM monsoon precipitation is −14.9% (Fig. 7), and the amplitude of tropical SST warming is 4.0 K in the total response of the CGCMs to CO2 quadrupling. Therefore, the total response of NAM rainfall is equivalent to −3.7% K−1, which is consistent with and slightly greater than the projection under the transient SSP5–8.5 scenario, suggesting that the projected change under global warming is driven primarily by the increase in CO2 concentration instead of aerosols or other factors. The amplitude of the NAM monsoon precipitation change is −7.0% and −7.3% for the fast and the slow responses based on the MMM, and more than 70% of the individual models agree on the reduction (Fig. 7). The evidence based on the abrupt-4xCO2 experiment suggests that direct CO2 forcing and SST change are almost equally important for the reduced NAM monsoon rainfall.
Quantitative changes of the NAM monsoon rainfall in the abrupt-4xCO2 experiment. The purple bullets stand for MMM-simulated fractional changes in precipitation (unit: %) with respect to the piControl experiment, and the thin error bar stands for the 30th and 70th percentiles among individual models.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
b. AGCM responses to quadrupling CO2 and SST change
The relative contributions from quadrupling CO2 and SST change simulated by nine available AGCMs in CMIP6 are analyzed in this subsection, to address the following two questions: 1) Can the fast response of CGCMs be reproduced by AGCMs forced by quadrupling CO2 with exactly fixed SST? 2) Can the reduction of NAM monsoon rainfall due to SST change be reproduced by AGCMs without air–sea interaction? Figure 8 shows the responses of the precipitation and atmospheric circulation to quadrupling CO2 as the difference between the AMIP-4xCO2 and AMIP experiments, and the responses to SST warming as the difference between the AMIP-future4K and AMIP experiments.
MMM of the AGCM response to (a),(b) CO2 quadrupling and(c),(d) change in SST, and (e),(f) their sum. Columns show (left) the changes in precipitation (shading; unit: mm day−1), with the monsoon regions indicated by purple curves, and (right) the change in 500-hPa vertical velocity (shading; unit: 10−2 Pa s−1), 850 hPa wind (vectors), and sea level pressure (contours; purple contours indicate 0.5, 1.0, 1.5, and 2.5 hPa, and green contours indicate −0.5, −1.0, −1.5, and −2.0 hPa). Significant changes on which more than 70% of the individual models agree are stippled.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
Forced by CO2 quadrupling with fixed SST, AGCMs also show increased precipitation over the ASN and NAF monsoon regions, and decreased rainfall from the NAM monsoon region to the subtropical North Atlantic (Fig. 8a), similar to the pattern of precipitation change in the fast response of the CGCMs. As in the fast response of the CGCMs, the SLP decreases over the Eurasian–African continent but increases over other regions in the subtropical Northern Hemisphere (contours in Fig. 8b), and an anomalous cyclone appears over the Eurasian–African continental area, with anomalous ascending motion over the ASN–NAF monsoon region and anomalous subsidence over the subtropical North Atlantic (Fig. 8b). The anomalous subsidence over the subtropical North Atlantic overlaps with the anomalous northeasterly wind on the western side of the anomalous cyclone (shading and vectors in Fig. 8b), also suggesting that the anomalous subsidence is stimulated by the negative moist enthalpy advection associated with the northeasterly wind anomaly (Wu et al. 2017). The AGCM results confirm that the fast response of CGCMs is dominated by the direct forcing of increased CO2 concentration, despite a weak SST change in the fast response of CGCMs. The amplitude of the response to quadrupling CO2 is slightly greater in the AGCMs than in the fast response of the CGCMs, possibly because the small increase in global SST in the CGCMs reduces the land–ocean thermal contrast compared with the AGCM simulations.
The AGCM-simulated response to SST change captures the sharp increase in rainfall over the equatorial Pacific and the planetary-scale responses of precipitation and atmospheric circulation (Figs. 8c,d), but it shows an increase in NAM rainfall (Fig. 8c) and an insignificant change in local vertical velocity (Fig. 8d), which is different from the slow response of the CGCMs (Figs. 5c,d). Although He and Soden (2016) showed that AGCMs can reproduce the planetary-scale pattern of precipitation response to global warming without air–sea coupling, the results here show that AGCMs fail to simulate the regional response of NAM monsoon rainfall to SST change. As a result, the amplitude of decrease in NAM monsoon rainfall is underestimated by the sum of the AGCM responses to CO2 quadrupling and SST change (Fig. 8e), although the change in the planetary-scale atmospheric circulation is captured (Fig. 8f). The decomposition of CO2 direct forcing and SST change by AGCMs works well at a planetary scale and over the ASN monsoon region (Bony et al. 2013; He and Soden 2016; Endo et al. 2018), but it cannot capture the regional precipitation change over the NAM monsoon region. Further decomposition into the relative effects of uniform SST warming and SST pattern change (e.g., Chadwick 2016) by AGCMs is not suitable for investigating the NAM monsoon rainfall.
