1.4.2.3 Monsoons

Monsoon circulations are large-scale seasonal changes in the direction and strength of prevailing winds driven by insolation (incoming solar radiation) and local temperature differences between land and ocean. Their dynamics are strongly influenced by the seasonal migration of the Intertropical Convergence Zone (ITCZ), the regional band in the tropics where the trade winds from the northern and southern hemisphere converge and rise as part of the tropical atmospheric overturning circulation (see Figure 1.4.1). The term ‘monsoon’ was historically associated with summer precipitation over South Asia; however, monsoon systems affect other parts of the globe such as East Asia, Africa, Australia and the Americas. 

Historically, monsoons were seen as large-scale sea breeze circulations driven by land-sea heating differences due to seasonal changes in incoming solar radiation (Figure 1.4.9). Currently, a perspective of a global monsoon has emerged (Trenberth et al., 2000; Wang & Ding, 2008), where the monsoon systems are seen as interconnected and driven by localised seasonal and more extreme migrations of the ITCZ (Gadgil, 2018; Geen et al., 2020, and references within). Monsoon regions in the world experience heavy precipitation in the summer months, and the global monsoon system is an integral part of the global hydrological cycle, contributing ~31 per cent of total precipitation over the globe (Wang and Ding, 2008). 

There is a recent intensification trend in global monsoon precipitation, mainly due to enhanced northern hemisphere summer monsoon (Wang et al., 2012). It will likely continue in the future (high confidence, IPCC (2021), by ~1-3% per °C warming) because of increased water vapour related to warming driven by increased CO₂ in the atmosphere (Hsu et al., 2013; Lee and Wang, 2014; Chen et al., 2020; Ha et al., 2020; Wang et al., 2020), although a few studies conversely show that climate warming may lead to a weakened global monsoon circulation (Hsu et al., 2012, 2013). Climate simulations also project expansion of global monsoon domain areas with increasing CO₂ (Wang et al., 2020; Paik et al., 2023) and increased frequency of monsoon precipitation extremes in the 21st Century (Chevuturi et al., 2018; Ali et al., 2019; Ha et al., 2020; Katzenberger et al., 2021). 

Monsoon precipitation is vital for agrarian populations and livelihoods in vast areas of South Asia, Africa and South America, and changes to it could expose almost two thirds of the global population to disastrous effects (Wang et al,. 2021). Hence it is crucial to understand the dynamics and potential nonlinear changes or tipping behaviour of monsoon systems under a changing climate. Here the ‘tipping’ of monsoon systems refers to a significant, feedback-driven shift in the precipitation state of the monsoon, with implications for the regional and global climate and ecosystems. In this discussion we assess if the major regional monsoon systems (West African, Indian and South American) show any evidence of nonlinear (tipping or abrupt) responses to climate forcings based on available literature. 

Indian summer monsoon (ISM)

During the summer season over South Asia (June-September), winds from the south west carry large amounts of water vapour from the Indian Ocean to the Indian subcontinent and cause heavy precipitation in the region, providing ~80 per cent of the total annual precipitation (Figure 1.4.9). ISM precipitation shows considerable intra-seasonal, interannual and decadal variability, many times with precipitation extremes (leading to droughts, floods) during the season, and years and decades with above and below (in drought years) normal precipitation. Indian monsoon variability is strongly influenced by ocean-atmosphere interactions such as El Niño Southern Oscillation (ENSO, see Chapter 1.4.2.5), Indian Ocean Dipole events (irregular changes in the temperature gradients in the Indian Ocean, Cherchi et al., 2021; Chaudhary et al., 2021; Hrudya et al., 2021), and cooler temperatures in the North Atlantic (Borah et al., 2020). 

ISM precipitation declined in the second half of the 20th Century, attributed mainly to human-driven aerosol loading (Bollasina et al., 2011) and strong Indian Ocean warming (Roxy et al., 2015). Recent studies (Jin and Wang, 2017) suggest it has revived since 2002, linked to enhanced warming over the Indian subcontinent due to reduced low clouds, resulting in an increased land-ocean thermal gradient. Future projections suggest increases in the ISM precipitation in future warming scenarios (by 5.3% per celsius of global warming, according to CMIP6 models, Katzenberger et al., 2021) and a longer monsoon duration (Ha et al., 2020).

