A novel technique for nowcasting extreme rainfall events using early microphysical signatures of cloud development

Sinan Nizar; Jobin Thomas; P.J. Jainet; K.P. Sudheer

doi:10.1371/journal.pclm.0000497

Abstract

Accurate forecasting of extreme rainfall events (EREs) at a regional scale with higher lead times is challenging due to the uncertainties in weather model predictions. This study introduces a novel technique to nowcast heavy- and extreme-rainfall events by analyzing early microphysical signatures in mesoscale convective clouds. The method primarily utilizes the cloud top temperature (T) - cloud effective radius (r_e) profiles derived using remote sensing. We estimate the probability of the occurrence of heavy- and extreme-rainfall events using a logistic regression model with attributes extracted from the T-r_e profile and cloud droplet size distribution. Our analysis indicates that the T-r_e profiles for normal-, heavy-, and extreme-rainfall events exhibit distinct microphysical characteristics, with a prominent diffusional zone during EREs. Applying this model to nowcast recent EREs in the southern Western Ghats (Kerala, India) demonstrates an overall skill score of 93% and a lead time of at least six hours, underscoring the effectiveness of the approach for nowcasting EREs at a regional scale.

Citation: Nizar S, Thomas J, Jainet P, Sudheer K (2025) A novel technique for nowcasting extreme rainfall events using early microphysical signatures of cloud development. PLOS Clim 4(5): e0000497. https://doi.org/10.1371/journal.pclm.0000497

Editor: Ahmed Kenawy, Mansoura University, EGYPT

Received: June 6, 2024; Accepted: April 7, 2025; Published: May 12, 2025

Copyright: © 2025 Nizar et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The rainfall events were extracted from the India Meteorological Department (IMD) daily gridded rainfall dataset, which has a spatial resolution of 0.25° x 0.25° (https://www.imdpune.gov.in/cmpg/Griddata/Rainfall_25_NetCDF.html). The hourly cloud properties were acquired from the NASA Langley Cloud and Radiation Research Group (https://satcorps.larc.nasa.gov/). The half-hourly gauge calibrated rainfall product from the Integrated Multi-Satellite Retrievals (IMERG; Huffman et al., 2019) was acquired through the GES DISC platform (https://disc.gsfc.nasa.gov/) to validate the accuracy and lead time of the nowcasting model.

Funding: This work was conducted under the DST project “Setting up of State Climate Change Knowledge Cell Under National Mission” at the Institute for Climate Change Studies, Kottayam, Kerala, India. The authors acknowledge funding support from the Department of Science and Technology (DST), Government of India (Sanction No. DST/SPLICE/CCP/NMSKCC/PR-62/2016 (G)).

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

Extreme rainfall events (EREs), resulting in floods and landslides, pose a significant threat to communities, affecting natural and anthropogenic resources with substantial environmental and economic impacts. It is recognized that EREs occur globally, exhibiting significant variations based on the area, depth, and rain rate, with intense EREs observed across tropical land areas, predominantly in South Asia and South America [1]. Recent reports from the Intergovernmental Panel on Climate Change have highlighted the alarming rise in the frequency and intensity of extreme weather events, indicative of climate anomalies, over recent decades [2]. The World Meteorological Organization (WMO) estimated that climate- and water-related extreme events have caused nearly 12,000 disasters from 1970 to 2021, resulting in economic losses exceeding US$ 4.3 trillion [3]. Advances in early warning systems and coordinated disaster management have slashed human casualties over the past 50 years. According to the WMO, over 90% of reported deaths occurred in developing countries, emphasizing the necessity to strengthen disaster early warning systems in vulnerable communities. The recent EREs in the southern Western Ghats (Kerala), India, vividly illustrate this urgency, as these unprecedented and widespread EREs triggered catastrophic floods and landslides, inundating vast areas, displacing millions of people and devastating communities [4]. Since 2018, the windward slopes of the southern Western Ghats (in Kerala) have experienced frequent, widespread EREs, causing extensive flooding (in the lowlands) and numerous landslides (in the highlands) across the region (S1 Fig). The high-magnitude floods of 2018 and 2019, as well as the major landslides - Kavalappara in Malappuram district (2019), Puthumala and Mundakkai-Chooralmala in Wayanad district (2019 and 2024, respectively), Pettimudy in Idukki district (2020) and Kootikkal-Kokkayar in Kottayam and Idukki districts (2021) - were the result of such EREs. These extreme weather-related disasters led to numerous human casualties and substantial economic losses. Given the frequent occurrence of EREs and their societal impacts, an urgent need is to develop advanced disaster early warning systems to safeguard vulnerable communities and critical infrastructure.

Vital to this endeavor is the ability to forecast EREs with sufficient lead time, facilitating timely and targeted interventions to mitigate their disaster potential and impacts. Despite scientific advances in numerical weather prediction capabilities, accurate forecasting of EREs at a regional scale and higher lead times remains challenging as the predictions are embedded with uncertainties generated from different sources [5–7]. Forecasting EREs presents formidable challenges in tropical regions, such as the southern Western Ghats, due to complex ocean-atmospheric interactions, as well as the significant influence of the high-elevation passive margin, facilitating orographic rainfall during the Indian summer monsoon season [8]. Given these challenges, the focus has shifted towards nowcasting, a proactive approach that leverages near real-time observational data to provide short-term forecasts, typically spanning a few hours. Unlike forecasting methods, nowcasting offers the advantage of rapid response, enabling authorities to issue timely warnings and implement mitigation measures prior to the occurrence of EREs. Numerical weather prediction [9–11], radar extrapolation [12–17] and stochastic models [18–20] are some of the standard approaches for nowcasting EREs. These methods rely on high-resolution observation networks, such as Doppler weather radars, automatic weather stations, wind profilers, and radiometers that monitor and deliver reliable weather data at fine spatial and temporal intervals.

Remote sensing-based observations of active meteorological systems have demonstrated great potential for nowcasting EREs in regions with limited ground observations [21–23]. These techniques utilize satellite measurements to monitor cloud development, with cloud top temperature (T) often serving as a critical indicator. For instance, Shukla and coauthors [23] introduced a nowcasting technique that relies on T as a key signature of cloud development. Similarly, Anderson and coauthors [24] developed a probabilistic nowcasting approach in the Sahel region, incorporating historical and real-time satellite data of T to predict convective activity associated with EREs. Further advancements, including the use of brightness temperature and brightness temperature difference methods employed on observations from Himawari-8, have also shown promise for short-term predictions of EREs by selecting optimal thresholds for various rainfall intensities [25]. Despite these advances, achieving high accuracy in nowcasting EREs remains challenging due to the complex dynamics of mesoscale convective systems, which are influenced by numerous factors beyond T [26,27]. One critical aspect overlooked in the current nowcasting methods is the interaction between aerosols and clouds. Aerosols function as cloud condensation nuclei (CCN) for cloud droplets, influencing cloud albedo and cloud coverage and altering cloud microphysics and precipitation processes [28]. The activation of CCN into cloud droplets is a key microphysical characteristic of convective clouds, as it influences the rate at which droplets grow with increasing cloud thickness and their transformation into precipitation-sized particles [29]. Numerous studies have reported the influence of aerosols on precipitation formation. For instance, aerosol pollution/enrichment influences the cloud development and precipitation formation processes [30,31] and modifies the cloud microphysical processes [32]. Thus, understanding aerosol-cloud interactions and their influence on cloud microphysical processes is critical for accurate forecasting/nowcasting of EREs [5]. However, the existing remote sensing-based methods do not consider aerosol-cloud interactions (and consequent changes in cloud microphysical evolution), which contain vital signatures for nowcasting EREs. This study addresses this critical gap in the literature by incorporating knowledge of aerosol-cloud interactions and cloud microphysical processes into remote sensing-based nowcasting of EREs.

Rosenfeld and Lensky [33] suggested that the vertical evolution of cloud effective radius (r_e) indicates the stages of cloud development, with the shape of the T-r_e profile characterizing different cloud microphysical processes. Accordingly, five microphysical stages were identified: (a) condensation-dominated growth of cloud droplets, (b) coalescence-dominated growth of cloud droplets, (c) rainout phase, (d) ice-water mixed phase, and (e) glaciated phase. A deep condensation-dominated zone with a higher concentration of CCN at the cloud base marks aerosol-rich (polluted) conditions. Such aerosol-enriched conditions, combined with moisture convergence, are prime mechanisms leading to EREs [34]. In contrast, cleaner environments are characterized by the coalescence of cloud droplets [34]. Zhu and coauthors [35] reported that the simple yet powerful relationship in the T-r_e profile can be used for parameterization of the effect of aerosols on cloud microstructure and precipitation-forming processes. Thus, we argue that the temporal evolution of the T-r_e profile during cloud development inherits the microphysical characteristics of a particular rainfall event. Further, the microphysical characteristics pertaining to normal- and heavy/extreme-rainfall events are distinguishable with sufficient lead time, thereby facilitating the nowcasting of EREs.

The present study proposes a method that uses satellite-derived T-r_e profiles to nowcast EREs on a regional scale. Our research addresses the following questions: (i) Do the T-r_e profile characteristics differ between normal-, heavy-, and extreme-rainfall events? (ii) Can the microphysical stages of cloud development be determined from the T-r_e profile prior to the event? (iii) Can the characteristic attributes of the T-r_e profile be used to nowcast the probability of occurrence of an ERE? We addressed these research questions by analyzing cloud microphysical data from recent EREs in the southern Western Ghats (Kerala, India) and subsequently developed a nowcasting model. Although the proposed nowcasting model is unbounded by regional extent, its performance evaluation in such a densely populated region experiencing heavy monsoonal rains could contribute towards managing natural disasters by reducing the risks.

2. Material and methods

The manuscript is organized into three sections: (1) a general characterization of EREs over Kerala; (2) an assessment of T-r_e profiles for different rainfall events; and (3) nowcasting EREs using the variables derived from the T-r_e profiles and droplet size distribution of clouds. The details of the data and methodology are described in the subsequent sections.