Idealized experiments are performed by CAM5, by prescribing global SST* anomalies projected by the MMM of the CMIP6 models under the SSP5–8.5 scenario (see Fig. 4b for the SST*). The response of CAM5 to CMIP6-projected change in global SST* is also characterized by increased precipitation along the equatorial Pacific and the NAM monsoon region (Fig. 9a), similar to Fig. 8c based on the AMIP-future4K experiment of the CMIP6 models. This increase in NAM monsoon rainfall is probably forced by the local warm SST*, because a large increase in NAM rainfall is simulated by CAM5 if it is forced by the CMIP6-projected positive SST* over the western coast of Central America (Fig. 9b). The changes in NAM monsoon precipitation show a weak intermodel negative correlation with the changes in local SST* (or local SST* and LST*) under the SSP5–8.5 scenario of CMIP6 models (Fig. S6), suggesting that decreased NAM monsoon rainfall in the coupled model projection is not a response to local SST* change (Wu et al. 2009). However, the NAM monsoon rainfall does respond to the prescribed local SST* change in the AGCMs, which explains the failure of AGCM-simulated NAM monsoon rainfall due to global SST (or SST*) change.
CAM5-simulated precipitation (shading; unit: mm day−1) and 850-hPa wind (vectors; unit: m s−1) responses to the prescribed forcing adopted from the change in SST* shown in Fig. 4b. The changes in SST* as a forcing are prescribed (a) globally and over (b) the NAM region, (c) the equatorial Pacific, and (d) the subtropical North Atlantic. The SST* prescribed as forcing is shown as contours (black, red, and blue contours indicate zero, positive, and negative values, respectively, with a contour interval of 0.3 K), and the prescribed SST* forcing is located within the region enclosed by the black contour in (b)–(d). The precipitation anomaly significant at the 95% confidence level is stippled, and only the wind anomalies significant at the 95% confidence level are shown.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
As previous studies have proposed that the enhanced SST warming over the equatorial Pacific and the relative cooling over the subtropical North Atlantic are responsible for the decrease of NAM monsoon rainfall (Rauscher et al. 2011; Wang et al. 2020a,b), experiments are performed by CAM5 to re-examine which of these two regions is more important in suppressing the NAM monsoon rainfall. Forced by the positive SST* over the equatorial Pacific (within the black contour in Fig. 9c) adopted from Fig. 4b, substantially decreased rainfall over the NAM monsoon region can be reproduced by CAM5, associated with an easterly wind anomaly over the NAM monsoon region and a northerly wind anomaly across the equator (Fig. 9c), similar to the slow response of the CGCMs (Figs. 5c,d). As the formation of abundant NAM monsoon rainfall in summer is associated with the northward migration of the ITCZ, it is reasonable that the enhanced SST warming over the equatorial Pacific prevents the ITCZ from shifting northward into the NAM monsoon region (Seth et al. 2019). Forced by the negative SST* over the subtropical North Atlantic (within the black contour in Fig. 9d) adopted from Fig. 4b, the change in NAM monsoon rainfall is weak and insignificant (Fig. 9d). The above evidence suggests that the slow response of NAM monsoon rainfall is dominated by the El Niño–like SST warming over the equatorial Pacific, but the relative SST cooling over the subtropical North Atlantic has no detectable contribution. This does not contradict Johnson et al. (2020), since the negative SST* over the subtropical western North Atlantic is weaker and narrower than the positive SST* over the equatorial Pacific in this study (Fig. 4b).
c. LBM response to change in diabatic heating
To investigate the possible precipitation–circulation feedback process, the response of the atmospheric circulation to CMIP6-projected changes in atmospheric diabatic heating is investigated by using LBM (Watanabe and Kimoto 2000). Forced by enhanced diabatic heating (mainly latent heating, red shading in Fig. 10a) over the land area within 5°–30°N, 0°–120°E projected by the MMM of CMIP6 models, an anomalous cyclone appears in the lower troposphere from North Africa to Europe (Fig. 10a), reminiscent of a Rossby wave response to enhanced latent heating over the ASN-NAF monsoon region (Gill 1980). This confirms that the enhanced diabatic heating over the continental monsoon region can enhance the anomalous cyclone to its west, which further enhances the responses of precipitation and circulation initiated by CO2 direct forcing. As the mean state moist enthalpy in the lower troposphere over the subtropical North Atlantic decreases northeastward (gray shading in Fig. 10a), the anomalous northeasterly wind on the western flank of the anomalous cyclone induces a negative moist enthalpy advection over the subtropical North Atlantic, acting to suppress the precipitation by creating subsidence (Wu et al. 2017).
