The South Asian and East Asian monsoon systems are of key importance for the region's economy, agriculture, and industry. The monsoons bring the majority of the annual rainfall over the summer months. The seasonality of rainfall is a major controlling factor for agricultural activities and for natural ecosystems. Significant deviations from the normal onset dates and amount of summer rainfall associated with the monsoon have far-reaching consequences, especially for farmers . Crop production, which contributes substantially to the regions gross domestic product, is particularly adversely affected during deficit rainfall years . The occurrence of flooding or droughts , typically linked to strong or weak monsoon years, also has widespread socio-economic impacts and is expected to become more frequent in the future .

There remains considerable uncertainty in how the monsoons will evolve in a changing climate, both in terms of the response of the monsoons to an applied forcing and in the nature of future climate forcing itself. Despite many improvements to state-of-the-art global climate models (GCMs) , difficulties persist in accurately portraying the physical processes and interactions associated with the South Asian and East Asian monsoon systems . Variability between models remains an issue, with large inter-model spreads reducing the confidence in future climate projections. The multi-model mean usually performs better than any single model . Systematic biases in GCMs typically arise from poor representations of orographic rainfall and the various oceanic phenomena affecting sea surface temperature . Precipitation over northern India and the Yangtze River valley is underestimated, while precipitation on the leeward side of the Western Ghats and the mountainous eastern Indian states is overestimated . Biases in precipitation are linked with biases in the moisture transport; models typically have too much moisture over the southern Indian peninsula and too little moisture over central India . The role of irrigation is largely ignored, despite its importance in determining the strength and spatial extent of local-scale precipitation . It is widely accepted that a reasonable representation of the large-scale circulation is required for accurate simulation of precipitation .

In general, it is agreed upon that in the future monsoonal precipitation in the Northern Hemisphere will increase, with more frequent extreme rainfall events, despite a weakening of the large-scale circulation . Changes to forcings, such as greenhouse gases and aerosols, can have significant impacts on local moisture availability and affect the hydrology. In combination with a strong extratropical event, such as prolonged atmospheric blocking or advection of a low-pressure system, this can lead to extreme floods and heat waves . In the near-term future, internal variability is expected to remain the dominant mode in modulating the monsoonal precipitation, but by the end of the 21st century, the effects of external forcing will become prevalent. The main components of external forcing are greenhouse gases and aerosols, which impact the optical properties of the atmosphere over a variety of spectral ranges. Phenomena contributing to the internal variability on annual–decadal timescales include the Indian Ocean Dipole, the El Niño–Southern Oscillation (ENSO), the Pacific decadal oscillation, and the Atlantic multidecadal oscillation. ENSO is the most important contributor on interannual timescales, and it is expected that the precipitation variability due to ENSO will increase . Here, we focus on the South Asian and East Asian monsoon responses to external forcing, specifically to globally increased greenhouse gas concentration and regional absorbing aerosol loading.

In terms of the Asian monsoons, aerosols and greenhouse gases have competing effects. Greenhouse gases act to warm the surface, enhancing the contrast between land and sea temperatures and enabling greater moisture uptake (e.g. ). This leads to stronger monsoons and increased precipitation, despite a weakening of the large-scale circulation . Generally, aerosols have a stabilising effect on the atmosphere through cooling the surface, increasing the stratification of the atmosphere, and causing a drying trend that results in a weaker monsoon . Historically, aerosol forcing has dominated, linked with a declining rainfall trend over the latter half of 20th century , but moving forward, greenhouse gas forcing is expected to dominate, which is associated with a likely increase in monsoonal rainfall in the Northern Hemisphere . In the near term, changes in aerosol concentrations over the Asian region will continue to impact the monsoons, with increases in anthropogenic emissions acting to weaken the large-scale circulation and hydrological cycle, thus weakening the monsoons, while decreases in emissions, perhaps through the enforcement of air quality policies, will likely intensify the monsoons.

