Predicting seasonal precipitation at regional scales with coupled models is challenging but essential for many applications. In this study, we compare the skills of the Monsoon Mission Coupled Forecast System version 2 (MMCFSv2) model in reproducing precipitation over five homogeneous regions of India; Northeast India (NEI), Northwest India (NWI), Central Northeast India (CNEI), West Central India (WCI), and South Peninsula India (SPI) during 1998–2022, compared with the previous version (MMCFSv1). We examine why MMCFSv2 better captures interannual variability and correlation than MMCFSv1, as well as the remaining challenges in regional forecasting. MMCFSv2 (MMCFSv1) shows seasonal mean precipitation skill of 0.72 (0.54) in the All-India Region (AIR), 0.38 (0.09) in CNEI, 0.66 (0.60) in WCI, 0.55 (0.45) in SPI, -0.18 (-0.42) in NEI, and 0.21 (0.53) in NWI. Rainfall over homogeneous regions is strongly teleconnected to Indo-Pacific Sea Surface Temperatures (SST), with the NWI influenced by the Arabian Sea (AS) and the others by tropical Pacific and Southern Indian Ocean (SIO) SSTs. MMCFSv2 simulates SIO and tropical Pacific SSTs well, particularly in the Niño 3.4, and Indian Ocean Dipole regions; however, it fails to capture AS SSTs, resulting in lower skill over NWI. Despite improved performance over these regions and AIR rainfall, MMCFSv2 underestimates southwestern WCI and northern SPI rainfall, shows a high July and August dry bias, and struggles with extremes along the Western Ghats and deficiencies over NWI. Seasonal mean precipitation bias values for MMCFSv2 (MMCFSv1) are − 1.50 (− 1.03) mm/day in WCI and 0.96 (0.55) mm/day in SPI. MMCFSv2 exhibits a stronger wet bias over NEI (June –September) and over northern CNEI near the Himalayan foothills (July–August). The standard deviation of rainfall is underestimated over northern CNEI (July), western NEI (June), and NWI (July-September). These findings highlight the limitations of MMCFSv2 in complex orographic regions and NWI, emphasizing the challenges for future development of regional rainfall prediction.

Over the last decade, tropical cyclone (TC) track and intensity predictions have improved by nearly 50% in the Atlantic and Northern Indian Ocean, driven by advancements in ocean-coupled numerical models, data assimilation techniques, and an expanding network of observations. However, the prediction of severe weather events driven by convection, particularly those associated with heavy precipitation over land, has not kept pace with these improvements in TC forecasting. While 1-2 km horizontal resolutions are crucial for capturing convection over land and ocean, seamless prediction across scales demands an accurate representation of the coupled evolution of ocean, land, and atmospheric states. To address the complex problem of severe weather across a spectrum of atmospheric motions-including TCs over the ocean and severe convective systems over coastal and inland regions-we have developed the Indian Ocean-Land-Atmosphere (IOLA) Coupled Mesoscale Prediction Framework. This Framework integrates the well-tested nonhydrostatic model (NMM) dynamical core with advanced nesting techniques from the hurricane weather research and forecast (HWRF) system. It further incorporates ocean coupling from HWRF and physics packages adopted from the WRF community model. This represents the first-ever coupled modeling system explicitly designed to tackle extreme weather events across multiple domains and scales. Extensive testing of this novel modeling framework demonstrates that a high-resolution (1-2 km) "all-purpose" severe weather prediction system can effectively address the challenges of forecasting extreme weather over the Indian region. One of the key focuses of this work is the application of 1-km horizontal resolution moving nests over the monsoon region, where synoptic-scale interactions play a critical role in modulating severe weather and heavy precipitation events. With this configuration, the model provides a high equitable threat score (ETS) > 0.18 for heavy to extreme rainfall events for 48 h and above lead times. This framework enables a unified approach to simulating severe weather phenomena accurately and flexibly. Also, it sets a new benchmark for seamless prediction of extreme weather, paving the way for improved resilience against coastal hazards and inland severe weather events.