Response of 850-hPa wind (vectors) to prescribed diabatic heating simulated by LBM, and the vertically averaged three-dimensional diabatic heating prescribed as forcing (red and blue shading; unit: K day−1). The prescribed heating is (a) the enhanced land monsoon heating within 5°–30°N, 0°–120°E and (b) the reduced NAM monsoon heating. Gray shading (interval of 104 J kg−1, starting from 3 × 104 J kg−1) shows the 20C mean-state moist enthalpy at 925 hPa based on the MMM of CMIP6 models, in which lighter (darker) shading indicates lower (higher) moist enthalpy.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
Forced by the reduced latent heating over the NAM monsoon region projected by the MMM of CMIP6 models (blue shading in Fig. 10b), an anomalous anticyclone appears over the eastern coast of North America, with a low-level northeasterly wind anomaly over the subtropical western Atlantic (Fig. 10b). This northeasterly wind anomaly may further create local subsidence and suppress precipitation via negative moist enthalpy advection (Wu et al. 2017). Therefore, there is also a positive feedback between the reduction in NAM monsoon rainfall and the changes in the atmospheric circulation, similar to the “convection-circulation feedback” proposed by Xiang et al. (2013). Once the NAM monsoon rainfall is suppressed by either direct CO2 radiative forcing or SST change over the equatorial Pacific, the reduction in NAM monsoon rainfall will be amplified through the above positive feedback.
5. Conclusions and discussion
In this study, the response of NAM monsoon rainfall to global warming is investigated based on the output of CMIP6 and CMIP5 models, in comparison with the ASN and NAF monsoons. To understand the mechanism for the distinct response of NAM monsoon rainfall, a hierarchy of model experiments is analyzed. The relative effects from direct CO2 forcing and SST change are investigated, based on the fast and slow responses of CMIP6 CGCMs to abrupt quadrupling of CO2, and the responses of CMIP6 AGCMs to quadrupling of CO2 and changed SST. Idealized experiments were carried out by CAM5 to identify the key oceanic region responsible for the response of NAM monsoon rainfall, and LBM experiments were also performed to investigate the feedback process between changes in precipitation and atmospheric circulation. The major findings are summarized as follows, and the mechanism for the reduction in NAM monsoon rainfall is illustrated schematically in Fig. 11.
As consistently projected by the CMIP5 and CMIP6 models under high-emission scenarios, the NAM monsoon rainfall in summer will decrease under a warmer climate, in sharp contrast to the increased ASN–NAF monsoon rainfall. The quantitative changes in NAM monsoon rainfall are about −2.3% and −3.0% K−1 based on the MMMs of the CMIP5 and CMIP6 models, respectively, with more than 70% of the individual models in agreement. Moisture budget analysis shows that the distinct response of NAM monsoon rainfall to global warming is a result of the dynamic component of vertical moisture advection, associated with the weakening of upward motion over the NAM monsoon region.
Substantial reduction in NAM monsoon rainfall is seen in both the fast and slow responses of the CGCMs to CO2 quadrupling, suggesting that the direct radiative forcing of CO2 is equally as important as the SST change for the reduction in NAM monsoon rainfall. As the Eurasian–African continent is the largest continent in the subtropical Northern Hemisphere, a large-scale cyclone forms over the subtropical continent due to the increased land–sea thermal contrast under increased CO2 concentration. The enhanced ASN–NAF monsoon precipitation associated with the continental cyclone anomaly may further enhance this anomaly via a positive precipitation–circulation feedback. The northerly wind anomaly on the western flank of the continental cyclone anomaly induces subsidence over the subtropical North Atlantic via negative moist enthalpy advection, suppressing the NAM monsoon rainfall and the rainfall over the subtropical North Atlantic.
Based on the sensitivity experiments performed by CAM5, the El Niño–like SST warming pattern plays an essential role in the further reduced NAM monsoon rainfall due to SST change, whereas the relative cooling of SST over the subtropical North Atlantic has no contribution. As tropical convection follows relatively higher SST, the warmer equatorial SST on the southern flank of the NAM monsoon region shifts the tropical convection southward, which enhances the precipitation along the equatorial Pacific and suppresses the precipitation over the NAM monsoon region. There is also a positive feedback between decreased NAM monsoon rainfall and the change in atmospheric circulation, which amplifies the response of NAM monsoon rainfall to direct CO2 forcing and SST change. The reduction in NAM monsoon rainfall due to SST pattern change is strongly dependent on local air–sea interaction, and it cannot be reproduced by forcing AGCMs with global SST change.