Aerosols are often referred to as scattering or absorbing, depending on how the aerosol interacts with shortwave radiation. Scattering aerosols, such as sulfates, cause surface cooling through scattering radiation, which weakens the land–sea temperature gradient. This inhibits moisture advection from the Arabian Sea and reduces low-level southwesterly wind speeds, resulting in decreased precipitation . Absorbing aerosols, the most important of which is black carbon, cause both surface cooling and localised mid-level heating by absorbing some of the incoming radiation. The effect of black carbon on the South Asian and East Asian monsoons is debated. found that while black carbon contributed to increased precipitation during the pre-monsoon period, there was a decrease in monsoonal precipitation. In partial agreement, also found a link between increased black carbon (and dust) loading and pre-monsoon precipitation but that subsequent precipitation decreased over India and increased over East Asia. noted a near-linear relationship between black carbon loading and summer precipitation in the South Asian monsoon, with enhanced monsoonal circulation, stronger meridional tropospheric gradient, and increased precipitable water as the levels of black carbon are raised. Others have found that black carbon weakens the Asian monsoons or that the results are inconclusive . Although suggest that sulfates are the dominant aerosol species for impacting South Asian monsoon rainfall, further work is needed to unify the individual effects of different aerosol species on the vertical profile of the atmosphere and the interactions between them .

The COVID-19 pandemic in 2020 offered an unprecedented opportunity to observe the effects of reduced aerosol loading, resulting from nationwide lockdowns, on the South Asian and East Asian monsoons. The decrease in aerosols was found to correspond to a 4 W m-2 increase in surface solar radiation , resulting in warmer surface temperatures over land. The heightened contrast in land–sea temperature and stronger low-level winds enhanced moisture inflow to the monsoon system, leading to increased precipitation and a more intense monsoon . In particular, spatial coherence between reduced aerosol loading and precipitation has been noted over western and central India and the Yangtze River valley .

The South Asian summer monsoon has been identified as a possible tipping point , highlighting the importance of investigating the impact of changing radiative forcing. Radiation plays a major role in the evolution of the monsoon, with seasonal changes in temperature and circulation triggering monsoon onset. The onset itself can be considered a regime change, with a simplified model able to reproduce the transition from low to high relative humidity as the low-level wind speed is increased . Modelling of the Indian monsoon by suggested that an abrupt transition to a low rainfall regime with a destabilised circulation is plausible given changes to the planetary albedo, the insolation, and/or the carbon dioxide (CO2) concentration. The possibility of an abrupt transition (on intraseasonal timescales) is supported by , who used a similar conceptual model based on heat and moisture conservation to show the bistability of the monsoon system given net radiative influx above a critical threshold. took this conceptual model a step further so that it can be applied on orbital timescales and demonstrate a series of abrupt transitions of the East Asian monsoon that correspond to a proxy record of the penultimate glacial period. In contrast, used a simplified energetic theory in combination with GCM simulations to show that the intensity of summer monsoons in the tropics has a near-linear dependence on a range of radiative forcings. They argue that previous tipping point theory neglected the static stability of the troposphere and that no threshold response to large radiative forcings is observed in a GCM simulation on decadal timescales. also found no abrupt transitions of the South Asian monsoon in GCM simulations with very high black carbon loading despite significant surface cooling and a decrease in global precipitation.

The main goal of this study is to understand the impact of absorbing aerosol forcing on the monsoon in India, Southeast Asia, and eastern China, focusing on the precipitation and the large-scale circulation using a simplified modelling framework. We will investigate the level of radiative forcing required for a breakdown of the monsoon circulation and capture the transition from a high-precipitation regime to a low-precipitation regime. Additionally, we wish to explore the combined effect of regional absorbing aerosol forcing with globally increased carbon dioxide concentration on the South Asian and East Asian monsoons. The so-called Silk Road pattern indicates the presence of westward-propagating signal linking the Indian monsoon with atmospheric patterns over China as a result of Rossby waves . We will also consider whether localised forcings applied in each of the regions of interest have noticeable remote impacts.

