Robust increase of Indian monsoon rainfall and its variability under future warming in CMIP-6 models

The Indian summer monsoon is an integral part of the global climate system. As its seasonal rainfall plays a crucial role in India’s agriculture and shapes many other aspects of life, it affects the livelihood of a fifth of the world’s population. It is therefore highly relevant to assess its change under potential future climate change. Global climate models within the Coupled Model Intercomparison Project Phase 5 (CMIP-5) indicated a consistent increase in monsoon rainfall and its variability under global warming. Since the range of the results of CMIP-5 was still large and the confidence in the models was limited due 5 to partly poor representation of observed rainfall, the updates within the latest generation of climate models in CMIP-6 are of interest. Here, we analyse 32 models of the latest CMIP-6 exercise with regard to their annual mean monsoon rainfall and its variability. All of these models show a substantial increase in June-to-September (JJAS) mean rainfall under unabated climate change (SSP5-8.5) and most do also for the other three Shared Socioeconomic Pathways analyzed (SSP1-2.6, SSP24.5, SSP3-7.0). Moreover, the simulation ensemble indicates a linear dependence of rainfall on global mean temperature with 10 high agreement between the models and independent of the SSP if global warming is the dominant forcing of the monsoon dynamics as it is in the 21st century; the multi-model mean for JJAS projects an increase of 0.33 mm/day and 5.3% per degree of global warming. This is significantly higher than in the CMIP-5 projections. Most models project that the increase will contribute to the precipitation especially in the Himalaya region and to the northeast of the Bay of Bengal, as well as the west coast of India. Interannual variability is found to be increasing in the higher-warming scenarios by almost all models. The 15 CMIP-6 simulations largely confirm the findings from CMIP-5 models, but show an increased robustness across models with reduced uncertainties and updated magnitudes towards a stronger increase in monsoon rainfall.

Within the latest studies using global coupled models, there is a widespread consensus that the Indian monsoon rainfall will increase due to climate change in the 21st century (Chaturvedi et al., 2012;Menon et al., 2013;Lee and Wang, 2014;Asharaf and Ahrens, 2 . This trend is found for various CMIP-5 models (Menon et al., 2013), the multi-model mean (Chaturvedi et al., 2012), the 150 mean of onle the four best models (Lee and Wang, 2014) or the model with the best deep convection scheme (Varghese et al., 2020) . Under Representative Concentration Pathway 8.5 (RCP-8.5) CMIP-5 models project a median increase in Indian monsoons rainfall of 2.3%/K (Menon et al., 2013). Also under SSP5-8.5, the amount of rainfall over India is projected to increase by 18.7% by the end of the 21st century compared to 1961-1999(Chaturvedi et al., 2012. This trend is expected to be the consequence of the warming of the Indian Ocean enhancing atmospheric moisture content and thus moisture convergence 155 (Cherchi et al., 2011;Seth et al., 2013;Mei et al., 2015;Sooraj et al., 2015). This so called thermodynamic effect dominates over the dynamic effect which refers to a reduced monsoon circulation due to a weakened walker circulation and an expected decrease of rainfall (Vecchi et al., 2006;Mei et al., 2015;Sooraj et al., 2015). The interannual variability is projected to increase in most models under the strongly forced scenarios as well as in models with good performance in capturing the mean seasonal cycle in the present climate (Kitoh et al., 1997;Menon et al., 2013;Jayasankar et al., 2015;Sharmila et al., 2015;Kitoh, 2017) 160 . The uncertain role of Here, we aim to update the CMIP projections for the Indian monsoon rainfall and its interannual variability for the 21st century by using 32 models of the latest climate model generation. For this purpose, we use the shared socioeconomic pathways and possible corresponding forcing levels as scenario framework (O'Neill et al., 2017). Section 2 gives a brief overview of the data used and processed. In section 3.1 we evaluate the participating models according to their capacity of modeling the Indian 165 summer monsoon in historic periods. Section 3.2 presents the results of the mean summer monsoon precipitation while section 3.3 focuses on the long-term trend of interannual variability. The results are discussed in section 4.