Western Himalayan (WH) glaciers provide critical water resources but are rapidly retreating under climate change. The WH has warmed at 0.31 degrees C decade(-1) over 1979-2019, twice the rate of the broader Indian landmass (0.16 degrees C decade(-1)) and higher than the Himalayan mean (0.24 degrees C decade(-1)). Projections indicate further warming of similar to 3.5 degrees C-6 degrees C by 2100 under SSP3-7.0. We assess the response of WH glaciers to global warming levels of +1.5, +2.0, +3.0, and +4.0 degrees C using the open global glacier model forced with bias-corrected multi-model coupled model intercomparison project phase projections. Glacier meltwater is projected to peak near +2 degrees C before declining, with the total WH meltwater flux falling by 65.4 +/- 10.7% at +4 degrees C. The response is strongly size-dependent, with small glaciers (<2 km(-2)) already declining in melt at +1.5 degrees C, medium glaciers (2-10 km(-2)) peaking near +2 degrees C, and large glaciers (>10 km(-2)) sustaining melt increases through +3 degrees C before declining by 20.3 +/- 21.3% at +4 degrees C. The hydrological cycle broadens, and the rain fraction rises from 0.34 to 0.52 across the warming range, with implications for water management. The results presented here are for the WH glacierized region and are susceptible to uncertainties arising from inter-model variability and unrepresented cryosphere processes. Nevertheless, they consistently emphasize the importance of limiting warming to below 2 degrees C to sustain glacier-fed water resources.

The National Monsoon Mission programme of the Ministry of Earth Sciences, Government of India, was launched in 2012 to improve Indian monsoon forecasts from short-range to seasonal timescales using dynamical models. Two models, namely, the Climate Forecast System version 2 (CFSv2) from the United States National Centers for Environmental Prediction, and United Kingdom's UKMET Office unified model, were implemented and improvements were made. The initial climatologies of the seasonal predictions showed many biases, including cold biases in the sea surface temperature over the tropical Indian Ocean. The Airsea flux parametrization schemes used in these models, which were not verified under the monsoonal conditions over the North Indian Ocean due to the lack of relevant in situ observations, were among the possible causes of the bias. To test this, an ocean moored buoy with instrument packages for complete surface energy balance, including eddy-covariance flux sensors, was deployed in the North Bay of Bengal from July 2019 to August 2020. The present article compares measured and model scheme predicted surface fluxes. The surface shear stress is well predicted in both models. The heat loss due to evaporation of water is over-estimated in the 8 to 38 W m-2 range depending on the parametrization scheme used. The present version of the scheme implemented in CFSv2 has the smallest bias. Whereas a bias of 8 W m-2 in evaporative heat loss falls in the range of its measurement uncertainty limits, scope for further refinement exists.

The years 2021 and 2022 witnessed negative Indian Ocean Dipole (nIOD) conditions, with the 2022 event being the strongest on record. The dipole mode index was negative since the summer of 2021 and remained negative until early winter 2022, an unprecedented duration of 19 months. This makes it the first such occurrence of a multi-year nIOD. It co-existed with a triple-dip La Ni & ntilde;a event during 2020-2022. In this study, we explore the dynamics behind the occurrence of this multi-year nIOD event. The tropical Indian Ocean (TIO) witnessed predominant westerly wind anomalies starting in the summer of 2021 and lasting till the end of 2022, with a record number and duration of westerly wind bursts (WWBs). The anomalous westerlies were possibly supported by the background La Ni & ntilde;a state and anomalous convection over the eastern TIO associated with tropical intra-seasonal oscillations. Occurrences of WWBs outside their preferred climatological months and strong westerly wind anomalies modulated the intensity of the zonal currents and the Wyrtki jets in the TIO. The associated heat and mass transfer caused the depression of the thermocline in the eastern TIO, resulting in the sustenance of nIOD conditions. Anomalous westerly wind activity in the TIO during the spring of 2022 served as a bridge between the two nIOD events and sustained it for a record duration. This multi-year nIOD event thus prevented the Indian summer monsoon rainfall from being in large excess, as the monsoon conducive modulation of the Walker circulation was counteracted by the anomalous subsidence over India by the nIOD-modulated regional Hadley circulation.