We use an intermediate-complexity climate model, the Planet Simulator (PLASIM) , to simulate the response of the South Asian and East Asian monsoons to forcing scenarios with differing levels of carbon dioxide and absorbing aerosols. Details of the experiment set-up and design are given in Sect. . There are several studies using idealised models based on conservation of heat and moisture to investigate the occurrence of regime transitions in the monsoons, and a multitude of literature using complex GCMs to simulate future climate response to different forcings scenarios , but little using intermediate-complexity models . Intermediate-complexity models are able to reproduce large-scale climate features with a fair degree of accuracy, while maintaining the flexibility to perform long simulations at low computational cost and to allow for parametric exploration. However, they are unable to capture small-scale convection or oceanic processes such as ENSO. Our goal is to help bridge the gap between conceptual and complex models. The results of the simulations are presented in Sects.  and , with the former focusing on the response of key variables such as precipitation and circulation to absorbing aerosol forcing, while the latter considers the combined effect of carbon dioxide and absorbing aerosol forcing on the South Asian and East Asian monsoons. Additionally, we consider the relationship between the added forcing and regionally averaged summer precipitation, trying to capture the signature of abrupt transitions. In Sect.  we show the results of applying absorbing aerosol forcing individually to regions of India, Southeast Asia, and eastern China rather than simultaneously (as in Sects.  and ), which allows attribution of responses to local and remote forcing within the Asian region. Furthermore, we can compare the linear combination of response due to local-scale absorbing aerosol forcing to the response from regional-scale forcing in the spirit of linear response theory .

The PLASIM model is an intermediate-complexity global climate model, with a dynamical core and parameterised physical processes. It includes parameterisation schemes for land–surface interactions, radiation, and convection. There are several versions of the PLASIM model, where adaptions have been made to allow inclusion of additional climatic components such as vegetation, sea ice, and ocean . We use the version developed by , which features a large-scale geostrophic ocean component. Vertically there are 10 atmospheric levels, and horizontally the spectral resolution is T42, which approximately corresponds to 2.8∘. When focusing on the large-scale features over long timescales, the PLASIM model is an invaluable tool for investigating various future scenarios at a low computational cost.

The PLASIM model has previously been used to investigate regime transitions on a range of timescales, including the arid-to-hyperarid transition of the Atacama Desert and the effect of a permanent El Niño state on susceptible tipping elements like the Amazon rainforest and the Sahel Belt . PLASIM has also been used to simulate global-scale transitions between a warm (current climate) and a snowball state , which has helped to advance understanding of the multistability properties of the Earth's climate system . In terms of future climate projections, the PLASIM model has been used as an emulator and to support the application of Ruelle's response theory .

There are multiple precedents for using both PLASIM and other models of a similar complexity for climate simulations regarding the Asian monsoons . In terms of lower-complexity models, use a box model of the tropical atmosphere to investigate the stability of the South Asian monsoon to changes in planetary albedo and carbon dioxide concentration. For higher-complexity models, there is an abundance of literature where CMIP5 and CMIP6 standard models are used to explore the response of the South Asian and East Asian monsoons to future climate scenarios . Here, we use PLASIM to quantify the responses of the South Asian and East Asian monsoons to a range of future climate scenarios entailing different levels of CO2 concentration and localised aerosol forcings.

The added absorbing aerosol forcing causes the surface to cool in the regions where the forcing is applied and also causes a warm anomaly around 700 hPa. The increased stratification of the atmosphere and the reduction of the land–sea temperature contrast suppresses precipitation and weakens the large-scale circulation. As the absorbing aerosol forcing is increased, the response of the monsoons becomes more pronounced.

Figures – (and Figs. S2–S7) are presented as panels of two rows by three columns for each variable, with the top-left panel showing the state of the system at approximately 30 W m-2 of forcing. The remaining five panels show the anomaly with respect to the control simulation (aerosol only – control) at forcings of approximately 30, 60, 90, 120, and 150 W m-2. With the exception of 150 W m-2, each panel in the figures is produced using the average of the two 10-year means centred around the respective forcing values. The panel for 150 W m-2 represents the single 10-year period centred on 150 W m-2. Since there is no significant hysteresis in the simulations, the average over both the ascending branch with forcing 0 to 150 W m-2 and the descending branch with forcing 150 to 0 W m-2 is taken. Stippling is added to show statistically significant changes, defined as where the anomaly (aerosol only – control) is greater than double the JJA interannual variability of the aerosol-only simulation. In the following text, we discuss the response of the system by looking at the mean summer (JJA) precipitation, evaporation, surface temperature, and winds at 850 and 200 hPa. Areas of high orography will be masked in grey for certain pressure levels.