2.1. Characterization of EREs

The rainfall events of Kerala between 2001 and 2018 were extracted from the India Meteorological Department (IMD) daily gridded rainfall dataset, which has a spatial resolution of 0.25° x 0.25°. The gridded dataset was generated using rainfall data measured over 6,955 rain gauge stations across the Indian mainland, covering a period of 120 years (https://www.imdpune.gov.in/cmpg/Griddata/Rainfall_25_NetCDF.html). Pai and coauthors [36] provide a detailed description of the dataset. The accuracy of the dataset was assessed by comparing the gridded dataset with other existing datasets (IMD 0.5° x 0.5° gridded data and APHRODITE), and the results indicate that the dataset provides a realistic spatial representation of the rainfall pattern across the Western Ghats region. Following the guidelines of the IMD (Forecasting Circular No. 5/2015/3.7), the rainfall events over Kerala were categorized into three groups based on the accumulated daily rainfall depth: normal-, heavy-, and extreme-rainfall events. According to the IMD, rainfall events over the Indian region with a daily rainfall depth exceeding 204.4 mm (corresponding to the 99.9^th percentile in Kerala) were classified as EREs. However, to specifically focus on heavy- and extreme-rainfall events, we classified daily rainfall depths below 115.6 mm (corresponding to the 99^th percentile in Kerala) as normal rainfall events and the transitional range between 115.6 mm and 204.4 mm as heavy rainfall events. This classification enables us to differentiate and analyze the characteristics and impacts of the most extreme rainfall events expected over the region.

2.2. Assessment of T-r_e profile of rainfall events

The evolution of cloud particles through various cloud microphysical stages leading to rainfall can be inferred from the characteristic T-r_e profile for convective systems [33]. The hourly cloud properties (of August 2018), such as T and r_e, were acquired from the National Aeronautics and Space Administration (NASA) Langley Cloud and Radiation Research Group (http://www-angler.larc.nasa.gov). The NASA-Langley cloud and radiation products are generated using the Visible Infrared Solar-infrared Split-Window Technique, Solar-infrared Infrared Split-Window Technique, and Solar-infrared Infrared Near-Infrared Technique [37]. These radiation products are subject to uncertainties primarily due to challenges in detecting thin clouds, resolving overlapping cloud layers, assumptions in cloud microphysics, instrument calibration, and the coarse spatial and temporal resolution, which can affect the accuracy of radiative flux estimates and cloud property retrievals. However, using the T-r_e profile characteristics for nowcasting (explained in subsequent sections) rather than relying on absolute values of cloud properties may help mitigate such uncertainties to some extent. Satellites retrieve r_e by measuring the statistical moment of cloud drop size distribution, as described by Nakajima and King [38] and represented in Eq. 1.

(1)

where n(r) is the concentration of cloud particles having radius, r. For a given temperature, r_e is assumed to be a conserved property as long as the precipitation has not developed [33]. Thus, the temporal evolution of individual cloud particles can be inferred from a single satellite image with multiple cloud particles at different stages of development. The T-r_e relationship was assessed using the following steps [33].

In this study, the sampling of cloud properties during different rainfall events (described in section 2.1) was carried out on spatially and/or temporally independent cloud clusters causing the rainfall. First, RGB image composites were compiled using red for visible reflectance, green for 3.9 µm brightness temperature, and blue for 10.8 µm brightness temperature to identify convective cloud clusters. Rosenfeld and Lensky [33] detailed this visual multispectral classification scheme using the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) observations. Later, Rosenfeld and coauthors [39] compared the microphysical rendering of deep convective clouds using the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the Suomi National Polar-Orbiting Partnership (NPP) and Moderate Resolution Imaging Spectroradiometer (MODIS) observations. The RGB composite image was then used to identify a window containing individual cloud clusters with cloud elements representing all stages of cloud growth, typically containing several thousand pixels. Thus, the sampling covered small, medium, and high-level cumulus clouds and some cloud-free pixels. The 10^th, 25^th, 50^th, 75^th, and 90^th percentiles of r_e for each 1°C interval of T were estimated (as described by [33]), indicating the sampling amount for each 1°C interval of T. The median curve (50^th percentile) may be skewed towards other percentiles if the sampling is limited to specific cloud particles and may not represent the r_e at that T for the entire cloud cluster. With sufficient cloud samples, various characteristics of the T-r_e profile, implying different microphysical stages of cloud development, were extracted from the median curve.

A typical T-r_e curve is illustrated in Fig 1. The specific characteristics of the T-r_e curve that aid in developing the nowcasting procedure are the cloud base temperature (T_b), cloud base particle size (r_b), depth of the diffusional zone (D_z), cloud height with r_e > 14 µm (T₁₄), and mixed-phase initiation height (T_L). In convective clouds, droplets primarily grow by diffusion near the base such that r_e is proportional to D_a, where D_a is the depth above the cloud base approximated by T_b. In polluted (aerosol-rich) environments, a higher number of activated CCN at the cloud base often results in deep diffusional zones, represented by D_z. Coalescence and cloud formation processes amplify particle growth rates, which are crucial for the rainfall process. The coalescence growth of droplets becomes efficient beyond the critical effective radius (≈ 14 µm), which is represented by T₁₄. During the mixed phase, cloud droplets grow faster with an increase in height, identified by a sudden change in the slope of the T-r_e curve. The contribution of mixed-phase processes is essential for an ERE, and the mixed-phase initiation height is represented by T_L.

Download:

Fig 1. Conceptual diagram illustrating various attributes extracted from the T-r_e curve:

Cloud base temperature (T_b), cloud base particle size (r_b), depth of the diffusional zone (D_z), cloud height in terms of temperature exceeding the 14 µm precipitation threshold (T₁₄), and mixed-phase initiation height (T_L).

https://doi.org/10.1371/journal.pclm.0000497.g001

Apart from the T-r_e profile, we also examined the size distribution of cloud droplets during various rainfall events. The analysis of the droplet size distribution of clouds during normal-, heavy-, and extreme-rainfall events revealed distinct distributions (Fig 2). The heavy- and extreme-rainfall events exhibit bimodal curves compared to the normal rainfall events, which show unimodal curves. While the first peak occurs at 14 µm (D₁₄) for all the events, a second peak is observed at 70 µm (D₇₀) for heavy- and extreme-rainfall events. Larger ice particles held by strong updrafts likely cause the second peak in the droplet size distribution. As the mode diameter of the raindrop size distribution of convective storms shifts towards larger drop sizes with increasing rain rates, there is a corresponding decrease in the number concentration of small-sized raindrops by approximately three orders of magnitude [40]. Accordingly, we argued that, apart from the characteristics extracted from the T-r_e profile, information on cloud droplet size distribution would also help in nowcasting. Hence, we used the ratio as an input variable in the nowcasting model.

Download:

Fig 2. Cloud droplet size distribution during normal-, heavy-, and extreme-rainfall events.

The heavy- and extreme-rainfall events exhibit dual peaks compared to normal rainfall events: the first peak is around 14 µm (D₁₄), and the second is around 70 µm (D₇₀).

https://doi.org/10.1371/journal.pclm.0000497.g002

2.3. Nowcasting of EREs

The approach for nowcasting ERE involves classifying convective systems into two groups based on their probability of occurrence, i.e., (i) the probability for the occurrence of an ERE and (ii) the probability for the non-occurrence of an ERE. This classification relies on the non-categorical dependent variables (T_b, r_b, D_z, T₁₄, and T_L, and ) as discussed in Section 2.2. Ideally, this classification is a multi-dimensional data clustering problem, where machine learning techniques are effective tools but require extensive data for model training [41]. Given the rare occurrence of extreme weather events (such as EREs) and the limited availability of observational data [42], pattern recognition tends to be challenging for clustering algorithms. Alternatively, logistic regression and discriminant analysis are simpler and appropriate statistical techniques. However, logistic regression is generally preferred, as discriminant analysis depends on the multivariate normality assumption, which is often unmet [43]. Hence, we used logistic regression modeling for the nowcasting technique in this study. The methodological framework is briefly outlined below:

Identifying normal-, heavy-, and extreme-rainfall events: The rainfall events were categorized, and the geostationary satellite observations of cloud properties, such as T and r_e, were obtained before the rain initiation to generate the corresponding T-r_e profile.
Fitting a logistic regression: Logistic regression (Eq. 2) was fitted to the attributes extracted from the T-r_e profile (T_b, r_b, D_z, T₁₄, and T_L) and the droplet size distribution curve () of all cases of rainfall events.

(2)

where b₀, b₁, b₂, b₃, b₄, and b₅ are the regression coefficients. Estimating the logistic regression coefficients is similar to that used in multiple regression. However, we used the maximum likelihood method instead of ordinary least squares to estimate the model parameters [43].

Validation: The parameters (regression coefficients) estimated in Step 2 were validated using independent events that were not part of the training or calibration of the regression model. The half-hourly gauge calibrated rainfall product from the Integrated Multi-Satellite Retrievals (IMERG [44]) was used to validate the accuracy of the nowcasting technique by comparing the probability of occurrence of EREs with the corresponding rainfall intensity on an hourly time scale.
Nowcasting: Once the coefficients are estimated and validated, the probability of occurrence of heavy- and extreme-rainfall at any given time can be calculated using Eq. 2 with the attributes extracted from the T-r_e profile and the cloud droplet size distribution curve at that time.

Fig 3 illustrates the workflow adopted for developing the nowcasting model, and Table 1 outlines the data used in this study. The accuracy and applicability of the nowcasting technique were demonstrated using data from the EREs in Kerala that occurred between 2018 and 2021, along with an additional analysis of selected events from 2023.