Prolonged periods of anomalously high sea surface temperatures (SSTs) are becoming increasingly frequent and intense due to climate change and regional climate variability. Accurate seasonal prediction of marine heatwaves (MHWs) is crucial for developing early warning systems and ensuring sustainable management of marine resources. This study evaluates the prediction skill of MHWs in the Indian Ocean and surrounding basins using the Monsoon Mission Coupled Forecast System version 1 (MMCFSv1) for the period 1982–2017. MHWs are detected using the 90th percentile SST threshold from the full ensemble distribution across model members. Forecast skill is assessed using the Pearson correlation coefficient and Mean Squared Skill Score (MSSS) for various lead times and initial conditions. Results show that seasonal forecasts possess good forecast skill across key regions, including the Western Arabian Sea, the North Bay of Bengal, and the North-East Pacific, particularly during the March-April-May (MAM) and June-July-August (JJA) seasons. Ensemble-based thresholds better capture the spatial and temporal variability of MHWs. The highest skill is observed with initializations in February and March, while later months show a gradual decline in performance. This study highlights the importance of ensemble spread in capturing MHW extremes and provides a robust framework for enhancing seasonal MHW forecasts. The findings contribute to advancing climate-resilient planning and marine hazard mitigation strategies in the Indian Ocean basin.

The mesoscale eddies present in many dynamic regions of the world ocean are known to modulate wind and precipitation in their vicinity. This study investigates the observed association between eddy activity in the western Bay of Bengal and seasonal rainfall over the monsoon core region during the boreal summer. A novel method is employed to characterise eddy variability in the Bay of Bengal using the Okubo-Weiss (OW) parameter. The anticyclonic (cyclonic) eddy activity in the western Bay correlates with increased (decreased) monsoon rainfall and westerly (easterly) wind anomalies across the monsoon core zone. During years when BoB dominated by anticyclonic eddies, the seasonal Indian summer monsoon rainfall (ISMR) in this region exceeds the climatological mean by up to 35 % of its standard deviation. A mechanism linking the anticyclonic eddies and the ISMR is proposed here based on the observational data for the 1993-2022 period. The presence of anticyclonic eddies appears to mitigate the detrimental effects of El Nino on east-central Indian rainfall by fostering atmospheric conditions favourable for monsoon low-pressure system formation. Furthermore, years dominated by anticyclonic eddies exhibit a stronger, narrow coastal SST gradient in the northwestern Bay of Bengal, which may further enhance LPS development. Methodically structured model sensitivity experiments using a standalone atmospheric model substantiate the impact of mesoscale SST structures and the coastal gradient in the BoB on rainfall, thus providing potential evidence regarding the interplay between eddies and the Indian monsoon system.

Lightning associated with thunderstorms (TS) represents a major natural hazard worldwide. The occurrence of lightning is primarily governed by the dynamic and microphysical characteristics of thunderclouds, while the environmental conditions in which these clouds develop play a crucial role in modulating lightning activity. Information on the spatial and temporal variability of lightning and convection provides valuable insight into cloud electrification processes and resultant lightning activity. Such information is also essential for assessing the contribution of lightning to the global electric circuit and for improving lightning safety and protection strategies. Accurate representation of microphysical processes in tropical convection remains a significant challenge in meteorology. Over the Indian subcontinent, strong regional variability exists in the relationships between thunderstorm activity and thermodynamical, microphysical and dynamical parameters, leading to heterogeneity in lightning flash density. This variability is largely attributed to the accumulation of large ice particles in the upper regions of convective clouds, which is positively correlated with convective available potential energy (CAPE). The study reveals a strong linkage between lightning activity and ice-related cloud properties, including snow, rainwater, cloud water and particularly ice Convective Precipitation flux (ICPF). Among these, ICPF exhibits the highest correlation with lightning, identifying it as a robust indicator of lightning activity. Lightning is strongly associated with convective rainfall produced by intense, vertically developed clouds, especially during the pre-monsoon season. A weaker yet statistically significant relationship is also observed with stratiform rainfall. Overall, the findings highlight the influence of regional meteorological conditions, cloud microphysics and seasonal variability on lightning activity, offering valuable insights for improving storm prediction and understanding thunderstorm behaviour.