Evidence for tipping dynamics

Many periods of abrupt ISM transitions have been identified in past monsoon records in association with high-latitude climate events (Schulz et al., 1998; Morrill et al., 2003) such as during Heinrich events (glacial outbursts that temporarily shut down the AMOC – see 1.4.2.1) (McManus et al., 2004; Stager et al., 2011), the Younger Dryas (a temporary return to more intense glacial conditions 12,900-11,700 years ago, (Cai et al., 2008; Carlson et al., 2013), and several periods during the more recent Holocene (Gupta et al., 2003; Berkelhammer et al., 2012; Yan and Liu, 2019). However, the mechanisms of such abrupt transitions are not clearly understood. Efforts have been made to identify any Indian monsoon tipping mechanisms using simplified models (Zickfeld et. al., 2005; Levermann et al., 2009). 

An internal feedback mechanism, a ‘positive moisture advection feedback’ (Zickfeld et. al., 2005; Levermann et al., 2009; Schewe et al., 2012), has been suggested as responsible for abrupt transitions simulated using these analytical models. In this feedback, the atmospheric temperature gradient between the land and cooler ocean in summer leads to the onshore transport of moist air (advection), which then rises, forms clouds and condenses into rain. The phase transition from vapour to liquid warms the surrounding air (through the release of latent heat, or ‘diabatic heating’), increasing the land-ocean temperature gradient and sustaining this monsoon circulation. Any forcing that weakens this pressure gradient can therefore lead to monsoon destabilisation (Zickfeld et al., 2005). If monsoon winds weaken, advection and condensation reduce, and the threshold for a monsoon tipping is reached when the diabatic heating fails to balance the heat advection away from the region (Levermann et al., 2009). 

Contrarily, follow-up studies (Boos and Storelvmo, 2016) challenge occurence of any tipping in these simplified models, and rule out any abrupt monsoon responses to human-driven forcings in the future, and instead attribute past monsoon shifts to rapid forcings or vegetation feedbacks. Simplified models omit key aspects and feedbacks in the monsoon system (specifically, static stability of the troposphere in the models that simulated the monsoon tipping, Boos and Storelvmo, 2016; Kumar and Seshadri, 2022). Hence, more studies using models that represent the complexities of the monsoon and palaeoclimate data are required for a clearer picture on any non-linear changes in the monsoon system.

Apart from climate change, aerosols pose another significant human-driven pressure on the Earth system. Aerosols influence the Earth’s radiative budget, climate and hydrological cycle by reflecting or absorbing solar radiation, changing the optical properties of clouds, and also by acting as cloud condensation nuclei. An increase in anthropogenic aerosols has been attributed as the major reason for the decline of Northern Hemispheric summer monsoon strength from the 1950s to 1980s (Cao et al., 2022), due to its dimming effect. 

A large increase in regional aerosol loading over South and East Asia (>0.25 Aerosol Optical Depth, AOD, Steffen et al., 2015) could potentially switch the Asian regional monsoon systems to a drier state. Further, hemispheric asymmetries in the aerosol loading (>0.15 AOD, Rockström et al., 2023), due to volcanic eruptions, human sources or intentional geoengineering, could lead to hemispheric temperature asymmetries and changes in the location of the ITCZ, significantly disrupting regional monsoons over West Africa and South Asia (Haywood et al., 2013; Rockström et al., 2023; Richardson et al., 2023). However, there is no direct evidence of aerosols causing a tipping of the monsoon systems, and uncertainties in threshold estimates are large due to complex aerosol microphysics and aerosol-cloud interactions. Hence, systematic observational and modelling approaches would be needed to reduce the uncertainties, as well as additional assessments of interhemispheric asymmetries in the aerosol distribution.

Assessment and knowledge gaps

The ISM system was earlier classified as one of the Earth’s tipping systems (Lenton et al., 2008), based on the threshold behaviour of the monsoon in the past and the moisture-advection feedback (Levermann et al., 2009), but this was refuted by later studies (Boos and Storelvmo, 2016; Seshadri, 2017). Most recently, Armstrong McKay et al,. (2022) categorise ISM as an “uncertain potential [climate] tipping element” as global warming is not likely to cause tipping behaviour directly in ISM precipitation. 

Based on this current literature, the chances for ISM exhibiting a tipping behaviour towards a new low-precipitation state under climate change are uncertain, warranting extensive studies on the subject. However, potential tipping behaviour in the AMOC (see Chapter 1.4.2.1, 1.5.2.5, and relation to global monsoon described in West African monsoon below) or increase in the interhemispheric asymmetry of aerosol loading in the atmosphere beyond potential threshold levels could lead to large disruptions to monsoon systems. This could cause calamitous effects on millions of people in the monsoon regions, even in the absence of tipping.