Download:

Table 1. Details of various data used in this study.

https://doi.org/10.1371/journal.pclm.0000497.t001

Download:

Fig 3. Overall workflow of the development of the proposed nowcasting model.

https://doi.org/10.1371/journal.pclm.0000497.g003

3. Results and discussion

3.1. Characterization of EREs over Kerala (2001–2018)

The analysis of the rainfall events over Kerala from 2001 to 2018 indicates that the region experienced 222 heavy rainfall events (daily rainfall depth > 115.6 mm) during the period, including 32 EREs (i.e., daily rainfall depth > 204.4 mm) (Fig 4A). Although the probability for the occurrence of EREs is very low (i.e., exceedance probability less than 0.1%), recent years (2018–2024) witnessed multiple EREs across Kerala due to large-scale moisture convergence below 800 hPa [45]. The EREs of August 2018 recorded a maximum daily rainfall of 319 mm, making it one of the most extreme events recorded in the last century. The monthly rainfall of Kerala in August 2018 was 96% greater than normal, and the region received 123% greater than normal rainfall during August 2019. Of the 32 EREs in Kerala between 2001 and 2018, roughly 65% of the events occurred during the Indian summer monsoon season (mainly in July and August) (Fig 4B). While most regions of Kerala experienced EREs during the period, frequent occurrences were clustered around two locations (i) 9° 30’ N and (ii) 12° N latitudes (Fig 4C). However, as a departure from this general pattern, Kerala witnessed widespread EREs in 2018 and 2019. These EREs were largely underestimated by some numerical weather prediction models, such as the European Centre for Medium-Range Weather Forecasts (ECMWF) (S2 Fig).

Download:

Fig 4. Rainfall characteristics over the study area.

(A) one-day maximum rainfall (2001 to 2018) based on the IMD gridded (0.25° x 0.25°) rainfall dataset for the Indian sub-continent, (B) temporal distribution* of normal-, heavy-, and extreme-rainfall events in terms of their frequencies of occurrence, and (C) frequency of EREs in the analyzed domain (enclosed by the black rectangle) during the period. *Only rainfall values exceeding 64.5 mm are plotted in (b).

https://doi.org/10.1371/journal.pclm.0000497.g004

3.2. Variation of the T-r_e profiles between normal, heavy, and extreme rainfall events

The daily plots of the T-r_e profile were analyzed to study the vertical evolution of cloud particles. Fig 5 shows three cases of cloud clusters resulting in normal- (Fig 5A), heavy- (Fig 5B), and extreme-rainfall events (Fig 5C) in August 2018. The T-r_e profile during EREs is characterized by a condensation-dominated diffusional zone above the cloud base (Fig 5F). The diffusional zone is formed when the expansion of rising moist air causes an increase in the ambient supersaturation, and the excess vapor, thereby condensing on the aerosol particles, to form activated cloud droplets. In the diffusional zone, the r_e is not sufficiently large enough (r_e < 14 µm) to support the warm rain process [46,47]. The number of activated cloud droplets (N_a) at the cloud base for a given updraft velocity is determined by the concentration of CCN [34]. In highly polluted conditions, a large number of aerosol particles provides a sufficient concentration of CCN for cloud formation. Each activated cloud droplet acts as a tiny sink, adsorbing excess water vapor. Higher N_a produces more but smaller cloud droplets with a narrow size distribution, resulting in a less efficient collision-coalescence process. The cloud droplets grow much slower in the diffusional zone, reaching colder altitudes with smaller r_e, delaying collision-coalescence. As D_z is primarily determined by N_a [48], the deep diffusional zone during EREs (Fig 5F) indicates higher N_a, which implies a polluted (aerosol-rich) scenario. Although shallower, the diffusional zone above the cloud base is associated with heavy rainfall events as well (Fig 5E).

Download:

Fig 5. Variation of the T-r_e profiles between normal-, heavy-, and extreme-rainfall events.

Cloud RGB composite image during (A) normal-, (B) heavy-, and (C) extreme-rainfall events. The color composition includes red for visible reflectance, green for 3.9 µm solar infrared temperature, and blue for 10.8 µm brightness temperature. The cloud clusters analyzed are bounded by the white polygon. Vertical cloud microphysical structures of (D) normal-, (E) heavy-, and (F) extreme-rainfall events are shown. The 10^th, 25^th, 50^th, 75^th, and 90^th percentiles of r_e are plotted. The vertical bars denote the vertical extent of the microphysical zones labeled as D - diffusion, C - collision-coalescence, R - warm rainout, M - mixed phase, and G - glaciation. The continuous red line is for visual representation only and does not carry any reference or implication related to the analysis or data presented.

https://doi.org/10.1371/journal.pclm.0000497.g005

Normal rainfall events are devoid of any diffusional zone and are characterized by the collision-coalescence zone above the cloud base. Faster growth in r_e is observed until it stabilizes around 14 µm. This scenario typically occurs in a clean environment with fewer concentrations of CCN, which produces low N_a. Konwar and coauthors [49] observed similar cases of clouds developed in less polluted conditions, leading to low N_a. In such environments, the limited activated cloud droplets absorb most of the excess vapor and grow quickly in size, resulting in fewer but larger cloud droplets with a higher likelihood of collision and coalescence. The stabilization of r_e at around 14 µm with further vertical growth during normal rainfall events is an indication of the warm rain process. Using a cloud parcel model, Rosenfeld and coauthors [50] showed that the rainwater fraction increases sharply for r_e above 14 µm. This is further supported by the simultaneous stabilization of the 75^th percentile of r_e at a radius of 25 µm (Fig 5D). The presence of cloud droplets with the r_e of 25 µm disrupts the colloidal stability of the cloud, resulting in an efficient collision-coalescence process [51]. The warm rain mechanism thus produces larger cloud droplets from the cloud tops, balancing additional coalescence and rendering the observed stabilization of r_e [34]. Such a warm rain process can also lead to rain washout of aerosols, thereby limiting CCN for further cloud growth.

Further evidence of a polluted (aerosol-enriched) scenario during EREs is indicated by the delayed onset of the coalescence zone above the freezing level (Fig 5F). Using aircraft-based measurements, Konwar and coauthors [49] reported similar cases of delayed coalescence zone up to the −5°C isotherm in highly polluted environments over the Indo-Gangetic plains. Supercooled raindrops between −12°C and −17°C are the primary initiators of precipitation in such clouds, suggesting the presence of ice nuclei or giant CCN seeding the clouds [34]. Insoluble aerosol particles, such as desert dust, are the major natural sources of ice nuclei. Konwar and coauthors [52] also noted the presence of giant CCN and dust particles (S3 Fig) in a moist environment within the convective clouds developed over the geographical extent of the present study.

A similarity is noted between the extreme- and normal-rainfall events in the glaciation zone (Fig 5), where clouds in both cases are glaciated near −20°C. However, the underlying processes driving this glaciation may differ between the cases. While a high concentration of desert dust facilitates glaciation in clouds near −20°C [53], secondary ice formation processes, such as ice multiplication, can also cause glaciation at this temperature range [54]. Ice multiplication becomes significant when cloud droplets grow larger than 12 µm (r_e) at a temperature range of −3°C to −8°C [55]. This is a characteristic of normal rainfall events, where early collision-coalescence promotes the cloud drops to a larger radius (r_e > 12 µm). In such cases, glaciation occurs at a much warmer temperature than those induced by desert dust [56]. The occurrence of smaller cloud droplets (within the temperature range of −3°C to −8°C) and glaciation colder than −20°C during EREs (Fig 5F) suggest the presence of dust aerosols and subsequent ice nuclei concentrations.

The updraft velocity does not affect D_z for a given N_a, but it can delay the increase of r_e with height above the height of rain initiation (h (r_e ~ 14 µm)) due to less time for raindrop formation in the fast-rising parcel [57]. Such strong updrafts are evident from the steeper collision-coalescence zone (Fig 5F) during EREs. Interestingly, Meenu and coauthors [58] also reported such strong updrafts within the deep convective clouds associated with the orographic lifting of airmass during the EREs of August 2018. Moreover, the deeper diffusional zone of EREs dominated by condensation increases the column loading of condensed water and releases latent heat, resulting in stronger updraft velocity. In such cases, the r_e reaches the rain threshold (14 µm) at heights colder than 0°C isotherms, making most cloud water available for mixed-phase processes. These processes release latent heat, further invigorating the updraft velocity [32,59]. Meenu and coauthors [58] highlighted similar latent heat-positive feedback while analyzing the extreme event on 9 August 2018 using multi-satellite observations and numerical simulations. A sudden decrease in the r_e at the temperature range between −20°C and −18°C is consistently observed in all the rainfall events (Fig 5). This could be due to the formation of ice crystals, which commonly appear when the temperature is between −10°C and −15°C. However, a detailed analysis is required to reinforce this inference. The relatively low vapor pressure of ice causes water vapor to diffuse from many cloud droplets to fewer ice crystals [51], thus reducing the cloud top r_e. Once the ice crystals attain sufficient sizes, cold cloud growth mechanisms are initiated, resulting in faster growth of r_e with a further increase in height.

3.3. Nowcasting of EREs

The logistic regression model (Eq. 2) was calibrated using the rainfall events of August 2018, a period during which Kerala recorded widespread occurrences of heavy- and extreme-rainfall events. A total of 45 rainfall events that are spatially and/or temporarily independent were identified and characterized in terms of their T-r_e profile and droplet size distribution, of which 19 are heavy- and extreme-rainfall events and 26 are normal rainfall events. All the data was used to train the logistic regression model. Although different parameters (T_b, r_b, D_z, T₁₄, and T_L, and ) were used in the modeling process; D_z is the most sensitive parameter (p < 0.05) during the model training. The estimated probabilities of normal- and heavy/extreme-rainfall events (by the calibrated model; Eq. 2) are shown in Fig 6A, where a clear distinction between the normal- and heavy/extreme-rainfall events is evident in their probabilities. A probability cut-off of 0.5 was applied, and the overall skill score was 93% for heavy/extreme rainfall events. Further, the model was validated with an entirely different set of rainfall events from 2018 to 2021, and the efficiency of the model performance is evident from the receiver operating characteristic curve (Fig 6B), which shows an area under the curve of 0.97. The results of the model validation confirm the appropriateness of the cloud microphysical variables used and the suitability of the logistic regression model for nowcasting EREs.