The application of the radiative forcing associated with absorbing aerosols leads to surface cooling and mid-level warming where the forcing is applied (Fig. S2). When the intensity of absorbing aerosol forcing reaches approximately 90 W m-2, some areas of India and South East Asia have cooled by 10 ∘C. Above 90 W m-2, the surface temperature drops more rapidly and becomes unrealistically cool at -15 ∘C in parts of eastern China when the forcing is close to 150 W m-2. The intense surface cooling is primarily responsible for activating the ice–albedo effect, which is a positive feedback that enhances surface cooling as sea ice and snow cover increases, causing a greater amount of radiation to be reflected back into space. The result is similar to the establishment of a nuclear winter, albeit via a different mechanism. In our modelling strategy, the effect of absorbing aerosols is emulated by vertically redistributing the heating, rather than removing it from the atmospheric column. At the 700 hPa level (Fig. S2, bottom two rows), a statistically significant warm anomaly develops over eastern China, which becomes stronger as the forcing increases. There is little change in 700 hPa temperature across northern India and Myanmar, with only the southern peninsula areas of India and Southeast Asia showing cool anomalies at high to extreme levels of forcing. The areas showing moderate temperature change at the 700 hPa level, namely northwestern India, eastern India–Myanmar, and eastern China, are also associated with anomalous high pressure (Fig. S3). The monsoon trough, a low-pressure region to the north of India that is associated with the formation of the Indian monsoon, intensifies as the forcing increases beyond 90 W m-2. When the absorbing aerosol forcing is greater than 90 W m-2, a statistically significant cold anomaly develops over the Middle East, creating an east–west temperature dipole. In general, the combination of surface cooling and mid-level warming, clearly seen in vertical cross sections of temperature (Fig. S5), leads to a strong temperature inversion and an increase in the static stability of the atmosphere, which suppresses moist convective processes and thus reduces precipitation.

Figure  (top two rows) illustrates the change in precipitation. The eastern China and Southeast Asia regions are the most affected and feature strong reduction in the precipitation. With the exception of the west coast and northeastern states, i.e. areas of higher orography, the majority of India does not follow the same trend of declining rainfall. At high to extreme levels of forcing, eastern Siberia experiences a reduction in precipitation, despite being outside the area where forcing is applied. To understand the variation in the precipitation response to absorbing aerosol forcing, other variables such as the vertically integrated precipitable water, specific humidity, evaporation, and circulation are considered.

Figure  (bottom two rows), showing the change in precipitable water with absorbing aerosol forcing, highlights a moist anomaly over the Middle East and dry anomalies over northeastern India, Southeast Asia, and eastern China. These anomalies only become statistically significant at 90 W m-2 forcing and above; this is in contrast to the changes in the precipitation, which become significant from 60 W m-2. The fact that the precipitation decreases at a faster rate than the precipitable water is primarily attributed to a reduction in precipitation efficiency, which is due to the increase in the static stability of the lower levels of the atmosphere, as opposed to a reduction in the moisture availability of the atmospheric column.

There is a slight moist anomaly over India up to 60 W m-2 forcing. Beyond 60 W m-2 forcing, there remains a region in central India that does not experience the significant reduction in precipitable water seen over Southeast Asia and eastern China. Similarly, looking at Fig. , the specific humidity at 925 hPa is higher over northern India up to 90 W m-2 forcing, and thereafter the specific humidity reduces significantly in the area of applied forcing, except for a small region in western central India. A vertical cross section of specific humidity along 20∘ N (Fig. S6, top two rows) also highlights the moist anomaly over northern India (70–90∘ E) at 30 and 60 W m-2 and the moist anomaly over the Middle East (around 50∘ E).

Figures  and S6 show that the greatest decline in specific humidity occurs near the surface, which corresponds with a reduction in the evaporation (Fig. S4). The absorbing aerosol forcing leads to a decrease in the evaporation because of the reduction of the solar radiation reaching the surface . There is a greater decrease in evaporation over eastern central and southern India than over northwestern India, in agreement with observed evapotranspiration trends in the period 1979–2008 .

Another key effect of the applied radiative forcing is to weaken the large-scale circulation. At low levels (Fig. , top two rows), there is a reduction in the strength of the southwesterly wind in the band 0–20∘ N, which is a key driver of both the South Asian and East Asian monsoons. With approximately 60 W m-2 of heating, the wind speed is reduced by 2–3 m s-1, and with 90 W m-2 heating, there is a 4–5 m s-1 reduction in wind speed. When the maximum forcing is applied, the speed of the southwesterly monsoon wind becomes close to zero. Stronger radiative forcing causes more pronounced surface cooling, weakening the land–sea temperature gradient and thus weakening the low-level monsoon flow.