West African monsoon (WAM)

The West African monsoon (WAM) controls hydroclimatic conditions, vegetation and mineral-dust emissions of northern tropical and subtropical Africa, up to the dry Sahel region at the southern edge of the Sahara Desert (Figure 1.4.11). The strength of the monsoon shows large variations over a range of timescales from interannual to decadal and longer. Albedo (reflectivity of the Earth’s surface) changes caused by human-driven land-cover changes and desertification (Otterman, 1974; Charney et al., 1975; Charney, 1975) can affect rainfall: a less vegetated surface with higher albedo increases radiative loss, thereby reducing temperature and suppressing the rising and condensation of moist air into rainfall (i.e. convective precipitation). Variations of sea surface temperatures (SSTs) in different oceanic basins can also drive interannual and decadal variability in WAM precipitation (Rodríguez-Fonseca et al., 2015). Other major factors that affect WAM variability are land surface variability such as variations in soil moisture (Giannini et al., 2013; Zeng et al., 1999), vegetation (Charney et al., 1975;Kucharski et al., 2013; Otterman, 1974; Wang et al., 2004; Xue, 1997), high-latitude cooling (Collins et al., 2017) and dust emissions (Konare et al., 2008; Solmon et al., 2008; Zhao et al., 2011).

Evidence for tipping dynamics

Palaeoclimate records underscore dramatic variations of the WAM in the more distant past, such as the periodic expansion of vegetation into the Sahara Desert during the so-called ‘African humid periods’ (AHPs) and linked to the emergence of ancient cultures along the Nile. Another example is the drought 200-300 years ago, which caused the water level of Lake Bosumtwi in Ghana to fall by almost four times as much as it did during the drought of the 1970s and 1980s. Large past variations of the WAM, such as those during the AHPs, raise the question of whether present-day anthropogenic global warming could have potentially significant impacts on the WAM. Although the nature and magnitude of radiative forcing were different during the AHPs than they are now (i.e. an external change in insolation due to orbital forcing versus an internal change from increased greenhouse gases), the fact that the AHPs occurred under a globally warmer climate than the pre-industrial period invites questions.

Some palaeoclimate archives show WAM precipitation changes that took place over several centuries (deMenocal et al., 2000; McGee et al., 2013), i.e. an order of magnitude faster than the orbital forcing. However, others show a much more gradual change (e.g. Kröpelin et al., 2008) with a time-varying withdrawal of the WAM from North to South following the insolation changes (Shanahan et al., 2015). Because of geographic variability of the African landscape and African monsoon circulation, abrupt changes can occur in several, but not all, regions at different times during the transition from the humid to arid climate (Dallmeyer et al., 2021). 

By inducing latitudinal movements of the ITCZ, change in the AMOC is considered to play a role in shifts of global monsoon systems. Palaeoclimate evidence suggests that glacial meltwater-induced weakening of the AMOC during Heinrich events in the last glacial period led to abrupt Asian and African monsoon weakening (Mohtadi et al., 2014; Mohtadi et al., 2016). Similarly, the Younger Dryas led to a cool and dry state over Northern Hemisphere tropical monsoon regions. North Atlantic fresh water-hosing simulations using climate models (Lewis et al., 2010; Pausata et al., 2011; Kageyama et al., 2013) confirm these shifts in ITCZ can occur as a result of substantial glacial meltwater release. These influences of AMOC on the monsoon systems have also been studied in the context of the South American monsoon (see below). Hence, a collapse of AMOC (see Chapter 1.4.2.1) has the potential to cause disruptions to the regional monsoon systems and other tropical precipitation systems over Asia, Africa and South America (Gupta et al., 2003; IPCC, 2021). 