Model comparison
To evaluate the models' quantitative capacities of capturing the Indian monsoon rainfall, we compare their projected seasonal 2 3 4 5 6 7 8 JJAS Mean Rainfall (mm/day) -8 1985-2015 1900-1930 1985-2015. 205 In order to identify models with a potentially realistic representation of the Indian monsoon rainfall, we also analyze the spatial precipitation distribution for 1985-2015. We choose this period since it is closer to present time and therefore closer India reaches rainfall values above 10 mm/day (Fig. 2). The spatial rainfall pattern for the CMIP-6 models in 1985-2015 is shown in Fig. 3. Models that captured the rainfall quantitatively well mostly simulate a spatial pattern close to the reference distribution e.g. NorESM2-MM, CNRM-CM6-1, FGOALS-f3-L. FIO-ESM-2-0 overestimates the rainfall in the Himalaya region. The models with the tendency to underestimate the rainfall as ACCESS-CM2, CanESM5-CanOE, CanESM5 mostly are not able to capture the spatial pattern. Especially the southwest coast of India and the Himalaya region are not reproduced 215 according to the reanalysis data by most of these models. An exemption for the models with low rainfall values are the models of EC-Earth-Consortium (EC-Earth3, EC-Earth3-Veg) which simulate a pattern very close to the reference distribution.
Presenting ::: For :::::::: presenting : and discussing the results of this study, we decided to focus on the models within mean plus/minus twice the standard deviation which also deliver a reasonable spatial rainfall pattern. Nevertheless, we will provide information for all 32 models.

Trend in Indian summer monsoon mean rainfall for the end of the 21st century
In order to determine the long-term trend in Indian monsoon rainfall, we first analyze the temporal time series between 1850-2100 for all models under SSP5-8.5 (Fig. 4) Table 1. project an increasing trend, too. The only models to project a decrease are the models of the National Center for Atmospheric

230
Research (CESM2-WACCM in SSP1-2.5 and SSP2-4.5 and CESM2 in SSP2-4.5). On average over all models an increase of 24,3% is projected under SSP5-8.5 (Fig. 6) and of +18,6% in SSP3-7.0 (Appendix Fig. A1), of +11,9% in SSP2-4.5 (Appendix Fig. B1) and of +9.7% in SSP1-2.6 (Appendix Fig. C1). CanESM5 and CanESM5-CanOE show the maximum relative increase in all scenarios by the end of the 21st century. But as shown in Fig. 1 and Fig. 3, they clearly underestimate the rainfall and do not capture a realistic pattern of the rainfall distribution. CESM2-WACCM shows the minimal increase of 7.8% under 235 SSP5-8.5. This model was able to capture the mean rainfall in 1985-2015 within twice the standard deviation and is able to capture a reasonable pattern of the rainfall. Focusing on the models that captured the mean rainfall in 1985-2015 within twice the standard deviation (upper panel in Fig. 6), the relative increase is 17.4% under SSP5-8.5, i.e. slightly less than the average over all models. Also in the other scenarios the trend is less for these models compared to the average over all models. In summary, a robust increase of seasonal rainfall between 1985-2015 and 2070-2100 can be derived under global warming.

240
Most models project that this increase will contribute to the precipitation especially in the Himalaya region and to the northeast of the Bay of Bengal, as well as the Western Ghats (Fig. 7). Individual models indicate decreasing rainfall along the southwest coast of India and around Myanmar.
an increase of 0.33 mm/day ranging from 0.11 mm/day to 0.54 mm/day. The relative dependence is 5.3% per degree of global warming ranging from 1.7%/K to 13.4%/K for SSP5-8.5 across models. Considering only the more realistic models, the projected mean change is 6.1%/K for SSP5-8.5.

Long-term trend of interannual variability 250
In order to analyze the future evolution of interannual variability, we removed the nonlinear trend obtained by a Singular Spectrum Analysis from the rainfall data as displayed in Fig. 4, and use the percentage changes in standard deviation for the period 2050-2100 with respect to 1900-1950. Under SSP5-8.5, 28 of 32 models indicate an increase of interannual variability ( Fig. 9); the multi-model mean in this scenario indicates an increase of 21.3%. The strongest increase of 56.2% is simulated by EC-Earth3-Veg which is a model that does not capture the quantitative rainfall of the Indian summer monsoon well. Four  projecting the Indian summer monsoon than some other models. Nevertheless, among the 16 models within twice the standard 260 deviation, 13 project an increase in interannual variability. In SSP3-7.0, 22 out of the available 27 models project an increase of interannual variability (See Appendix Fig D1). The signal in the scenarios with less forcing is less clear (See Appendix Fig. E1 and Fig. F1), but even in SSP1-2.6 still 21 out of 31 available models project an increase in interannual variability until the second half of the 21st century. For the purpose of comparison, we also calculated the change without removing the trend and found that for SSP5-8.5 all models project an increase in interannual variability in average 39.9%. Fig. 10 shows the 265 dependence of interannual variability on global mean temperature for all available models (after removing the trend). As the global mean temperature change grows with stronger forcing, the positive trend in interannual variability becomes clearer.