Accurate forecasts of the Indian Summer Monsoon (ISM) are crucial for climate monitoring and operational planning, yet they remain challenging due to model biases and complex land-ice-ocean-atmosphere interactions. The hindcast simulations of the ISM with the latest version of Monsoon Mission Climate Forecast System version 2(MMCFSv2) are assessed at different lead times to evaluate the seasonal skill of ISM at different lead times. Hindcast simulations for 28 years (1998–2024) using 10 ensemble members reveal that the highest ISMR skill (~ 0.72) is achieved with April initialization, with ISMR phase skill improving by 17% and amplitude by 20%. All three tercile (below normal, normal, and above normal) has better representation for April IC. The improvement in ISMR teleconnection with large-scale climate forcing is also noticed. Rainfall anomalies for excess and deficit years composites are better simulated in April IC, followed by February, May, and March IC. The impact of the differing initial conditions on forecast errors varies and provides predictive skill for up to three months, with February initial conditions being particularly crucial. The non-ENSO deficit years in April IC are captured better than in February IC, which is one of the important reasons for the highest ISMR skill. The pattern correlation is highest in April IC as compared to other ICs. MMCFSv2, with April IC having the highest skill shows improved prediction of extreme monsoon years.

This study evaluated the performance of various global storm-resolving models (GSRMs) in simulating the Indian Summer Monsoon (ISM), highlighting the limitations of current global kilometer-scale models in accurately representing precipitation patterns over the Indian subcontinent. The DYAMOND (DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains) project provides a critical intercomparison platform for these models. Participating models-such as ARPNH, GEOS, MPAS, SAM, NICAM, FV3, and UM-explicitly resolve deep convection, reducing dependence on parameterizations that often introduce biases in low-resolution models. Despite operating at high spatial resolutions (ranging from 1.5 to 5 km), these models exhibit notable discrepancies in simulating Indian Summer Monsoon (ISM) precipitation. For instance, NICAM and FV3 tend to overestimate precipitation across longitudes (65-100 degrees E) while the UM model significantly underestimates it over the central Indian region. These variations are largely due to differences in how convection statistics and rainfall probability distributions are represented in the models. Interestingly, these models simulated moderate to heavy rainfall events with comparatively less bias than other categories. However, this is insufficient on its own to deliver accurate kilometer-scale forecasts. Improvements in numerical schemes and the representation of physical processes, particularly cloud microphysics, are essential for advancing the simulation of ISM rainfall on a seasonal scale, both in mean and distribution.

Recent advances in machine learning offer new opportunities to enhance air–sea flux representations beyond traditional empirical and semi-empirical formulations. In this study, we develop a machine learning–based turbulent flux parameterization framework (SamudraFlux) for estimating latent heat, sensible heat, and wind stress. A comprehensive dataset of eddy covariance flux measurements from focused ship campaigns across the Bay of Bengal, Arabian Sea, tropical Pacific, and tropical Atlantic Oceans forms the basis for model development. These high-quality fluxes serve as targets, while a suite of meteorological and oceanic variables-including wind speed, humidity, air temperature, sea-surface temperature, stability indices, and radiative fluxes-constitutes the feature space. The SamudraFlux system employs an eXtreme Gradient Boosting (XGBoost) regression model, optimized via systematic hyperparameter tuning, to learn nonlinear, multi-regime relationships between surface forcing parameters and turbulent fluxes. This enables the model to capture variability and interactions that conventional bulk algorithms often miss. Model performance is evaluated against widely used physical parameterizations, including various bulk flux schemes, using standard statistical metrics (Coefficient of Determination, Root Mean Square Error, Mean Absolute Error, Correlation Coefficient). During in-sample testing, the SamudraFlux model significantly outperforms all physical formulations, yielding substantially lower errors and higher predictive accuracy. Under out-of-sample conditions, across independent oceanic regimes and unseen ship-based observations, SamudraFlux performs on par with the best physics-based parameterizations, demonstrating strong generalization. An extreme-event analysis based on the upper and lower 5