A schematic for the two mechanisms of NAM monsoon rainfall reduction under global warming. Green and yellow shading indicates increased and decreased precipitation, respectively, and arrows stand for the change in low-level wind. Mechanism 1: Under the direct forcing of increased CO2 concentration, a surface low pressure (the red ellipse) forms over the Eurasian–African continent due to enhanced land–sea thermal contrast, associated with a cyclone anomaly (blue arrows). The northerly wind anomaly on the western flank of this anomalous cyclone induces anomalous subsidence over the subtropical North Atlantic (including the NAM monsoon region) via negative moist enthalpy advection. A positive feedback may exist between the continental cyclone anomaly and enhanced ASN-NAF monsoon rainfall. Mechanism 2: The enhanced SST warming over the equatorial Pacific associated with an El Niño–like SST warming pattern shifts the tropical rainbelt southward, enhancing precipitation over the equatorial Pacific and suppressing the NAM monsoon rainfall.
Citation: Journal of Climate 33, 22; 10.1175/JCLI-D-20-0189.1
The simulation of the mean state climate by both CMIP5 and CMIP6 models is not realistic (Fig. 1), which may lead to bias in the model-projected future climate change. However, the possible impact of the 20C climate simulation on the future projection may not be straightforward. A preliminary intermodel correlation analysis shows that the projected change in precipitation is weakly correlated with the precipitation in 20C for these three monsoon regions (Fig. S9). The major aim of this study is to answer why the models project a reduction of NAM monsoon rainfall in contrast to the Asian–African monsoon, and more effort is needed in the future to disclose which aspects of the 20C climate simulation have an effect on the projected change in monsoon rainfall. The emergent constraint approach is a promising way to correct the impact of model bias on climate projection (Xie et al. 2015; Brient 2020).
In this study, monsoon regions are defined based on the observed climatology of precipitation, and monsoon regions based on observations are adopted to evaluate the model-projected monsoon rainfall changes. The monsoon regions identified based on the MMMs of the CMIP5 and CMIP6 models are somewhat different from those in the observations (Figs. S2c,d), with an excessive eastward extension of the ASN monsoon region and an excessive westward extension of the NAF monsoon region, which is a common bias in CMIP5 and CMIP6 models. Averaged within the monsoon regions simulated by CMIP5 or CMIP6 models, the regional averaged monsoon rainfall changes are still characterized by increased ASN–NAF monsoon rainfall and reduced NAM monsoon rainfall, despite some quantitative differences (Figs. S4c,d). No matter whether the monsoon domains are defined based on the CMIP5 MMM or CMIP6 MMM, the MMM of CMIP6 models shows a stronger response of monsoon rainfall in each monsoon region than CMIP5 models (Figs. S4c,d), but it is currently hard to judge whether this is a result of the update in the scenarios or the models.
It is worth mentioning that the NAM under the current definition deals with the summer precipitation over Central America and surrounding oceans following the concept of the global monsoon (Wang and Ding 2008; Lee and Wang 2014), but what has been traditionally defined as the core region of the North American monsoon is the region over northwestern Mexico extending into the southwestern United States (Adams and Comrie 1997; Pascale et al. 2017, 2019). Thus, a caution is needed to interpret the projected change of the NAM monsoon rainfall. The conclusions reached in this paper are mostly valid for Central America and surrounding oceans but less valid for the core region of the NAM monsoon. The response of the core region of NAM monsoon to a warming climate remains much more uncertain [see a recent review by Pascale et al. (2019)].
Acknowledgments
The authors wish to acknowledge three anonymous reviewers for their valuable comments and suggestions, and the modeling groups for sharing the model data. Dr. Chao He wishes to thank Prof. Masahiro Watanabe for offering the code of LBM. This work was supported by the National Key Research and Development Program of China (2017YFA0604601) and Hong Kong RGC General Research Fund (11335316).
Data availability statement
All the data adopted in this work can be accessed online. The CMIP5 and CMIP6 model data used in this study were accessed at the official website of the Earth System Grid Federation (https://esgf-node.llnl.gov), and the observational data were accessed at the official website of NOAA (
REFERENCES
Adams, D. K., and A. C. Comrie, 1997: The North American monsoon. Bull. Amer. Meteor. Soc., 78, 2197–2213, https://doi.org/10.1175/1520-0477(1997)078<2197:TNAM>2.0.CO;2.
Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4, 1147–1167, https://doi.org/10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.
Akinsanola, A. A., and W. Zhou, 2019a: Projections of West African summer monsoon rainfall extremes from two CORDEX models. Climate Dyn., 52, 2017–2028, https://doi.org/10.1007/s00382-018-4238-8.