Assessment and knowledge gaps

Abrupt changes in one region can be induced by abrupt changes in others, a process sometimes referred to as ‘induced tipping’. The AHP transition of the Sahara was slow with respect to timescales of individual humans and local ecosystems, but regionally rapid with respect to changes in the driver. Based on the record of large past variations of WAM precipitation patterns (including collapse), and the existence of positive amplifying feedbacks, we classify WAM as a tipping system with low confidence. This is in line with previous assessments (Armstrong McKay et al., 2022), in which a lower tipping threshold of 2°C global warming was estimated but attributed low confidence due to limited model resolution of vegetation shifts, and model disagreements in future trends. The timescale of abrupt shifts is estimated to range from decades as observed in CMIP5 models (Drijfhout et al., 2015) to centuries based on palaeorecords (Hopcroft and Valdes, 2021; Shanahan et al.; 2015). Potential additional destabilisation through AMOC weakening and atmospheric aerosol loading, and the far-reaching implications of WAM tipping, call for intensified research efforts on this system.

South American monsoon (SAM)

The South American monsoon (SAM) system is characterised by strong seasonality in precipitation, even though it does not show a reversal of low-level winds like in the Asian monsoon (Zhou and Lau, 1998; Vera et al., 2006; Liebmann and Mechoso, 2011; Carvalho et al., 2012). Studies are relatively few compared to the Asian and African monsoon systems, as it was not classified as a monsoon system until a couple of decades ago (Zhou and Lau, 1998). 

A mature SAM system (from December to February) shows features such as enhanced northeastern trade winds, increased land-ocean thermal gradient and the development of an active convective zone (the South Atlantic Convergence Zone) (Figure 1.4.12; Zhou and Lau, 1998). The SAM system affects vast areas of tropical South America all the way to southern Brazil, and provides more than 50 per cent of the annual precipitation to these regions (Vera et al., 2006) including most of the Amazon rainforest. SAM precipitation varies from interannual to orbital timescales (Chiessi et al., 2009; Liebmann and Mechoso, 2011; Carvalho and Cavalcanti, 2016; Hou et al., 2020).

The influence of anthropogenic climate change on the SAM precipitation is ambiguous (Douville et al., 2021), and many CMIP5/CMIP6 models are noted for their poor representation of SAM precipitation (Jones and Carvalho, 2013; Douville et al., 2021). IPCC AR6 finds high confidence in delayed onset of the SAM precipitation since the 1970s associated with climate change, which could worsen with increased CO₂ levels (Douville et al., 2021). However, the projected future change in total SAM precipitation is uncertain, as the models show low agreement on the projections (Douville et al., 2021). 

Evidence for tipping dynamics

Orbital timescale changes (i.e. over tens of thousands of years) in SAM precipitation seem to be largely controlled by changes in insolation and respond linearly to it (Cruz et al., 2005; Hou et al., 2020). Millennial-scale changes (i.e. over thousands of years) in the SAM are thought to be associated with variations in strength of the AMOC, as described for the West African monsoon above. In particular, palaeo evidence indicates that an increase in South American precipitation to the south of the equator followed weakening of the AMOC related to Heinrich events (Mulitza et al., 2017; Campos et al., 2019). Similarly, meltwater flux from the Laurentide Ice Sheet during the Younger Dryas may have led to a warm and wet state over tropical South America to the south of the equator (McManus et al., 2004; Broecker et al., 2010; Venancio et al., 2020; Brovkin et al., 2021). Earth system model projections of AMOC collapse impacts on the tropical rainfall in South America are model-dependent, but generally find a reduction in rainfall over northern South America and an increase over the southern Amazon (Bellomo et al., 2023; Nian et al., 2023; Orihuela-Pinto et al., 2022; Liu et al., 2020: see 1.5.2.4).

Further, deforestation over 30-50 per cent of the Amazon rainforest led to a tipping point in the SAM system in one model (Boers et al., 2017), causing precipitation reductions of up to 40 per cent in non-forested parts of the western Amazon. This reduction is caused by the breakdown of a positive amplifying feedback mechanism that involves latent heat of condensation over the Amazon rainforest due to transpiration (i.e. water lost from plants) and water vapour transport from the Atlantic. Reduced transpiration due to deforestation can no longer sufficiently provide water vapour to sustain the latent heat required, thereby reducing the inflow of oceanic water vapour, and leading to a monsoon tipping in this model (Boers et al., 2017). (see 1.3.2.1 for more on Amazon dieback)

Assessment and knowledge gaps

A combination of climate change and deforestation could lead to substantial changes in the SAM system, affecting many millions of people. Additionally, a decrease in AMOC strength could potentially trigger major changes in tropical South American precipitation (see 1.5.2.4). However, the current scarcity of research in the subject limits our ability to fully understand and assess the tipping potential of the system, and we classify the possibility of SAM tipping to be uncertain.