Quantitative prediction of the intensity of rainfall events (light or heavy) has remained a challenge in Numerical Weather Prediction (NWP) models. For the first time, the mean coefficient of diffusional growth rate ( c_m ) is calculated using a Eulerian-Lagrangian particle-based model on in situ airborne measurement data from the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) during monsoon over the Indian sub-continent. The results show that c_m varies in the range of ∼ 0.25× 10^-3 - 1.5× 10^-3 (cm s^-1 ). The generic problem of overestimation of light rain in NWP models might be related to the choice of c_m in the model. It is also shown from a direct numerical simulation (DNS) experiment using small-scale model that relative dispersion ( ϵ ) is constrained with average values in the range of ∼ 0.2 - 0.37 ( ∼ 0.1 - 0.26) in less humid (more humid) conditions. This is in agreement with in situ airborne observation ( ϵ ∼ 0.36) and previous studies over the Indian sub-continent. The linear relationship between relative dispersion ( ϵ ) and cloud droplet number concentration (NC) is obtained using CAIPEEX. The present study compares different exciting parameterizations for the cloud-to-rain “autoconversion” and effective radius using a sophisticated parcel-DNS model guided by CAIPEEX observation. The dispersion-based ‘autoconversion’ and effective radius parameterization schemes for the Indian region must be useful for the calculation of Indian summer monsoon precipitation in the general circulation model. The present study also provides valuable guidance for parameterizing the effective radius, which is important for the radiation scheme.

The present study investigates the decadal variation of summer monsoon extreme precipitation over India for the period 1901 to 2020 using the Generalised Extreme Value (GEV) distribution. The analysis focuses on the spatial and temporal variation of annual maximum rainfall (Yearmax), assessing the evolution of GEV parameters- shape (ξ), location (µ) and scale (σ) – over the three climatological epochs. GEV distribution effectively captures the patterns of extreme rainfall and provides further insights into the frequency, intensity, and duration through its parameters. Comparison with the earlier decades reveals a significant anomalous increase in the increases of extreme precipitation in the recent decades over west-central India, where rainfall extreme values have risen from 78 mm (1900–1940) to 88 mm (1980–2020), while a decrease has been observed in east-central India. The rise in extreme rainfall in west-central India is attributed to broader climate change feedback, such as persistent warming of SST over the Arabian Sea, an increase in land-sea contrast, anomalous stronger winds, enhanced moisture transport, westward shifts in monsoon circulation, local convective amplification due to changes in land surface, and broader climate change feedback. Conversely, the anomalous decline in extreme rainfall over east-central India is correlated with the weaker SST anomalies in the Bay of Bengal, reduced anomalous westerly winds, and fewer intense low-pressure systems. The study presents a comprehensive analysis of the bimodal distribution of extreme rainfall in recent decades, with the backdrop of recent factors related to climate change. These analyses are helpful for improving forecasts and developing effective mitigation strategies for the present scenario of extreme weather conditions.

In recent decades, the spread, intensity and frequency of extreme monsoon events over India have increased significantly. The year 2019 witnessed a widespread increase in extreme precipitation. Although oceanic variables have a role, the mechanism of rain formation and detailed cloud properties remain unclear for extremes. Using satellite and reanalysis datasets, we examined the interplay between cloud characteristics and dynamics in extreme and non-extreme rainfall events for the 2019 monsoon. The study reveals that extreme events depict higher cloud ice (260.84 g/m2) and liquid (209.60 g/m2) mass than non-extreme events (161.67 and 116.64 g/m(2)) averaged over central India (CI) due to increased atmospheric moisture availability. Extreme events were associated with greater ice particles of all sizes, while non-extreme events had lower number densities, primarily concentrated within the 21-35 mu m range. These findings underscore the essential role of cloud microphysics in influencing extreme rainfall. The study reveals distinct differences in dynamical features, including atmospheric circulation, specific humidity, outgoing longwave radiation (OLR), and vertical velocity between extreme and non-extreme events. The average specific humidity over CI is higher (9.01 g/kg) in extreme than non-extreme (7.93 g/kg). The lower OLR in extreme (197.95 W/m(2)) than non-extreme events (210.80 W/m(2)) indicates enhanced deep convection, which supports more stratiform rain, contributed significantly (similar to 45%) to the total precipitation and significant difference in convective rain during extreme and non-extreme events. In contrast, the Monsoon Mission Coupled Forecast System version 2.0 (MMCFSv2) model shows that the percentage of convective rainfall does not differ between extreme (76.56%) and non-extreme (75.5%) events. Our findings deepen our understanding of the complex relationships between ocean and atmospheric dynamics and cloud properties, emphasising their pivotal influence on quantified precipitation forecasts (QPF) to improve extreme rainfall simulation.