Akinsanola, A. A., and W. Zhou, 2019b: Dynamic and thermodynamic factors controlling increasing summer monsoon rainfall over the West African Sahel. Climate Dyn., 52, 4501–4514, https://doi.org/10.1007/s00382-018-4394-x.
Bala, G., K. Caldeira, and R. Nemani, 2010: Fast versus slow response in climate change: Implications for the global hydrological cycle. Climate Dyn., 35, 423–434, https://doi.org/10.1007/s00382-009-0583-y.
Bony, S., G. Bellon, D. Klocke, S. Sherwood, S. Fermepin, and S. Denvil, 2013: Robust direct effect of carbon dioxide on tropical circulation and regional precipitation. Nat. Geosci., 6, 447–451, https://doi.org/10.1038/ngeo1799.
Brient, F., 2020: Reducing uncertainties in climate projections with emergent constraints: Concepts, examples and prospects. Adv. Atmos. Sci., 37 (1), 1–15, https://doi.org/10.1007/s00376-019-9140-8.
Ceppi, P., G. Zappa, T. G. Shepherd, and J. M. Gregory, 2018: Fast and slow components of the extratropical atmospheric circulation response to CO2 forcing. J. Climate, 31, 1091–1105, https://doi.org/10.1175/JCLI-D-17-0323.1.
Chadwick, R., 2016: Which aspects of CO2 forcing and SST warming cause most uncertainty in projections of tropical rainfall change over land and ocean? J. Climate, 29, 2493–2509, https://doi.org/10.1175/JCLI-D-15-0777.1.
Chadwick, R., D. Ackerley, T. Ogura, and D. Dommenget, 2019: Separating the influences of land warming, the direct CO2 effect, the plant physiological effect, and SST warming on regional precipitation changes. J. Geophys. Res. Atmos., 124, 624–640, https://doi.org/10.1029/2018JD029423.
Chen, X., T. Zhou, and Z. Guo, 2014: Climate sensitivities of two versions of FGOALS model to idealized radiative forcing. Sci. China Earth Sci., 57, 1363–1373, https://doi.org/10.1007/s11430-013-4692-4.
Chen, Z., T. Zhou, L. Zhang, X. Chen, W. Zhang, and J. Jiang, 2020: Global land monsoon precipitation changes in CMIP6 projections. Geophys. Res. Lett., 47, e2019GL086902, https://doi.org/10.1029/2019GL086902.
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, https://doi.org/10.1175/2008JCLI2471.1.
Collins, M., and Coauthors, 2010: The impact of global warming on the tropical Pacific Ocean and El Niño. Nat. Geosci., 3, 391–397, https://doi.org/10.1038/ngeo868.
Colorado-Ruiz, G., T. Cavazos, J. A. Salinas, P. De Grau, and R. Ayala, 2018: Climate change projections from Coupled Model Intercomparison Project phase 5 multi-model weighted ensembles for Mexico, the North American monsoon, and the mid-summer drought region. Int. J. Climatol., 38, 5699–5716, https://doi.org/10.1002/joc.5773.
Deser, C., R. Knutti, S. Solomon, and A. S. Phillips, 2012: Communication of the role of natural variability in future North American climate. Nat. Climate Change, 2, 775–779, https://doi.org/10.1038/nclimate1562.
Endo, H., and A. Kitoh, 2014: Thermodynamic and dynamic effects on regional monsoon rainfall changes in a warmer climate. Geophys. Res. Lett., 41, 2013GL059158, https://doi.org/10.1002/2013GL059158.
Endo, H., A. Kitoh, and H. Ueda, 2018: A unique feature of the Asian summer monsoon response to global warming: The role of different land–sea thermal contrast change between the lower and upper troposphere. SOLA, 14, 57–63, https://doi.org/10.2151/SOLA.2018-010.
Eyring, V., S. Bony, G. A. Meehl, C. A. Senior, B. Stevens, R. J. Stouffer, and K. E. Taylor, 2016: Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev., 9, 1937–1958, https://doi.org/10.5194/gmd-9-1937-2016.
Eyring, V., and Coauthors, 2019: Taking climate model evaluation to the next level. Nat. Climate Change, 9, 102–110, https://doi.org/10.1038/s41558-018-0355-y.
Gadgil, S., and K. R. Kumar, 2006: The Asian monsoon—Agriculture and economy. The Asian Monsoon, 1st ed. B. Wang, Ed., Springer, 651–683.
Giannini, A., 2010: Mechanisms of climate change in the semiarid African Sahel: The local view. J. Climate, 23, 743–756, https://doi.org/10.1175/2009JCLI3123.1.
Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447–462, https://doi.org/10.1002/qj.49710644905.
Gleckler, P. J., K. E. Taylor, and C. Doutriaux, 2008: Performance metrics for climate models. J. Geophys. Res., 113, D06104, https://doi.org/10.1029/2007JD008972.
Han, Z., T. Su, B. Huang, T. Feng, S. Qu, and G. Feng, 2019: Changes in global monsoon precipitation and the related dynamic and thermodynamic mechanisms in recent decades. Int. J. Climatol., 39, 1490–1503, https://doi.org/10.1002/joc.5896.
He, C., and T. Li, 2019: Does global warming amplify interannual climate variability? Climate Dyn., 52, 2667–2684, https://doi.org/10.1007/s00382-018-4286-0.
He, C., T. Zhou, and T. Li, 2019: Weakened anomalous western North Pacific anticyclone during an El Niño–decaying summer under a warmer climate: Dominant role of the weakened impact of the tropical Indian Ocean on the atmosphere. J. Climate, 32, 213–230, https://doi.org/10.1175/JCLI-D-18-0033.1.
He, J., and B. J. Soden, 2016: Does the lack of coupling in SST-forced atmosphere-only models limit their usefulness for climate change studies? J. Climate, 29, 4317–4325, https://doi.org/10.1175/JCLI-D-14-00597.1.
Held, I. M., and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19, 5686–5699, https://doi.org/10.1175/JCLI3990.1.
Hill, S. A., 2019: Theories for past and future monsoon rainfall changes. Curr. Climate Change Rep., 5, 160–171, https://doi.org/10.1007/s40641-019-00137-8.
Hoell, A., C. Funk, M. Barlow, and S. Shukla, 2016: Recent and possible future variations in the North American monsoon. The Monsoons and Climate Change: Observations and Modeling, L. M. V. de Carvalho, and C. Jones, Eds., Springer International Publishing, 149–162.
Hsu, P.-C., T. Li, J.-J. Luo, H. Murakami, A. Kitoh, and M. Zhao, 2012: Increase of global monsoon area and precipitation under global warming: A robust signal? Geophys. Res. Lett., 39, L06701, https://doi.org/10.1029/2012GL051037.
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, https://doi.org/10.1038/ngeo1792.
Hurrell, J. W., J. J. Hack, D. Shea, J. M. Caron, and J. Rosinski, 2008: A new sea surface temperature and sea ice boundary dataset for the Community Atmosphere Model. J. Climate, 21, 5145–5153, https://doi.org/10.1175/2008JCLI2292.1.
Johnson, N. C., and S.-P. Xie, 2010: Changes in the sea surface temperature threshold for tropical convection. Nat. Geosci., 3, 842–845, https://doi.org/10.1038/ngeo1008.
Johnson, N. C., L. Krishnamurthy, A. T. Wittenberg, B. Xiang, G. A. Vecchi, S. B. Kapnick, and S. Pascale, 2020: The impact of sea surface temperature biases on North American precipitation in a high-resolution climate model. J. Climate, 33, 2427–2447, https://doi.org/10.1175/JCLI-D-19-0417.1.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1644, https://doi.org/10.1175/BAMS-83-11-1631.
Kitoh, A., H. Endo, K. Krishna Kumar, I. F. A. Cavalcanti, P. Goswami, and T. Zhou, 2013: Monsoons in a changing world: A regional perspective in a global context. J. Geophys. Res. Atmos., 118, 3053–3065, https://doi.org/10.1002/JGRD.50258.
Kosaka, Y., and S.-P. Xie, 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403–407, https://doi.org/10.1038/nature12534.
Lee, J.-Y., and B. Wang, 2014: Future change of global monsoon in the CMIP5. Climate Dyn., 42, 101–119, https://doi.org/10.1007/s00382-012-1564-0.
Li, G., S.-P. Xie, Y. Du, and Y. Luo, 2016: Effects of excessive equatorial cold tongue bias on the projections of tropical Pacific climate change. Part I: The warming pattern in CMIP5 multi-model ensemble. Climate Dyn., 47, 3817–3831, https://doi.org/10.1007/s00382-016-3043-5.
Li, T., L. Zhang, and H. Murakami, 2015: Strengthening of the Walker circulation under global warming in an aqua-planet general circulation model simulation. Adv. Atmos. Sci., 32, 1473–1480, https://doi.org/10.1007/s00376-015-5033-7.
Li, Y., Y. Ding, and W. Li, 2017: Interdecadal variability of the Afro-Asian summer monsoon system. Adv. Atmos. Sci., 34, 833–846, https://doi.org/10.1007/s00376-017-6247-7.