Discussion
In this study, the long-term trend of the Indian summer monsoon and its variability have been analyzed based on the latest global coupled model simulations under the SSP scenarios. Our approach addresses the question whether the results from 270 earlier studies can be confirmed or need to be adapted in their sign or magnitude.
By comparing the CMIP-6 projection results with the WFDE5 reanalysis data, we classified some models as probably more capable of simulating a realistic representation of the monsoon rainfall. The share of models that capture the reference rainfall within twice the standard deviation has slightly increased in CMIP-6 (16 out of 32) in comparison to the precursor models in CMIP-5 (9 out of 20) (Menon et al., 2013). But it has to be noted that the validation period and the used reanalysis data differ 275 between (Menon et al., 2013) and this study. The observation of quantitatively measurable improvement between CMIP-5 and  Table 1. Grey dashed lines indicate the slope (the hydrological sensitivity) for SSP5-8.5 CMIP-6 coincides with the results of Gusain et al. (2020). While all the models that were out of the two standard deviation range underestimated the mean in CMIP-5, thus revealing a very clear general tendency of underestimation, the 16 models outside of the range in CMIP-6 partly underestimated (13 models) and party overestimated (3 models) the observed mean in 1985-2015. Modeling centers whose models underestimated the rainfall within two standard deviations in our study mostly 280 underestimated the rainfall already in CMIP-5. Some models with realistic patterns in CMIP-6 are updates from CMIP-5 that already revealed a pattern relatively similar to reanalysis data, e.g. NorESM2-MM. As in CMIP-5, models with the tendency to underestimate the rainfall in the evaluation period are mostly not capable either of capturing the spatial rainfall pattern in CMIP-6. But there are also various models that improved their capacity in capturing the Indian monsoon, such as the models 2013). Additionally, we calculated the average multi-model trend of projected change in mean rainfall by the end of the 21st century. As some modeling centers provide several models and some of them are based on overlapping model components, the models cannot be regarded as independent from each other (See e.g., Knutti et al., 2017)). The results have to be interpreted against this background. The found average multi-model trend in CMIP-6 with an increase of +24.3% by 2100 seems stronger 295 in comparison to CMIP-5 (Chaturvedi et al., 2012;Menon et al., 2013). As some modeling centers provide several models and some of them are based on overlapping model components, the models cannot be regarded as independent from each other.
The median dependence of relative change in precipitation on GMT taking into account all models has increased from 3.2%/K 315 in CMIP-5 to 5.3%/K in CMIP-6. Considering only the models with a more realistic representation of the monsoon, the increase is even more noticeably from 2.3%/K in CMIP-5 to 6.1%/K in CMIP-6. It also has to be mentioned that the range of projected sensitivities has decreased remarkably from 1-19%/K in CMIP-5 to 2-13%/K in the latest generation of climate models, i.e.
the uncertainty in hydrological sensitivity has decreased with the model updates. Similar tendencies have been found for the equilibrium climate sensitivity in CMIP-6 Zelinka et al. (2020); Wyser et al. (2020). Which of the updated processes between 320 CMIP-5 and CMIP-6 described by Gusain et al. (2020) dominate in causing the increased sensitivity of the monsoon to global warming needs further investigation.
The increase in rainfall is projected to contribute to the precipitation in the Himalaya region, the northeast Bay of Bengal and the northwest coast of India. These regions coincide to a large extent with the existing monsoon rainfall pattern, leading to a wet regions get wetter pattern during June to September monsoon rainfall. The distribution of regions with projected increasing 325 precipitation in CMIP-6 confirms the projection of previous studies using CMIP-5 models (Chaturvedi et al., 2012;Menon et al., 2013;Sharmila et al., 2015). Furthermore, the increasing pattern is shared by a larger percentage of available models in CMIP-6 compared to CMIP-5. But our projection of increased rainfall over the Western Ghats does not coincide with the study of Varghese et al. (2020) projecting a decrease in this region. By focusing on high resolution models with the best deep convection scheme, their study reveals decreasing precipitation in the southwest coast of India, which is only captured by one 330 third of the CMIP-6 models in our study, including the CNRM-CM6-1-HR model. A finer resolution seems to be necessary to capture this trend which is not given for all CMIP-6 models.