In recent years, the scientific community has placed growing emphasis on the impact of aerosols on the Indian Summer Monsoon (ISM) system. The ISM, the world’s strongest monsoon system, delivers approximately 80% of India’s annual rainfall from June to September, sustaining ecosystems, agriculture, and millions of livelihoods. The Indian subcontinent, with its diverse geography, dense population, and industrial zones, is a major source of aerosols through short-range and long-range transportation. Aerosols are instrumental in the process of cloud formation, functioning as cloud condensation nuclei (CCN) and ice nuclei (IN), which are vital for the formation and evolution of cloud hydrometers. Numerous possible mechanisms by how aerosols influence rainfall have been suggested by recent research on the direct radiative effects of aerosols. However, the microphysical impact of aerosols on monsoonal rainfall in the Indian Summer Monsoon Region (ISMR) remains largely unexplored. Indian summer monsoon rainfall is influenced by the large-scale circulation and different monsoon drivers. The large-scale driver modulates clouds and indicates the sign of seasonal mean ISM rainfall anomalies and the role of aerosols are secondary. Our current understanding of both the direction and magnitude of aerosol-cloud interaction (ACI), which induced changes in rainfall is insufficient. Furthermore, changes in thermodynamic and climatic circumstances, precipitation types, their vertical distribution in the atmosphere, cloud and dynamics all have a significant impact on the ACI and give feedback to ISM rainfall. In this context, we will carry out a comprehensive analysis of multi-satellite observations and numerical model simulations to examine the role of aerosol on cloud properties and precipitation susceptibility. The analysis of multi-satellite data reveals considerable spatial and vertical variability of dust aerosols over the ISM region. Increased dust activity can modify the monsoon cloud system, leading to significant changes in the microphysics of both the liquid and ice phases over the ISM region. The process analysis of ACI is crucial for accurately predicting monsoonal rainfall and will help resolve discrepancies in aerosol-cloud-rainfall interactions between models and observations. While more aerosols tend to reduce the cloud drop size and delay the warm rain, during the Indian summer monsoon, this is overcome by invigoration in higher moisture environments and cold-rain processes. The observational and modeling studies will be helpful for in depth understanding the role of aerosols and their interactions with clouds on the hydrological cycle by modifying the cloud properties and monsoon intraseasonal oscillations. The process studies must be beneficial for the realistic parameterization of cloud processes in the NWP model and can provide a pathway for increasing the grid point ISM rainfall skill through fundamental basic research on cloud microphysical processes.

Synoptic-scale systems, such as monsoon low-pressure systems (LPSs), contribute significantly to seasonal mean monsoon rainfall. Therefore, realistic simulation of characteristics (i.e., intensity, frequency and propagation) of LPSs is crucial for reducing the dry rainfall biases in general circulation models. Recent studies have argued that the generation and propagation of LPSs are strongly modulated by the narrow coastal sea surface temperature (SST) fronts over the Bay of Bengal (BoB). Therefore, the present study addresses the improvement in LPS characteristics in a coupled model through better representation of coastal SST fronts over the BoB. LPSs over the Indian regions are tracked through the Tempest Extreme v2.1 tracking algorithm. Through a comparative analysis, the present study demonstrates that the recently developed coupled model, that is, the Monsoon Mission Climate Forecast System (MMCFS) version 2 (v2), simulates a higher number of LPSs generated with larger inland propagation as compared to the previous generation model (MMCFSv1). Correct location of positive vorticity anomaly, a larger extent of positive vorticity and circulation anomalies into the landmass and the stronger temporal evolution of sea level pressure or vorticity anomalies are a few of the favourable conditions that support the improved LPS characteristics in MMCFSv2.