Long, S.-M., S.-P. Xie, X.-T. Zheng, and Q. Liu, 2014: Fast and slow responses to global warming: Sea surface temperature and precipitation patterns. J. Climate, 27, 285–299, https://doi.org/10.1175/JCLI-D-13-00297.1.
Luo, Y., J. Lu, F. Liu, and W. Liu, 2015: Understanding the El Niño–like oceanic response in the tropical Pacific to global warming. Climate Dyn., 45, 1945–1964, https://doi.org/10.1007/s00382-014-2448-2.
Mall, R. K., R. Singh, A. Gupta, G. Srinivasan, and L. S. Rathore, 2006: Impact of climate change on Indian agriculture: A review. Climatic Change, 78, 445–478, https://doi.org/10.1007/s10584-005-9042-x.
Maloney, E. D., and Coauthors, 2014: North American climate in CMIP5 experiments: Part III: Assessment of twenty-first-century projections. J. Climate, 27, 2230–2270, https://doi.org/10.1175/JCLI-D-13-00273.1.
Martinez, C., L. Goddard, Y. Kushnir, and M. Ting, 2019: Seasonal climatology and dynamical mechanisms of rainfall in the Caribbean. Climate Dyn., 53, 825–846, https://doi.org/10.1007/s00382-019-04616-4.
Merlis, T. M., 2015: Direct weakening of tropical circulations from masked CO2 radiative forcing. Proc. Natl. Acad. Sci. USA, 112, 13 167–13 171, https://doi.org/10.1073/pnas.1508268112.
Neale, R. B., and Coauthors, 2012: Description of the NCAR Community Atmosphere Model (CAM 5.0). NCAR Tech. Note NCAR/TN-486+STR, 274 pp., www.cesm.ucar.edu/models/cesm1.0/cam/docs/description/cam5_desc.pdf.
Ni, Y., and P.-C. Hsu, 2018: Inter-annual variability of global monsoon precipitation in present-day and future warming scenarios based on 33 Coupled Model Intercomparison Project phase 5 models. Int. J. Climatol., 38, 4875–4890, https://doi.org/10.1002/joc.5704.
O’Neill, B. C., and Coauthors, 2016: The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6. Geosci. Model Dev., 9, 3461–3482, https://doi.org/10.5194/gmd-9-3461-2016.
Pascale, S., W. R. Boos, S. Bordoni, T. L. Delworth, S. B. Kapnick, H. Murakami, G. A. Vecchi, and W. Zhang, 2017: Weakening of the North American monsoon with global warming. Nat. Climate Change, 7, 806–812, https://doi.org/10.1038/nclimate3412.
Pascale, S., L. M. V. Carvalho, D. K. Adams, C. L. Castro, and I. F. A. Cavalcanti, 2019: Current and future variations of the monsoons of the Americas in a warming climate. Curr. Climate Change Rep., 5, 125–144, https://doi.org/10.1007/s40641-019-00135-w.
Rauscher, S. A., F. Kucharski, and D. B. Enfield, 2011: The role of regional SST warming variations in the drying of Meso-America in future climate projections. J. Climate, 24, 2003–2016, https://doi.org/10.1175/2010JCLI3536.1.
Schneider, U., A. Becker, P. Finger, A. Meyer-Christoffer, B. Rudolf, and M. Ziese, 2011: GPCC full data reanalysis version 6.0 at 0.5°: Monthly land-surface precipitation from rain-gauges built on GTS-based and historic data, accessed 21 May 2020, https://doi.org.10.5676/DWD_GPCC/FD_M_V7_050.
Seager, R., M. Cane, N. Henderson, D.-E. Lee, R. Abernathey, and H. Zhang, 2019: Strengthening tropical Pacific zonal sea surface temperature gradient consistent with rising greenhouse gases. Nat. Climate Change, 9, 517–522, https://doi.org/10.1038/s41558-019-0505-x.
Seth, A., S. A. Rauscher, M. Rojas, A. Giannini, and S. J. Camargo, 2011: Enhanced spring convective barrier for monsoons in a warmer world? Climatic Change, 104, 403–414, https://doi.org/10.1007/s10584-010-9973-8.
Seth, A., A. Giannini, M. Rojas, S. A. Rauscher, S. Bordoni, D. Singh, and S. J. Camargo, 2019: Monsoon responses to climate changes—Connecting past, present and future. Curr. Climate Change Rep., 5, 63–79, https://doi.org/10.1007/s40641-019-00125-y.
Shaw, T. A., and A. Voigt, 2015: Tug of war on summertime circulation between radiative forcing and sea surface warming. Nat. Geosci., 8, 560–566, https://doi.org/10.1038/ngeo2449.