Download:

Fig 6. Modelling the occurrence of EREs.

(A) modeled probability for normal- and heavy/extreme-rainfall events during August 2018, and (B) the receiver operating characteristic curve for the logistic regression model.

https://doi.org/10.1371/journal.pclm.0000497.g006

3.3.1. Model evaluation.

The robustness of the nowcasting model was evaluated using different evaluation metrics and by assessing the model uncertainty. Bootstrapping was performed to characterize model uncertainty within the 95% confidence intervals for the parameters of the logistic regression model (Table 2). Subsequently, Monte Carlo simulations were performed to simulate the model. The results of the simulations, based on 10,000 samples, reveal that the model could accurately predict true positive (heavy/extreme rainfall events) for 94% of the cases, implying minimal model parametric uncertainty. Further, various statistical evaluation metrics, such as accuracy, precision, recall, F1-score, Brier Score (BS), and ranked probability score (RPS), were computed to assess the model performance (Table 3). All these metrics also suggest the capability of the model to accurately nowcast heavy/extreme rainfall events.

Download:

Table 2. 95% confidence intervals for the parameters of the nowcasting model.

https://doi.org/10.1371/journal.pclm.0000497.t002

Download:

Table 3. Metrics used to assess the accuracy of the nowcasting model.

https://doi.org/10.1371/journal.pclm.0000497.t003

Hourly rainfall data of multiple EREs between 2018 and 2023 was used to assess the efficiency of the model and lead time in nowcasting the EREs. It has been observed that the nowcasting model performed well in all cases. However, we presented the results of the model performance for a couple of events in 2018 and a recent one in 2023 for brevity (Fig 7 and S4 Fig). On 8 August 2018, Kerala received heavy rainfall, followed by an ERE on 9 August. The rainfall rate of IMERG data indicates that the rainfall intensified around 9:00 PM on 7 August (Fig 7A). The modeled probability for the occurrence of a heavy/extreme rainfall event is also shown in Fig 7A. The probability for a heavy/extreme rainfall event remains low until 2:00 PM on 7 August 2018, after which it shows a high probability for occurrence. The modeled probability decreases during the early hours of 8 August, followed by a decrease in the rainfall intensity. However, from 8:00 AM onwards, the model predicts a high probability until 9 August, which is indicative of the ERE on 9 August 2018.

Download:

Fig 7. Nowcasting of EREs.

Nowcasting the heavy/extreme rainfall events of (A) 8 August 2018 and (B) 15 August 2018. The half-hourly gauge calibrated precipitation product from the IMERG was used to plot the temporal variation of rainfall intensity.

https://doi.org/10.1371/journal.pclm.0000497.g007

Similarly, the rainfall rate of IMERG data indicates that Kerala received heavy rainfall on 14 August 2018, followed by EREs on 15, 16, and 17 August. The temporal pattern of the rainfall rate shows intensification around 3:00 AM on 14 August (Fig 7B). The modeled probability for a heavy/extreme rainfall event remains low until 12:00 PM on 13 August 2018, after which the model predicts a high probability for such an event. The temporal variability of the modeled probability values for the first ERE (8–9 August 2018) shows the potential for forecasting the occurrence of a heavy/extreme rainfall event seven hours before the rainfall intensification. Similarly, from 15–17 August 2018, the technique provides a warning for the occurrence of heavy/extreme rainfall events 15 hours before the rainfall intensification. The performance of the nowcasting model for a recent ERE (on 14 August 2023) in Thiruvananthapuram, the southernmost district of Kerala, is included in the supplementary material (S4 Fig). The correlation between the rainfall intensity and the modeled probability indicates that the proposed method can nowcast EREs sufficiently ahead of their intensification, making it suitable for developing an early warning system for EREs. However, we noted a few limitations in nowcasting EREs associated with tropical cyclonic storms. The model accurately predicted heavy/extreme rainfall events during the tropical cyclonic storm in Kerala in 2021, but the nowcast lead time was very short (one hour). However, the model can be improved by incorporating remotely sensed real-time data of atmospheric moisture content and wind convergence. Further, the absence of satellite data during nighttime restricts the nowcasting only to daytime.

3.4. Further evidence of the influence of aerosols and moisture availability on EREs

The EREs of August 2018 were analyzed in detail to understand the effect of moisture availability and aerosol loading on their occurrences. For that, the rainfall events from 2001-2018 were analyzed along with the TPWV and aerosol loading over the study region. The TPWV data serves as a proxy for impending heavy rainfall activity. The TPWV data was extracted from the MERRA-2 [60] reanalysis data, available at a spatial resolution of 0.5° x 0.5° and a temporal resolution of 1 hour (Table 1). The role of aerosols in the occurrence of EREs was investigated by analyzing the AOD data from the MODIS instrument onboard Terra. Numerous researchers have extensively validated the MODIS AOD products [61–64]. We used the MODIS-derived daily AOD at 550 nm from Collection 6.1 (C6.1) level 3 products (1° x 1°) (Table 1). The results of the analysis are presented in Fig 8.

Download:

Fig 8. Effect of moisture availability and aerosol loading on EREs.

(A) variability of moisture availability (in terms of TPWV) during normal-, heavy-, and extreme-rainfall events from 2001 to 2018, (B) daily mean values of TPWV, rainfall, and AOD during August 2018, and (C) mean rainfall profile for the different percentile classes of AOD and TPWV.

https://doi.org/10.1371/journal.pclm.0000497.g008

Moisture availability, governed by various synoptic-scale features, is one of the major ingredients conducive to the occurrence of EREs [65]. Temporal variability of TPWV during normal-, heavy-, and extreme-rainfall events (during the period 2001–2018) is shown in Fig 8A. It is observed that heavy- and extreme-rainfall events are characterized by relatively larger TPWV values than normal rainfall events regardless of the overlaps between them. Obviously, a substantial moisture influx is necessary to form precipitating clouds that could yield heavy rainfall. Previous studies suggest that low-level jets (especially when originating from moisture-laden regions like warm oceans) and atmospheric rivers (formed when moisture-laden air is concentrated into narrow, elongated bands by large-scale atmospheric dynamics, such as wind patterns and pressure systems) are the major large-scale meteorological features responsible for large-scale atmospheric moisture transport during heavy- and extreme-rainfall events in the region [66,67]. Intensification of the monsoon low-level jet, westerly wind depth, zonal water vapor flux, horizontal wind shear, and cyclonic vorticity can bring a sufficiently large supply of moisture and trigger strong convection, resulting in the occurrence of EREs during the Indian summer monsoon season [67]. The geographic setting of Kerala in the Indian peninsula along the western coast ensures sufficient moisture supply during the monsoon season, with an increasing trend of near-surface water vapor over the tropical Indian Ocean [68]. The significance of the anomalously large amount of moisture supply from the neighboring oceans in the EREs of August 2018 in Kerala was demonstrated by various researchers [66,69]. However, all the events with higher TPWV values did not result in heavy- or extreme-rainfall events (Fig 8A), suggesting that other factors, such as the concentration of CCN in the cloud formation processes, also play a crucial role in producing heavy- or extreme-rainfall events.

The analysis of the wind trajectories of August 2018 indicates the transport of aerosols from the neighboring ocean, the Arabian Peninsula, and the Indo-Gangetic plains (S3 Fig). This implies the crucial role of aerosols, such as dust, mixed (dust-anthropogenic) aerosols, and sea salt, in triggering EREs on a regional scale. Atmospheric aerosols can alter the cloud microphysical properties, including droplet concentration, size distribution, and cloud lifetime, ultimately influencing rainfall formation [70,71]. It is important to note that droplet formation driven by available atmospheric moisture depends on the fraction of aerosol particles that serve as CCN. However, the interaction between moisture availability and aerosol concentration is nonlinear and complex.

Temporal variability of the AOD, TPWV, and rainfall during August 2018 is shown in Fig 8B. Higher concentration of aerosols combined with relatively low moisture supply tends to form a large number of smaller cloud droplets, inhibiting processes leading to the occurrence of rainfall. Such conditions were noted on 3 and 4 August 2018 (Fig 8B). Rosenfeld [72] described a similar suppressing effect: increased CCN shuts off the warm rain process in tropical convective clouds. Indeed, such cases lead to low collision rates delaying raindrop formation [73]. On the contrary, smaller AOD with sufficient moisture supply leads to the formation of fewer, larger droplets. In those cases, the ambient supersaturation is weakly consumed by the limited CCN, resulting in faster growth of limited droplets that fall quickly as weak rain [74]. Examples of scenarios leading to normal rainfall events include 1, 7, 11, and 13 August 2018. High rainfall events are characterized by significantly higher AOD and sufficient moisture supply. An increase in rainfall is observed with an increase in the AOD under higher TPWV conditions (Fig 8C), while under low TPWV conditions, an increase in the AOD results in low rainfall due to the formation of a larger number of smaller droplets. Increased rainfall with an increase in aerosol concentrations in deep clouds having high liquid-water content was reported by Li and coauthors [59]. Scenarios of increased surface rainfall over polluted (aerosol-rich) regions during the Indian summer monsoon season were also shown by numerous researchers [75,76].

The independent and interaction effects of moisture availability and aerosol loading on the heavy/extreme rainfall were assessed using the two-way analysis of variance (ANOVA) on the 18 years (2001–2018) data during the Indian summer monsoon season (Table 4). The results show the significant role of the two variables (AOD and TPWV) and their interaction on the occurrence of rainfall activity in the region (p < 0.001; Table 4). Importantly, aerosol concentration over Kerala shows a significant increasing trend (S5 Fig), which has the potential to invigorate cloud formation and intensify rain rates [77]. The increase in aerosol loading can impact the radiation budget [78,79], cloud lifetime, and microphysics [80,81]. Hence, the occurrence of EREs over the region is expected to increase with the increasing atmospheric water holding capacity [82] and the increasing trend of TPWV over the Indian Ocean [83] in response to global climate warming.