The Indian summer monsoon (ISM) transition is characterized by the seasonal reversal of winds from northeasterly in winter to southwesterly in summer over the north Indian Ocean. Physical processes that govern ISM dynamics are complex, with major characteristics being strong monsoon winds that pick up and transport moisture eastward, resulting in rain over the Indian subcontinent. The accumulation of moisture across the northern Indian Ocean depends on processes occurring on a range of scales across the air-sea transition zone, encapsulating the oceanic and atmospheric boundary layers and the air-sea interface. Many small-scale processes across the air-sea transition layer are not resolved by coupled prediction models and are instead represented by parameterizations, introducing uncertainty. The Enhancing Knowledge of the Arabian Sea Marine Environment through Science and Advanced Training (EKAMSAT) program aims to improve the understanding and parameterization of critical, unresolved small-scale processes that will improve monsoon prediction. This is accomplished through a team-based approach led by Indian and U.S. institutions from the research and operational communities that combines in situ process and large-scale remote observations, multiscale modeling, cross-scale synthesis, innovative training, and capacity building. This is being achieved through intensive multiplatform observational programs across the Arabian Sea and Bay of Bengal alongside a hierarchy of numerical simulations spanning process, regional, and global circulation models. With a focus on the northern Indian Ocean, the program aims to identify processes governing momentum, heat, and freshwater exchange across the air-sea interface during the ISM transition.

Prediction and study of Weather and Climate has been a challenging topic for scientists worldwide. These are muti-component and multi-phase systems with high degree of complexity. Simulating/predicting such a complex system needs significant computational resources. Using conventional computing techniques has always been expensive, and was more so in the 80s. Narasimha recognized the importance of the role of Parallel Computing /HPC (High Performance Computing) for weather and climate modelling in the 1980s. He encouraged the use of parallel computing /HPC techniques to study climate especially the Indian Monsoon. Using Flosolver, the indigenously developed parallel computing system at the CSIR (Council for Scientific and Industrial Research)—National Aerospace Laboratories (NAL), a climate model was implemented during 1986–1988. This was perhaps one of the earliest such implementations worldwide. This was followed by the implementation of a Medium Range Forecast (MRF the opearational model then being used by National Centre for Medium Range Weather Forecasting, NCMRWF) model on an upgraded version of Flosolver during early 1990s. The uniqueness of Flosolver was that identifying the needs of spectral models, a new type of intercommunication switch, FloSwitch, was designed and developed that could not only facilitate communications but would also process information and thus reduce communication overheads and improve scalability. The knowledge so gained by implementation of the the MRF model also helped in a complete remoulding of this model and was followed up with the development of an indigenous model viz. VARSHA. This model had many unique features that incorporated research in various aspects of atmospheric modelling by Indian researchers. Narasimha also played a pivotal role in the establishment of Ministry of Earth Sciences (MoES). Today, the institutes under this ministry have contributed significantly in improving the forecasts of monsoons and other weather events. Thanks to the strong foundations laid by Narasimha, India has been successfully exploiting HPC for cutting edge simulations in climate and weather predictions at all scales. Here we trace the evolution of Climate and Weather studies/prediction using parallel processing, from its beginnings to the present, especially in the Indian context and also speculate on the future directions.

The Indian summer monsoon (ISM) and associated monsoon intraseasonal oscillations (MISOs) influence the billions of people living in the Indian subcontinent. This study explores the role of autoconversion parameterization in microphysical schemes for the simulation of MISO with the coupled climate model, for example, the Climate Forecast System version 2 (CFSv2), by conducting sensitivity experiments in two resolutions (∼100 and ∼38 km). Results reveal that the modified autoconversion parameterization better simulates the active‐break spells of the ISM rainfall. The main improvements include the contrasting features of rainfall over land and ocean and the MISO index, representing MISO periodicity. The improvements are qualitatively and quantitatively more significant in the higher‐resolution simulations, particularly regarding rainfall spatial patterns over the Indian subcontinent during active spells. The MISO monitoring index in the revised CFSv2 also shows improvement compared with the control run. This study concludes that proper autoconversion parameterization in the coupled climate model can lead to enhanced representation of active‐break spells and sub‐seasonal variability of ISM.