There is a strengthening of the low-level wind in eastern China (20–40∘ N), causing more dry air to be advected from southwest to northeast towards eastern Siberia and corresponding to a reduction in precipitation in the region. The mid-level wind field (not shown) experiences changes of a similar magnitude to the low-level wind field. A narrow region stretching from northwestern India to Bangladesh and continuing northeastwards features increased wind speeds, a process associated with the formation of three atmospheric highs (Fig. S3).

At higher levels (Fig. , bottom two rows), from Southeast Asia to the eastern coast of Somalia there is a significant reduction in the speed of the tropical easterly jet. Higher absorbing aerosol forcing corresponds with greater declines in easterly wind speeds. Additionally, there is a slight weakening of the westerly subtropical jet, located north of India, at high forcing rates (>90 W m-2).

With light (30 W m-2) absorbing aerosol forcing, there is vertical ascent over land for the regions of southern India, Southeast Asia, and eastern China (Fig. S7), corresponding to the locations of peak convective rainfall (Fig. S7 – dotted lines). Note that Fig. S7 shows the vertical velocity in pressure coordinates (ω), meaning that negative vertical velocity corresponds to ascent and positive vertical velocity to descent. As the absorbing aerosol forcing increases, the upwards vertical velocity is reduced. This is consistent with increased atmospheric stratification and increased static stability, leading to suppression of convective precipitation.

In Fig. S7 (top two rows), there is an area of vertical ascent around 20∘ N, 70–80∘ E, which becomes stronger as the absorbing aerosol forcing increases. Although there is not an equivalent increase in the convective precipitation, in Fig.  there is a slight increase in total precipitation for central India at high and extreme levels of forcing. Thus, the anomalous precipitation over central India comes from large-scale rather than convective processes and is linked to changes in the circulation, likely related to non-local effects (explored further in Sect. ).

Enhanced carbon dioxide levels lead to higher surface temperatures, higher absolute humidity levels and weakening of the large-scale circulation , as discussed in Sect. . Here, we investigate the response of the South Asian and East Asian monsoons to combined absorbing aerosol and carbon dioxide forcing and whether doubling the carbon dioxide levels is sufficient to offset the aerosol effects of surface cooling and reduced precipitation.

Doubling the carbon dioxide levels leads to warmer surface temperatures, with regions of South Asia being several degrees warmer in the aerosol-with-2 × CO2 simulation than in the aerosol-only simulation (Fig. S8, top row); however, for absorbing aerosol forcing over 60 W m-2, the net effect of combined absorbing aerosol and carbon dioxide forcing is a surface cooling. At the 700 hPa level (Fig. S8, bottom row), the higher level of carbon dioxide also leads to several degrees of warming compared to absorbing aerosol forcing alone (aerosol-only simulation). From Fig. S10, the warming from enhanced carbon dioxide levels is greater at mid-levels than at the surface relative to the aerosol-only simulation.

Given the warmer surface temperatures in the aerosol-with-2 × CO2 simulation, we would expect to see greater atmospheric moisture content and higher rates of evaporation and precipitation, compared to the aerosol-only simulation. Both the precipitable water and specific humidity are significantly greater when carbon dioxide levels are doubled (Fig. d–f, Figs. S9 and S11). The change in specific humidity is most pronounced close to the surface (Figs. S9 and S11) and over the Middle East (around 20∘ N, 50∘ E), enhancing the moist anomaly there. In contrast, there is no significant change in the rate of evaporation (not shown) over land for the aerosol-with-2 × CO2 simulation compared to aerosol only. There is an increase in both the evaporation (significant) and the precipitation (moderate) over the Indian Ocean when carbon dioxide levels are doubled.

In terms of the precipitation, there are some differences in the spatial pattern between the aerosol-with-2 × CO2 and aerosol-only simulations. Areas of high orography such as the Western Ghats and the Himalayas experience slightly less rainfall under combined absorbing aerosol and carbon dioxide forcing than with absorbing aerosol forcing alone (Fig. a–c). Moreover, the leeward side of the Himalayas receives slightly more rainfall, suggesting a downwind shift of precipitation in response to carbon-dioxide-induced warming . For absorbing aerosol forcing larger than 60 W m-2, in combination with the doubled carbon dioxide concentration, the absolute response of the precipitation is to decrease.