From the 32 available models, 28 models project an increase in interannual variability. This result is not directly comparable to the study of Menon et al. (2013) since the removal of the trend in our study has a relevant influence on the results. Without the removal of the trend, i.e. following the method of Menon et al. (2013), all 32 models project an increase in interannual 335 variability which shows that the signal has become clearer in comparison to the results in CMIP-5 models. The projected increase in interannual variability coincides with other studies (Kitoh et al., 1997;Jayasankar et al., 2015;Sharmila et al., 2015;Kitoh, 2017). A dominant role in shaping the interannual variability is taken by the El Niño Southern oscillation (ENSO) (Turner and Annamalai, 2012). As El-Niño events typically coincide with dry monsoon years and La-Niña years are often accompanied by strong monsoon rainfall (Kumar et al., 2006), changes in the emergence of these events have a relevant impact 340 on the Indian summer monsoon. (Azad and Rajeevan, 2016) applied spectral analysis and found a shortening of the spectral periods of ENSO which might lead to a shift in the relationship of ENSO and monsoon rainfall.
In this study, we 5 :::::::::: Conclusion ::: We used 32 CMIP-6 models to analyse the Indian summer monsoon's response to climate change. In order to identify models 345 with a good representation of the Indian monsoon, we compared the models' simulations in the past to WFDE5 reanalysis data. We found that there are 16 out of 32 models that are able to capture the monsoon rainfall within twice the standard deviation in the period 1985-2015. This is a slight increase compared to CMIP-5. The models outside that range in CMIP-6 still have a tendency to underestimate the amount of precipitation in this period. This was already observed in CMIP-5 where all of the models out of the range underestimated the rainfall. In our analysis, we focused on the models with the more realistic 350 representation of the Indian monsoon. We found that all models show an increase in mean summer monsoon rainfall under SSP5-8.5 and SSP3-7.0 by the end of the 21st century. An increase also was found in SSP2-4.5 and SSP1-2.6 by all models apart from two models in SSP2-4.5 and one model in SSP1-2.6. Under SSP5-8.5, the models exceed the envelope of the baselines variability on average in 2045 in . An multi-model mean increase of rainfall of 24,3% is projected under SSP5-8.5 and of +18,6% in SSP3-7.0, of +11,9% in SSP2-4.5 and of +9.7% in SSP1-2.6. The majority of models project that 355 that the increase will contribute to the precipitation especially in the Himalaya region, the northeast of the Bay of Bengal and to the west coast of India. Besides, the simulation ensemble indicates a linear dependence of rainfall on global mean temperature independent of the SSP :: in ::: the :::: 21st :::::: century; the multi-model mean for JJAS projects an increase of 0.33mm/day and 5.3% per degree of global warming. Furthermore, under SSP5-8.5 a majority of 28 out of 32 models project an increase in interannual variability by the end of the 21st century after removing the trend with Singular Spectrum Analysis. 360 We have seen in this study that low resolution models did not capture the spatial pattern of the monsoon rainfall in historic periods well. Small scale topography and its atmosphere feedback influence the rainfall to a relevant extent. Thus, the ongoing effort to improve the resolution of the individual CMIP models should be continued. Since other rainfall features such as extremes and the variability of rainfall on a subseasonal scale are beyond the scope of this study, they need to be analyzed in further studies owing to their high relevance e.g. for high-risk flooding events.

365
The projected increase in summer monsoon rainfall in combination with the projected longterm increase in interannual variability will be accompanied by an increased number of extremely wet years and potentially more high rainfall events (Turner and Slingo, 2009;Sharmila et al., 2015). While crops need water especially in the initial growing period, high rainfall events during other growing states can harm the plants Revadekar and Preethi (2012). Thus, the projected development might have serious consequences for the agriculture in India and neighbouring regions. Since the change differs from the decreasing 370 tendency in the second half of the 20th century, the development of adaptation strategies for the 21st century is required.
Appendix A: Appendix A: Change in Indian mean summer monsoon rainfall Appendix D: Appendix B: Change in Indian summer monsoon rainfall interannual variability Table 1. Overview of data availability for the 32 models used in the study (precipitation/temperature). Only those models are selected for which data for historic period and SSP5-8.5 was available at the time of the study.
Modeling Center (Group) Model SSP1-2.6 SSP2-4.5 SSP3-7.0 SSP5-8.5 Alfred Wegener Institute (AWI) Commonwealth Scientific and Industrial Research Organisation, ARC Centre of Excellence Japan Agency for Marine-Earth Science and