Download:

Table 4. Independent and interaction effects of aerosols and moisture availability on rainfall over Kerala.

https://doi.org/10.1371/journal.pclm.0000497.t004

4. Summary and conclusions

This study proposed a method to nowcast heavy- and extreme-rainfall events by analyzing the early microphysical signatures of mesoscale convective clouds. By leveraging various characteristics derived from the T-r_e profile (representing different microphysical stages of cloud development), cloud droplet size distribution, and logistic regression modeling, the nowcasting method effectively estimates the probability of the occurrence of EREs. The robustness of the technique was demonstrated by accurately nowcasting different EREs across the southern Western Ghats (Kerala, India) in recent years. The analysis of the T-r_e profiles of the normal-, heavy-, and extreme-rainfall events of August 2018 indicated a polluted (aerosol-enriched) scenario. The deep diffusional zone and the delayed onset of the coalescence zone above the freezing level also indicate the significance of aerosols during the EREs. The proposed model showed distinctive capability for nowcasting EREs at a regional scale with an overall skill score of 93% and a lead time of at least six hours. Arguably, the study highlighted the significant role of the interactions among aerosol loading and moisture availability (in addition to their independent effects) in the occurrence of EREs. The results of this study emphasize the requirement for systematic and long-term monitoring and characterization of aerosols to understand their role in the EREs of the region. The results of this study help implement advanced early warning systems and better policies to mitigate the impacts of EREs and associated natural hazards.

Supporting information

S1 Fig. Image of the flood-affected areas over Kerala derived from the Sentinel-1 SAR data during August 2018.

https://doi.org/10.1371/journal.pclm.0000497.s001

(DOCX)

S2 Fig. Daily rainfall over a grid centered at 9.5°N and 76.75°E during August 2018.

https://doi.org/10.1371/journal.pclm.0000497.s002

(DOCX)

S3 Fig. Wind trajectory clusters from a central location (10.5°N, 76.3°E) over Kerala during 2001.

https://doi.org/10.1371/journal.pclm.0000497.s003

(DOCX)

S4 Fig. Nowcasting the heavy/extreme rainfall events of 14 October 2023 in Thiruvananthapuram, Kerala.

https://doi.org/10.1371/journal.pclm.0000497.s004

(DOCX)

S5 Fig. Temporal variation of AOD over Kerala from 2001 to 2018.

https://doi.org/10.1371/journal.pclm.0000497.s005

(DOCX)