The effect of the combined forcing is to further weaken the large-scale circulation with respect to the aerosol-only run. In particular, there is a additional weakening of the low-level southwesterly wind (Fig. a–c), reducing the moisture influx from the Arabian Sea to India. At mid-levels (not shown), the anti-cyclonic wind over the Middle East is stronger in the aerosol-with-2 × CO2 simulation than the aerosol-only simulations, corresponding to an area of high pressure. Figure d–f, for the high level wind field, shows a further weakening of the tropical easterly jet compared to the aerosol-only run. On the other hand, there is a slight strengthening of the westerly subtropical jet (located between 30–45∘ N), likely related to the warmer mid-tropospheric temperatures in the aerosol-with-2 × CO2 simulation and associated increase in meridional temperature gradient in the upper troposphere/lower stratosphere. There are no statistically significant changes in the vertical velocity for the aerosol-with-2 × CO2 simulation compared to the aerosol-only simulation, as shown by the lack of stippling in Fig. S12.

The previous experiments (Sects.  and ) involved applying a radiative forcing simultaneously across three regions: India, Southeast Asia, and eastern China (as per Fig. ). Several aspects of the monsoon response, such as the anomalous summer precipitation over northern India, merit further investigation. The method employed here is to apply the radiative forcing to each region separately in order to decompose the local and non-local effects of the forcing. This allows better understanding and attribution of the different spatial responses to specific areas of applied forcing. The focus is on an absorbing aerosol forcing value of 60 W m-2, which corresponds to an intense, yet not physically unreasonable, forcing.

Figure  shows the precipitation, surface temperature, and wind speed anomalies with respect to the control run for simulations with forcing applied to all regions (first column), India (second column), eastern China (third column), and Southeast Asia (fourth column). The JJA results are averaged over a 100-year period. The local effects of applying the absorbing aerosol forcing are similar for each region; namely, cooler surface temperatures and a reduction in precipitation. Over eastern China and Southeast Asia, the changes in precipitation are statistically significant, while the local reduction in surface temperature is statistically significant for all regions. Although there is a reduction in precipitation over India under locally applied forcing, it does not show as significant due to the high interannual variability of summer rainfall for the Indian region.

The strongest non-local effects are observed when applying forcing over eastern China, which leads to a slight reduction in surface temperature and a slight increase in precipitation over India (Fig. , top two rows). The precipitation response of India to forcing applied over eastern China is nearly as strong as when the forcing is applied locally, albeit with opposing trends. This results in net change of nearly zero for the precipitation over India at 60 W m-2 forcing, as seen in Fig.  (top row) and discussed in Sect. . Similar asymmetry in the teleconnection between eastern China and India in relation to local absorbing aerosol forcing has been shown in . There is a reasonably well-established link between the Indian monsoon and China, via the Silk Road pattern, which is associated with eastward-propagating upper-level Rossby waves . However, our case seems to be the reverse, with the effects from forcing over eastern China propagating westwards towards India. Thus, further study is required to fully understand the underlying mechanisms.

In terms of the low-level circulation (Fig. , third row), applying forcing to India leads to a slight increase in wind speed from the west coast to central India. Applying forcing to Southeast Asia causes a slight reduction in the low-level monsoon flow from the Arabian Sea to India, while applying forcing to eastern China causes a reduction in the low-level flow from the Bay of Bengal to Southeast Asia and an increase from eastern China to northeastern China. At high levels (Fig. , fourth row), forcing in Southeast Asia or eastern China induces an increase in the westerly subtropical jet speed and a decrease in the tropical easterly jet wind speed. The change in the tropical easterly jet is statistically significant when applying the forcing to eastern China.

Generally, the local response to regionally applied absorbing aerosol forcing is consistent with the response to the large-scale forcing (Sect. ). India is the most sensitive to remote forcing, with the local precipitation, surface temperature, and circulation being affected when forcing is applied to eastern China. We find that for Southeast Asia and eastern China, applying the absorbing aerosol forcing locally gives the greatest response, while for India, the responses to local and remote forcing are comparable. This is in agreement with and , who note the importance of both Chinese and Indian emissions on precipitation over India, but in disagreement with and , who suggest prioritising the influence of local aerosol sources. On the other hand, find that the biggest contributors to precipitation changes over India are from remote sources. Uncertainty in future emissions scenarios, in terms of location and intensity, remains a barrier to predicting the response of the South Asian and East Asian monsoons.