Stager, J. C., D. B. Ryves, B. M. Chase, and F. S. R. Pausata, 2011: Catastrophic drought in the Afro-Asian monsoon region during Heinrich event 1. Science, 331, 1299–1302, https://doi.org/10.1126/science.1198322.
Thompson, D. W. J., E. A. Barnes, C. Deser, W. E. Foust, and A. S. Phillips, 2015: Quantifying the role of internal climate variability in future climate trends. J. Climate, 28, 6443–6456, https://doi.org/10.1175/JCLI-D-14-00830.1.
Trenberth, K. E., D. P. Stepaniak, and J. M. Caron, 2000: The global monsoon as seen through the divergent atmospheric circulation. J. Climate, 13, 3969–3993, https://doi.org/10.1175/1520-0442(2000)013<3969:TGMAST>2.0.CO;2.
van Vuuren, D., and Coauthors, 2011: The representative concentration pathways: An overview. Climatic Change, 109, 5–31, https://doi.org/10.1007/s10584-011-0148-z.
Vecchi, G. A., and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20, 4316–4340, https://doi.org/10.1175/JCLI4258.1.
Wang, B., and Q. H. Ding, 2006: Changes in global monsoon precipitation over the past 56 years. Geophys. Res. Lett., 33, L06711, https://doi.org/10.1029/2005GL025347.
Wang, B., and Q. H. Ding, 2008: Global monsoon: Dominant mode of annual variation in the tropics. Dyn. Atmos. Oceans, 44, 165–183, https://doi.org/10.1016/j.dynatmoce.2007.05.002.
Wang, B., and Coauthors, 2020a: Monsoons climate change assessment. Bull. Amer. Meteor. Soc., https://doi.org/10.1175/BAMS-D-19-0335.1, in press.
Wang, B., C. Jin, and J. Liu, 2020b: Understanding future change of global monsoons projected by CMIP6 models. J. Climate, 33, 6471–6489, https://doi.org/10.1175/JCLI-D-19-0993.1.
Wang, C., 2007: Variability of the Caribbean low-level jet and its relations to climate. Climate Dyn., 29, 411–422, https://doi.org/10.1007/s00382-007-0243-z.
Watanabe, M., and M. Kimoto, 2000: Atmosphere–ocean thermal coupling in the North Atlantic: A positive feedback. Quart. J. Roy. Meteor. Soc., 126, 3343–3369, https://doi.org/10.1002/qj.49712657017.
Webb, M. J., and Coauthors, 2017: The Cloud Feedback Model Intercomparison Project (CFMIP) contribution to CMIP6. Geosci. Model Dev., 10, 359–384, https://doi.org/10.5194/gmd-10-359-2017.
Wu, B., T. J. Zhou, and T. Li, 2009: Contrast of rainfall–SST relationships in the western North Pacific between the ENSO-developing and ENSO-decaying summers. J. Climate, 22, 4398–4405, https://doi.org/10.1175/2009JCLI2648.1.
Wu, B., T. Zhou, and T. Li, 2017: Atmospheric dynamic and thermodynamic processes driving the western North Pacific anomalous anticyclone during El Niño. Part I: Maintenance mechanisms. J. Climate, 30, 9621–9635, https://doi.org/10.1175/JCLI-D-16-0489.1.
Xiang, B., B. Wang, W. Yu, and S. Xu, 2013: How can anomalous western North Pacific subtropical high intensify in late summer? Geophys. Res. Lett., 40, 2349–2354, https://doi.org/10.1002/grl.50431.
Xie, S.-P., C. Deser, G. A. Vecchi, J. Ma, H. Y. Teng, and A. T. Wittenberg, 2010: Global warming pattern formation: Sea surface temperature and rainfall. J. Climate, 23, 966–986, https://doi.org/10.1175/2009JCLI3329.1.
Xie, S.-P., and Coauthors, 2015: Towards predictive understanding of regional climate change. Nat. Climate Change, 5, 921–930, https://doi.org/10.1038/nclimate2689.
Yanai, M., and T. Tomita, 1998: Seasonal and interannual variability of atmospheric heat sources and moisture sinks as determined from NCEP–NCAR reanalysis. J. Climate, 11, 463–482, https://doi.org/10.1175/1520-0442(1998)011<0463:SAIVOA>2.0.CO;2.
Zhang, L., and T. Zhou, 2011: An assessment of monsoon precipitation changes during 1901–2001. Climate Dyn., 37, 279–296, https://doi.org/10.1007/s00382-011-0993-5.
Zhou, T. J., L. X. Zhang, and H. M. Li, 2008: Changes in global land monsoon area and total rainfall accumulation over the last half century. Geophys. Res. Lett., 35, L16707, https://doi.org/10.1029/2008GL034881.