References

1. Kumar S, Silva Y, Moya-Álvarez AS, Martínez-Castro D. Seasonal and Regional Differences in Extreme Rainfall Events and Their Contribution to the World’s Precipitation: GPM Observations. Advances in Meteorology. 2019;2019:1–15.
- View Article
- Google Scholar
2. IPCC. Weather and Climate Extreme Events in a Changing Climate. In: Intergovernmental Panel on Climate C, editor. Climate Change 2021 – The Physical Science Basis: Working Group I Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press; 2023. p. 1513–766.
3. WMO. Economic costs of weather-related disasters soars but early warnings save lives. 2023 [13 March 2025. ]. Available from: https://wmo.int/news/media-centre/economic-costs-of-weather-related-disasters-soars-early-warnings-save-lives
- View Article
- Google Scholar
4. KSCSTE. Report of Committee Examine the Causes of Repeated Extreme Heavy rainfall events, Subsequent floods and Landslides and to Recommend Appropriate Policy Responses. Thiruvananthapuram: Kerala State Council for Science, Technology and Environment; 2019.
5. Mao J, Ping F, Yin L, Qiu X. A Study of Cloud Microphysical Processes Associated With Torrential Rainfall Event Over Beijing. JGR Atmospheres. 2018;123(16):8768–91.
- View Article
- Google Scholar
6. Srinivas CV, Yesubabu V, Hari Prasad D, Hari Prasad KBRR, Greeshma MM, Baskaran R, et al. Simulation of an extreme heavy rainfall event over Chennai, India using WRF: Sensitivity to grid resolution and boundary layer physics. Atmospheric Research. 2018;210:66–82.
- View Article
- Google Scholar
7. Wang R, Zhu Y, Qiao F, Liang X-Z, Zhang H, Ding Y. High-resolution Simulation of an Extreme Heavy Rainfall Event in Shanghai Using the Weather Research and Forecasting Model: Sensitivity to Planetary Boundary Layer Parameterization. Adv Atmos Sci. 2021;38(1):98–115.
- View Article
- Google Scholar
8. Ray P, Zhang C, Moncrieff M, Dudhia J, Li T. Tropical channel model. In: Druyan L, editor. Climate models. Rijeka: IntechOpen; 2012.
9. Radhakrishna B, Zawadzki I, Fabry F. Predictability of Precipitation from Continental Radar Images. Part V: Growth and Decay. Journal of the Atmospheric Sciences. 2012;69(11):3336–49.
- View Article
- Google Scholar
10. Sun J, Xue M, Wilson JW, Zawadzki I, Ballard SP, Onvlee-Hooimeyer J, et al. Use of NWP for Nowcasting Convective Precipitation: Recent Progress and Challenges. Bulletin of the American Meteorological Society. 2014;95(3):409–26.
- View Article
- Google Scholar
11. Wilson JW, Crook NA, Mueller CK, Sun J, Dixon M. Nowcasting Thunderstorms: A Status Report. Bull Amer Meteor Soc. 1998;79(10):2079–99.
- View Article
- Google Scholar
12. Fox NI, Wikle CK. A Bayesian Quantitative Precipitation Nowcast Scheme. Weather and Forecasting. 2005;20(3):264–75.
- View Article
- Google Scholar
13. He S, Raghavan SV, Nguyen NS, Liong S. Ensemble rainfall forecasting with numerical weather prediction and radar‐based nowcasting models. Hydrological Processes. 2013;27(11):1560–71.
- View Article
- Google Scholar
14. Li PW, Lai EST. Applications of radar‐based nowcasting techniques for mesoscale weather forecasting in Hong Kong. Meteorological Applications. 2004;11(3):253–64.
- View Article
- Google Scholar
15. Ravuri S, Lenc K, Willson M, Kangin D, Lam R, Mirowski P, et al. Skilful precipitation nowcasting using deep generative models of radar. Nature. 2021;597(7878):672–7. pmid:34588668
- View Article
- PubMed/NCBI
- Google Scholar
16. Sivasubramaniam K, Sharma A, Alfredsen K. Merging radar and gauge information within a dynamical model combination framework for precipitation estimation in cold climates. Environmental Modelling & Software. 2019;119:99–110.
- View Article
- Google Scholar
17. Xiang Y, Ma J, Wu X. A Precipitation Nowcasting Mechanism for Real-World Data Based on Machine Learning. Mathematical Problems in Engineering. 2020;2020:1–11.
- View Article
- Google Scholar
18. Metta S, von Hardenberg J, Ferraris L, Rebora N, Provenzale A. Precipitation Nowcasting by a Spectral-Based Nonlinear Stochastic Model. Journal of Hydrometeorology. 2009;10(5):1285–97.
- View Article
- Google Scholar
19. Pulkkinen S, Chandrasekar V, Harri A-M. Stochastic Spectral Method for Radar-Based Probabilistic Precipitation Nowcasting. Journal of Atmospheric and Oceanic Technology. 2019;36(6):971–85.
- View Article
- Google Scholar
20. Sirangelo B, Versace P, De Luca DL. Rainfall nowcasting by at site stochastic model P.R.A.I.S.E. Hydrol Earth Syst Sci. 2007;11(4):1341–51.
- View Article
- Google Scholar
21. Fortelli A, Scafetta N, Mazzarella A. Nowcasting and real-time monitoring of heavy rainfall events inducing flash-floods: an application to Phlegraean area (Central-Southern Italy). Nat Hazards. 2019;97(2):861–89.
- View Article
- Google Scholar
22. Kamarianakis Y, Feidas H, Kokolatos G, Chrysoulakis N, Karatzias V. Evaluating remotely sensed rainfall estimates using nonlinear mixed models and geographically weighted regression. Environmental Modelling & Software. 2008;23(12):1438–47.
- View Article
- Google Scholar
23. Shukla BP, Kishtawal CM, Pal PK. Satellite-Based Nowcasting of Extreme Rainfall Events Over Western Himalayan Region. IEEE J Sel Top Appl Earth Observations Remote Sensing. 2017;10(5):1681–6.
- View Article
- Google Scholar
24. Anderson SR, Cole SJ, Klein C, Taylor CM, Diop CA, Kamara M. Nowcasting convective activity for the Sahel: A simple probabilistic approach using real‐time and historical satellite data on cloud‐top temperature. Quart J Royal Meteoro Soc. 2024;150(759):597–617.
- View Article
- Google Scholar
25. Jumianti N, Marzuki M, Yusnaini H, Ramadhan R, Harjupa W, Saufina E, et al. Prediction of extreme rain in Kototabang using Himawari-8 satellite based on differences in cloud brightness temperature. Remote Sensing Applications: Society and Environment. 2024;33:101102.
- View Article
- Google Scholar
26. Ahmed F, Schumacher C. Convective and stratiform components of the precipitation‐moisture relationship. Geophysical Research Letters. 2015;42(23).
- View Article
- Google Scholar
27. Roca R, Fiolleau T. Extreme precipitation in the tropics is closely associated with long-lived convective systems. Commun Earth Environ. 2020;1(1).
- View Article
- Google Scholar
28. Andreae MO, Rosenfeld D, Artaxo P, Costa AA, Frank GP, Longo KM, et al. Smoking rain clouds over the Amazon. Science. 2004;303(5662):1337–42. pmid:14988556
- View Article
- PubMed/NCBI
- Google Scholar
29. Freud E, Rosenfeld D, Kulkarni JR. Resolving both entrainment-mixing and number of activated CCN in deep convective clouds. Atmos Chem Phys. 2011;11(24):12887–900.
- View Article
- Google Scholar
30. Guo L, Turner AG, Highwood EJ. Impacts of 20th century aerosol emissions on the South Asian monsoon in the CMIP5 models. Atmos Chem Phys. 2015;15(11):6367–78.
- View Article
- Google Scholar
31. Zhao C, Wang Y, Yang Q, Fu R, Cunnold D, Choi Y. Impact of East Asian summer monsoon on the air quality over China: View from space. J Geophys Res. 2010;115(D9).
- View Article
- Google Scholar
32. Rosenfeld D, Lohmann U, Raga GB, O’Dowd CD, Kulmala M, Fuzzi S, et al. Flood or drought: how do aerosols affect precipitation?. Science. 2008;321(5894):1309–13. pmid:18772428
- View Article
- PubMed/NCBI
- Google Scholar
33. Rosenfeld D, Lensky IM. Satellite–Based Insights into Precipitation Formation Processes in Continental and Maritime Convective Clouds. Bull Amer Meteor Soc. 1998;79(11):2457–76.
- View Article
- Google Scholar
34. Rosenfeld D. Chapter 6 - Cloud-Aerosol-Precipitation Interactions Based of Satellite Retrieved Vertical Profiles of Cloud Microstructure. In: Islam T, Hu Y, Kokhanovsky A, Wang J, editors. Remote sensing of aerosols, clouds, and precipitation. Amsterdam: Elsevier. 2018. p. 129–52.
35. Zhu Y, Rosenfeld D, Yu X, Li Z. Separating aerosol microphysical effects and satellite measurement artifacts of the relationships between warm rain onset height and aerosol optical depth. JGR Atmospheres. 2015;120(15):7726–36.
- View Article
- Google Scholar
36. Pai DS, Rajeevan M, Sreejith OP, Mukhopadhyay B, Satbha NS. Development of a new high spatial resolution (0.25° × 0.25°) long period (1901-2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. MAUSAM. 2014;65(1):1–18.
- View Article
- Google Scholar
37. Minnis P, Nguyen L, Palikonda R, Heck P, Spangenberg D, Doelling D, et al. Near-real time cloud retrievals from operational and research meteorological satellites. SPIE; 2008.
38. Nakajima T, King MD. Determination of the Optical Thickness and Effective Particle Radius of Clouds from Reflected Solar Radiation Measurements. Part I: Theory. J Atmos Sci. 1990;47(15):1878–93.
- View Article
- Google Scholar
39. Rosenfeld D, Liu G, Yu X, Zhu Y, Dai J, Xu X, et al. High-resolution (375 m) cloud microstructure as seen from the NPP/VIIRS satellite imager. Atmos Chem Phys. 2014;14(5):2479–96.
- View Article
- Google Scholar
40. Thomas A, Kanawade VP, Chakravarty K, Srivastava AK. Characterization of raindrop size distributions and its response to cloud microphysical properties. Atmospheric Research. 2021;249:105292.
- View Article
- Google Scholar
41. Sarker IH. Machine Learning: Algorithms, Real-World Applications and Research Directions. SN Comput Sci. 2021;2(3):160. pmid:33778771
- View Article
- PubMed/NCBI
- Google Scholar
42. WMO. Guidelines for Nowcasting Techniques. Geneva: World Meteorological Organization; 2017.
43. Hair JF Jr., Black WC, Babin BJ, Anderson RE. Multivariate Data Analysis. 7 ed. New York: Pearson Prentice Hall; 2010.
44. Huffman G, Stocker E, Bolvin D, Nelkin E, Tan J. Gpm imerg final precipitation L3 half hourly 0.1 degree x 0.1 degree v06. 2019.
45. Mukhopadhyay P, Bechtold P, Zhu Y, Murali Krishna RP, Kumar S, Ganai M, et al. Unraveling the Mechanism of Extreme (More than 30 Sigma) Precipitation during August 2018 and 2019 over Kerala, India. Weather and Forecasting. 2021;36(4):1253–73.
- View Article
- Google Scholar
46. Freud E, Rosenfeld D. Linear relation between convective cloud drop number concentration and depth for rain initiation. J Geophys Res. 2012;117(D2).
- View Article
- Google Scholar
47. Suzuki K, Nakajima TY, Stephens GL. Particle Growth and Drop Collection Efficiency of Warm Clouds as Inferred from Joint CloudSat and MODIS Observations. Journal of the Atmospheric Sciences. 2010;67(9):3019–32.
- View Article
- Google Scholar
48. Rosenfeld D, Andreae MO, Asmi A, Chin M, de Leeuw G, Donovan DP, et al. Global observations of aerosol-cloud-precipitation-climate interactions. Rev Geophys. 2014;52(4):750–808.
- View Article
- Google Scholar
49. Konwar M, Maheskumar RS, Kulkarni JR, Freud E, Goswami BN, Rosenfeld D. Aerosol control on depth of warm rain in convective clouds. J Geophys Res. 2012;117(D13).
- View Article
- Google Scholar
50. Rosenfeld D, Lahav R, Khain A, Pinsky M. The role of sea spray in cleansing air pollution over ocean via cloud processes. Science. 2002;297(5587):1667–70. pmid:12183635
- View Article
- PubMed/NCBI
- Google Scholar
51. Lamb D. Cloud microphysics. In: Holton JR, editor. Encyclopedia of Atmospheric Sciences. Oxford: Academic Press; 2003. p. 459–67.
52. Konwar M, Das SK, Deshpande SM, Chakravarty K, Goswami BN. Microphysics of clouds and rain over the Western Ghat. JGR Atmospheres. 2014;119(10):6140–59.
- View Article
- Google Scholar
53. Ansmann A, Tesche M, Althausen D, Müller D, Seifert P, Freudenthaler V, et al. Influence of Saharan dust on cloud glaciation in southern Morocco during the Saharan Mineral Dust Experiment. J Geophys Res. 2008;113(D4).
- View Article
- Google Scholar
54. Rosenfeld D, Woodley W. Deep convective clouds with sustained supercooled liquid water down to −37.5 degrees C. Nature. 2000;405(6785):440–2. pmid:10839535
- View Article
- PubMed/NCBI
- Google Scholar
55. Hallett J, Mossop SC. Production of secondary ice particles during the riming process. Nature. 1974;249(5452):26–8.
- View Article
- Google Scholar
56. Rosenfeld D, Yu X, Liu G, Xu X, Zhu Y, Yue Z, et al. Glaciation temperatures of convective clouds ingesting desert dust, air pollution and smoke from forest fires. Geophys Res Lett. 2011;38(21):n/a-n/a.
- View Article
- Google Scholar
57. Rosenfeld D, Fromm M, Trentmann J, Luderer G, Andreae MO, Servranckx R. The Chisholm firestorm: observed microstructure, precipitation and lightning activity of a pyro-cumulonimbus. Atmos Chem Phys. 2007;7(3):645–59.
- View Article
- Google Scholar
58. Meenu S, Gayatri K, Malap N, Murugavel P, Samanta S, Prabha TV. The physics of extreme rainfall event: An investigation with multisatellite observations and numerical simulations. Journal of Atmospheric and Solar-Terrestrial Physics. 2020;204:105275.
- View Article
- Google Scholar
59. Li Z, Niu F, Fan J, Liu Y, Rosenfeld D, Ding Y. Long-term impacts of aerosols on the vertical development of clouds and precipitation. Nature Geosci. 2011;4(12):888–94.
- View Article
- Google Scholar
60. (GMAO) GMaAO. MERRA-2 inst1_2d_asm_Nx: 2d,1-Hourly,Instantaneous,Single-Level,Assimilation,Single-Level Diagnostics V5.12.4. In: DISC) GESDaISCG, editor. 5.12.4 ed. Greenbelt, MD, USA; 2015.
61. Choudhry P, Misra A, Tripathi SN. Study of MODIS derived AOD at three different locations in the Indo Gangetic Plain: Kanpur, Gandhi College and Nainital. Ann Geophys. 2012;30(10):1479–93.
- View Article
- Google Scholar
62. Misra A, Jayaraman A, Ganguly D. Validation of Version 5.1 MODIS Aerosol Optical Depth (Deep Blue Algorithm and Dark Target Approach) over a Semi-Arid Location in Western India. Aerosol Air Qual Res. 2015;15(1):252–62.
- View Article
- Google Scholar
63. Sayer AM, Hsu NC, Bettenhausen C, Jeong M ‐J. Validation and uncertainty estimates for MODIS Collection 6 “Deep Blue” aerosol data. JGR Atmospheres. 2013;118(14):7864–72.
- View Article
- Google Scholar
64. Tripathi SN, Dey S, Chandel A, Srivastava S, Singh RP, Holben BN. Comparison of MODIS and AERONET derived aerosol optical depth over the Ganga Basin, India. Ann Geophys. 2005;23(4):1093–101.
- View Article
- Google Scholar
65. Breugem AJ, Wesseling JG, Oostindie K, Ritsema CJ. Meteorological aspects of heavy precipitation in relation to floods – An overview. Earth-Science Reviews. 2020;204:103171.
- View Article
- Google Scholar
66. Lyngwa RV, Nayak MA. Atmospheric river linked to extreme rainfall events over Kerala in August 2018. Atmospheric Research. 2021;253:105488.
- View Article
- Google Scholar
67. Xavier A, Kottayil A, Mohanakumar K, Xavier PK. The role of monsoon low‐level jet in modulating heavy rainfall events. Intl Journal of Climatology. 2018;38(S1).
- View Article
- Google Scholar
68. Rajeevan M, Bhate J, Jaswal AK. Analysis of variability and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data. Geophysical Research Letters. 2008;35(18).
- View Article
- Google Scholar
69. Mohandas S, Francis T, Singh V, Jayakumar A, George JP, Sandeep A, et al. NWP perspective of the extreme precipitation and flood event in Kerala (India) during August 2018. Dynamics of Atmospheres and Oceans. 2020;91:101158.
- View Article
- Google Scholar
70. Kaufman YJ, Tanré D, Boucher O. A satellite view of aerosols in the climate system. Nature. 2002;419(6903):215–23. pmid:12226676
- View Article
- PubMed/NCBI
- Google Scholar
71. Ramanathan V, Crutzen PJ, Kiehl JT, Rosenfeld D. Aerosols, climate, and the hydrological cycle. Science. 2001;294(5549):2119–24. pmid:11739947
- View Article
- PubMed/NCBI
- Google Scholar
72. Rosenfeld D. TRMM observed first direct evidence of smoke from forest fires inhibiting rainfall. Geophysical Research Letters. 1999;26(20):3105–8.
- View Article
- Google Scholar
73. Khain A, Rosenfeld D, Pokrovsky A. Aerosol impact on the dynamics and microphysics of deep convective clouds. Quart J Royal Meteoro Soc. 2005;131(611):2639–63.
- View Article
- Google Scholar
74. Koren I, Dagan G, Altaratz O. From aerosol-limited to invigoration of warm convective clouds. Science. 2014;344(6188):1143–6. pmid:24904161
- View Article
- PubMed/NCBI
- Google Scholar
75. Choudhury G, Tyagi B, Vissa NK, Singh J, Sarangi C, Tripathi SN, et al. Aerosol-enhanced high precipitation events near the Himalayan foothills. Atmos Chem Phys. 2020;20(23):15389–99.
- View Article
- Google Scholar
76. Sarangi C, Tripathi SN, Kanawade VP, Koren I, Pai DS. Investigation of the aerosol–cloud–rainfall association over the Indian summer monsoon region. Atmos Chem Phys. 2017;17(8):5185–204.
- View Article
- Google Scholar
77. Koren I, Altaratz O, Remer LA, Feingold G, Martins JV, Heiblum RH. Aerosol-induced intensification of rain from the tropics to the mid-latitudes. Nature Geosci. 2012;5(2):118–22.
- View Article
- Google Scholar
78. Ramanathan V, Chung C, Kim D, Bettge T, Buja L, Kiehl JT, et al. Atmospheric brown clouds: impacts on South Asian climate and hydrological cycle. Proc Natl Acad Sci U S A. 2005;102(15):5326–33. pmid:15749818
- View Article
- PubMed/NCBI
- Google Scholar
79. Schwartz SE. The whitehouse effect—Shortwave radiative forcing of climate by anthropogenic aerosols: an overview. Journal of Aerosol Science. 1996;27(3):359–82.
- View Article
- Google Scholar
80. Steinfeld JI. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. Environment: Science and Policy for Sustainable Development. 1998;40(7):26–26.
- View Article
- Google Scholar
81. Twomey S. The Influence of Pollution on the Shortwave Albedo of Clouds. J Atmos Sci. 1977;34(7):1149–52.
- View Article
- Google Scholar
82. IPCC. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK: Intergovernmental Panel on Climate Change, 2007.
83. Trenberth KE, Fasullo J, Smith L. Trends and variability in column-integrated atmospheric water vapor. Climate Dynamics. 2005;24(7–8):741–58.
- View Article
- Google Scholar