The strength of the South Asian and East Asian monsoons is largely determined by processes affecting the sea surface temperature, greenhouse gases, and aerosols. Here, we have conducted a parametric study with an intermediate-complexity climate model to assess the roles of absorbing aerosol and greenhouse gas forcing on the South Asian and East Asian monsoons. In addition, we have identified the level of regional forcing at which the monsoon system breaks down in terms of a significant reduction in precipitation.

Absorbing aerosol forcing, which we apply through a combination of mid-tropospheric heating and surface cooling in our model, causes decreasing surface temperatures, mid-level warming, weakening circulation, and a reduction in (convective) precipitation. As the forcing increases, the precipitation declines much faster than the precipitable water, indicating that it is due to a lack of precipitation efficiency related to changes in the stratification of the atmosphere, rather than due to a lack of moisture. Advection of dry air from eastern China leads to a reduction in precipitation in eastern Siberia, which is outside of the area being forced. On removal of the absorbing aerosol forcing, we find that the monsoon system recovers fully, indicating that there is no hysteresis in our model simulations. Doubling carbon dioxide concentration partially mitigates the effects of the absorbing aerosol forcing, through warmer surface temperatures enabling greater moisture take-up, but further weakens the large-scale circulation. We find that, when considering realistic ranges of applied forcings, the precipitation response is more sensitive to absorbing aerosol than greenhouse gas forcing, highlighting the importance of air quality policies and the impact they can have on the future state of the South Asian and East Asian monsoons.

The strongest regional responses, particularly in regards to the circulation, are attributed to absorbing aerosol loading over eastern China. Although the precipitation decline for each region directly corresponds to applying forcing to that region, there is a remote connection between eastern China and India. Forcing applied to eastern China leads to a slight increase in precipitation over India, which is in contrast with the response when forcing is applied to India. When both regions are forced simultaneously, there is a reduction in precipitation over India, but the reduction is much less than for Southeast Asia or eastern China. Comparing simulations where the regions have been forced separately to the simulation where the regions have been forced simultaneously, the results are qualitatively similar, indicating a fair degree of linearity in the response.

We have characterised regional regimes in terms of area-averaged precipitation, precipitable water, and surface temperature. India is separated into northern and southern regions due to the significant variance in their responses. Southern India is the least affected region, likely due to its peninsular nature. The precipitation response of northern India and Southeast Asia to increasing absorbing aerosol forcing is an approximately linear decline, with Southeast Asia showing a stronger negative sensitivity than northern India. For eastern China, there is an sharp transition at around 60 W m-2 to a regime where the precipitation is close to zero, indicating tipping behaviour. In terms of the precipitable water, it remains relatively constant for all regions until 60 W m-2; thereafter the precipitable water linearly declines with further increases in forcing. The surface temperature behaves similarly to the precipitable water but becomes much more sensitive to the forcing beyond 60 W m-2 and declines non-linearly.

We note the importance of aerosol loading over eastern China and the competition between aerosol loading over India in determining the response of the Indian region to future climate scenarios. At approximately 60 W m-2 of absorbing aerosol forcing, there is a clear transition from a high-precipitation regime to a low-precipitation regime for the eastern China region, while Southeast Asia and southern India show a more gradual linear decline in precipitation with increasing forcing. There is a compensating effect from eastern China absorbing aerosol forcing on precipitation over India, resulting in a lesser decline of precipitation over India compared with Southeast Asia or eastern China. For area-averaged precipitable water and surface temperature, there is transition at around 60 W m-2 for all regions from near-constant to decreasing levels. To maintain a safe operating space for the South Asian and East Asian monsoons, our results suggest keeping the absorbing aerosol forcing below 60 W m-2 through air quality policies and collaboration between Asian countries.

Our results are limited by the low resolution and lack of explicit aerosol interactions and chemistry, but future work will aim to address these issues by repeating similar experiments using a more complex global climate model. We would consider the effects of scattering aerosols and absorbing aerosols, both in combination and in isolation. Additionally, we would like to investigate the impact of applying forcing over a shorter seasonal period rather than perennially, particularly with regard to pre-monsoon conditions and the frequency of extreme weather events.