[ref1] 1. Kumar S, Silva Y, Moya-Álvarez AS, Martínez-Castro D. Seasonal and Regional Differences in Extreme Rainfall Events and Their Contribution to the World’s Precipitation: GPM Observations. Advances in Meteorology. 2019;2019:1–15.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. IPCC. Weather and Climate Extreme Events in a Changing Climate. In: Intergovernmental Panel on Climate C, editor. Climate Change 2021 – The Physical Science Basis: Working Group I Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press; 2023. p. 1513–766.

[ref3] 3. WMO. Economic costs of weather-related disasters soars but early warnings save lives. 2023 [13 March 2025. ]. Available from: https://wmo.int/news/media-centre/economic-costs-of-weather-related-disasters-soars-early-warnings-save-lives
View Article
Google Scholar

[6] View Article

[7] Google Scholar

[ref4] 4. KSCSTE. Report of Committee Examine the Causes of Repeated Extreme Heavy rainfall events, Subsequent floods and Landslides and to Recommend Appropriate Policy Responses. Thiruvananthapuram: Kerala State Council for Science, Technology and Environment; 2019.

[ref5] 5. Mao J, Ping F, Yin L, Qiu X. A Study of Cloud Microphysical Processes Associated With Torrential Rainfall Event Over Beijing. JGR Atmospheres. 2018;123(16):8768–91.
View Article
Google Scholar

[10] View Article

[11] Google Scholar

[ref6] 6. Srinivas CV, Yesubabu V, Hari Prasad D, Hari Prasad KBRR, Greeshma MM, Baskaran R, et al. Simulation of an extreme heavy rainfall event over Chennai, India using WRF: Sensitivity to grid resolution and boundary layer physics. Atmospheric Research. 2018;210:66–82.
View Article
Google Scholar

[13] View Article

[14] Google Scholar

[ref7] 7. Wang R, Zhu Y, Qiao F, Liang X-Z, Zhang H, Ding Y. High-resolution Simulation of an Extreme Heavy Rainfall Event in Shanghai Using the Weather Research and Forecasting Model: Sensitivity to Planetary Boundary Layer Parameterization. Adv Atmos Sci. 2021;38(1):98–115.
View Article
Google Scholar

[16] View Article

[17] Google Scholar

[ref8] 8. Ray P, Zhang C, Moncrieff M, Dudhia J, Li T. Tropical channel model. In: Druyan L, editor. Climate models. Rijeka: IntechOpen; 2012.

[ref9] 9. Radhakrishna B, Zawadzki I, Fabry F. Predictability of Precipitation from Continental Radar Images. Part V: Growth and Decay. Journal of the Atmospheric Sciences. 2012;69(11):3336–49.
View Article
Google Scholar

[20] View Article

[21] Google Scholar

[ref10] 10. Sun J, Xue M, Wilson JW, Zawadzki I, Ballard SP, Onvlee-Hooimeyer J, et al. Use of NWP for Nowcasting Convective Precipitation: Recent Progress and Challenges. Bulletin of the American Meteorological Society. 2014;95(3):409–26.
View Article
Google Scholar

[23] View Article

[24] Google Scholar

[ref11] 11. Wilson JW, Crook NA, Mueller CK, Sun J, Dixon M. Nowcasting Thunderstorms: A Status Report. Bull Amer Meteor Soc. 1998;79(10):2079–99.
View Article
Google Scholar

[26] View Article

[27] Google Scholar

[ref12] 12. Fox NI, Wikle CK. A Bayesian Quantitative Precipitation Nowcast Scheme. Weather and Forecasting. 2005;20(3):264–75.
View Article
Google Scholar

[29] View Article

[30] Google Scholar

[ref13] 13. He S, Raghavan SV, Nguyen NS, Liong S. Ensemble rainfall forecasting with numerical weather prediction and radar‐based nowcasting models. Hydrological Processes. 2013;27(11):1560–71.
View Article
Google Scholar

[32] View Article

[33] Google Scholar

[ref14] 14. Li PW, Lai EST. Applications of radar‐based nowcasting techniques for mesoscale weather forecasting in Hong Kong. Meteorological Applications. 2004;11(3):253–64.
View Article
Google Scholar

[35] View Article

[36] Google Scholar

[ref15] 15. Ravuri S, Lenc K, Willson M, Kangin D, Lam R, Mirowski P, et al. Skilful precipitation nowcasting using deep generative models of radar. Nature. 2021;597(7878):672–7. pmid:34588668
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref16] 16. Sivasubramaniam K, Sharma A, Alfredsen K. Merging radar and gauge information within a dynamical model combination framework for precipitation estimation in cold climates. Environmental Modelling & Software. 2019;119:99–110.
View Article
Google Scholar

[42] View Article

[43] Google Scholar

[ref17] 17. Xiang Y, Ma J, Wu X. A Precipitation Nowcasting Mechanism for Real-World Data Based on Machine Learning. Mathematical Problems in Engineering. 2020;2020:1–11.
View Article
Google Scholar

[45] View Article

[46] Google Scholar

[ref18] 18. Metta S, von Hardenberg J, Ferraris L, Rebora N, Provenzale A. Precipitation Nowcasting by a Spectral-Based Nonlinear Stochastic Model. Journal of Hydrometeorology. 2009;10(5):1285–97.
View Article
Google Scholar

[48] View Article

[49] Google Scholar

[ref19] 19. Pulkkinen S, Chandrasekar V, Harri A-M. Stochastic Spectral Method for Radar-Based Probabilistic Precipitation Nowcasting. Journal of Atmospheric and Oceanic Technology. 2019;36(6):971–85.
View Article
Google Scholar

[51] View Article

[52] Google Scholar

[ref20] 20. Sirangelo B, Versace P, De Luca DL. Rainfall nowcasting by at site stochastic model P.R.A.I.S.E. Hydrol Earth Syst Sci. 2007;11(4):1341–51.
View Article
Google Scholar

[54] View Article

[55] Google Scholar

[ref21] 21. Fortelli A, Scafetta N, Mazzarella A. Nowcasting and real-time monitoring of heavy rainfall events inducing flash-floods: an application to Phlegraean area (Central-Southern Italy). Nat Hazards. 2019;97(2):861–89.
View Article
Google Scholar

[57] View Article

[58] Google Scholar

[ref22] 22. Kamarianakis Y, Feidas H, Kokolatos G, Chrysoulakis N, Karatzias V. Evaluating remotely sensed rainfall estimates using nonlinear mixed models and geographically weighted regression. Environmental Modelling & Software. 2008;23(12):1438–47.
View Article
Google Scholar

[60] View Article

[61] Google Scholar

[ref23] 23. Shukla BP, Kishtawal CM, Pal PK. Satellite-Based Nowcasting of Extreme Rainfall Events Over Western Himalayan Region. IEEE J Sel Top Appl Earth Observations Remote Sensing. 2017;10(5):1681–6.
View Article
Google Scholar

[63] View Article

[64] Google Scholar

[ref24] 24. Anderson SR, Cole SJ, Klein C, Taylor CM, Diop CA, Kamara M. Nowcasting convective activity for the Sahel: A simple probabilistic approach using real‐time and historical satellite data on cloud‐top temperature. Quart J Royal Meteoro Soc. 2024;150(759):597–617.
View Article
Google Scholar

[66] View Article

[67] Google Scholar

[ref25] 25. Jumianti N, Marzuki M, Yusnaini H, Ramadhan R, Harjupa W, Saufina E, et al. Prediction of extreme rain in Kototabang using Himawari-8 satellite based on differences in cloud brightness temperature. Remote Sensing Applications: Society and Environment. 2024;33:101102.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref26] 26. Ahmed F, Schumacher C. Convective and stratiform components of the precipitation‐moisture relationship. Geophysical Research Letters. 2015;42(23).
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref27] 27. Roca R, Fiolleau T. Extreme precipitation in the tropics is closely associated with long-lived convective systems. Commun Earth Environ. 2020;1(1).
View Article
Google Scholar

[75] View Article

[76] Google Scholar

[ref28] 28. Andreae MO, Rosenfeld D, Artaxo P, Costa AA, Frank GP, Longo KM, et al. Smoking rain clouds over the Amazon. Science. 2004;303(5662):1337–42. pmid:14988556
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref29] 29. Freud E, Rosenfeld D, Kulkarni JR. Resolving both entrainment-mixing and number of activated CCN in deep convective clouds. Atmos Chem Phys. 2011;11(24):12887–900.
View Article
Google Scholar

[82] View Article

[83] Google Scholar

[ref30] 30. Guo L, Turner AG, Highwood EJ. Impacts of 20th century aerosol emissions on the South Asian monsoon in the CMIP5 models. Atmos Chem Phys. 2015;15(11):6367–78.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref31] 31. Zhao C, Wang Y, Yang Q, Fu R, Cunnold D, Choi Y. Impact of East Asian summer monsoon on the air quality over China: View from space. J Geophys Res. 2010;115(D9).
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref32] 32. Rosenfeld D, Lohmann U, Raga GB, O’Dowd CD, Kulmala M, Fuzzi S, et al. Flood or drought: how do aerosols affect precipitation?. Science. 2008;321(5894):1309–13. pmid:18772428
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref33] 33. Rosenfeld D, Lensky IM. Satellite–Based Insights into Precipitation Formation Processes in Continental and Maritime Convective Clouds. Bull Amer Meteor Soc. 1998;79(11):2457–76.
View Article
Google Scholar

[95] View Article

[96] Google Scholar

[ref34] 34. Rosenfeld D. Chapter 6 - Cloud-Aerosol-Precipitation Interactions Based of Satellite Retrieved Vertical Profiles of Cloud Microstructure. In: Islam T, Hu Y, Kokhanovsky A, Wang J, editors. Remote sensing of aerosols, clouds, and precipitation. Amsterdam: Elsevier. 2018. p. 129–52.

[ref35] 35. Zhu Y, Rosenfeld D, Yu X, Li Z. Separating aerosol microphysical effects and satellite measurement artifacts of the relationships between warm rain onset height and aerosol optical depth. JGR Atmospheres. 2015;120(15):7726–36.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref36] 36. Pai DS, Rajeevan M, Sreejith OP, Mukhopadhyay B, Satbha NS. Development of a new high spatial resolution (0.25° × 0.25°) long period (1901-2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. MAUSAM. 2014;65(1):1–18.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref37] 37. Minnis P, Nguyen L, Palikonda R, Heck P, Spangenberg D, Doelling D, et al. Near-real time cloud retrievals from operational and research meteorological satellites. SPIE; 2008.

[ref38] 38. Nakajima T, King MD. Determination of the Optical Thickness and Effective Particle Radius of Clouds from Reflected Solar Radiation Measurements. Part I: Theory. J Atmos Sci. 1990;47(15):1878–93.
View Article
Google Scholar

[106] View Article

[107] Google Scholar

[ref39] 39. Rosenfeld D, Liu G, Yu X, Zhu Y, Dai J, Xu X, et al. High-resolution (375 m) cloud microstructure as seen from the NPP/VIIRS satellite imager. Atmos Chem Phys. 2014;14(5):2479–96.
View Article
Google Scholar

[109] View Article

[110] Google Scholar

[ref40] 40. Thomas A, Kanawade VP, Chakravarty K, Srivastava AK. Characterization of raindrop size distributions and its response to cloud microphysical properties. Atmospheric Research. 2021;249:105292.
View Article
Google Scholar

[112] View Article

[113] Google Scholar

[ref41] 41. Sarker IH. Machine Learning: Algorithms, Real-World Applications and Research Directions. SN Comput Sci. 2021;2(3):160. pmid:33778771
View Article
PubMed/NCBI
Google Scholar

[115] View Article

[116] PubMed/NCBI

[117] Google Scholar

[ref42] 42. WMO. Guidelines for Nowcasting Techniques. Geneva: World Meteorological Organization; 2017.

[ref43] 43. Hair JF Jr., Black WC, Babin BJ, Anderson RE. Multivariate Data Analysis. 7 ed. New York: Pearson Prentice Hall; 2010.

[ref44] 44. Huffman G, Stocker E, Bolvin D, Nelkin E, Tan J. Gpm imerg final precipitation L3 half hourly 0.1 degree x 0.1 degree v06. 2019.

[ref45] 45. Mukhopadhyay P, Bechtold P, Zhu Y, Murali Krishna RP, Kumar S, Ganai M, et al. Unraveling the Mechanism of Extreme (More than 30 Sigma) Precipitation during August 2018 and 2019 over Kerala, India. Weather and Forecasting. 2021;36(4):1253–73.
View Article
Google Scholar

[122] View Article

[123] Google Scholar

[ref46] 46. Freud E, Rosenfeld D. Linear relation between convective cloud drop number concentration and depth for rain initiation. J Geophys Res. 2012;117(D2).
View Article
Google Scholar

[125] View Article

[126] Google Scholar

[ref47] 47. Suzuki K, Nakajima TY, Stephens GL. Particle Growth and Drop Collection Efficiency of Warm Clouds as Inferred from Joint CloudSat and MODIS Observations. Journal of the Atmospheric Sciences. 2010;67(9):3019–32.
View Article
Google Scholar

[128] View Article

[129] Google Scholar

[ref48] 48. Rosenfeld D, Andreae MO, Asmi A, Chin M, de Leeuw G, Donovan DP, et al. Global observations of aerosol-cloud-precipitation-climate interactions. Rev Geophys. 2014;52(4):750–808.
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref49] 49. Konwar M, Maheskumar RS, Kulkarni JR, Freud E, Goswami BN, Rosenfeld D. Aerosol control on depth of warm rain in convective clouds. J Geophys Res. 2012;117(D13).
View Article
Google Scholar

[134] View Article

[135] Google Scholar

[ref50] 50. Rosenfeld D, Lahav R, Khain A, Pinsky M. The role of sea spray in cleansing air pollution over ocean via cloud processes. Science. 2002;297(5587):1667–70. pmid:12183635
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref51] 51. Lamb D. Cloud microphysics. In: Holton JR, editor. Encyclopedia of Atmospheric Sciences. Oxford: Academic Press; 2003. p. 459–67.

[ref52] 52. Konwar M, Das SK, Deshpande SM, Chakravarty K, Goswami BN. Microphysics of clouds and rain over the Western Ghat. JGR Atmospheres. 2014;119(10):6140–59.
View Article
Google Scholar

[142] View Article

[143] Google Scholar

[ref53] 53. Ansmann A, Tesche M, Althausen D, Müller D, Seifert P, Freudenthaler V, et al. Influence of Saharan dust on cloud glaciation in southern Morocco during the Saharan Mineral Dust Experiment. J Geophys Res. 2008;113(D4).
View Article
Google Scholar

[145] View Article

[146] Google Scholar

[ref54] 54. Rosenfeld D, Woodley W. Deep convective clouds with sustained supercooled liquid water down to −37.5 degrees C. Nature. 2000;405(6785):440–2. pmid:10839535
View Article
PubMed/NCBI
Google Scholar

[148] View Article

[149] PubMed/NCBI

[150] Google Scholar

[ref55] 55. Hallett J, Mossop SC. Production of secondary ice particles during the riming process. Nature. 1974;249(5452):26–8.
View Article
Google Scholar

[152] View Article

[153] Google Scholar

[ref56] 56. Rosenfeld D, Yu X, Liu G, Xu X, Zhu Y, Yue Z, et al. Glaciation temperatures of convective clouds ingesting desert dust, air pollution and smoke from forest fires. Geophys Res Lett. 2011;38(21):n/a-n/a.
View Article
Google Scholar

[155] View Article

[156] Google Scholar

[ref57] 57. Rosenfeld D, Fromm M, Trentmann J, Luderer G, Andreae MO, Servranckx R. The Chisholm firestorm: observed microstructure, precipitation and lightning activity of a pyro-cumulonimbus. Atmos Chem Phys. 2007;7(3):645–59.
View Article
Google Scholar

[158] View Article

[159] Google Scholar

[ref58] 58. Meenu S, Gayatri K, Malap N, Murugavel P, Samanta S, Prabha TV. The physics of extreme rainfall event: An investigation with multisatellite observations and numerical simulations. Journal of Atmospheric and Solar-Terrestrial Physics. 2020;204:105275.
View Article
Google Scholar

[161] View Article

[162] Google Scholar

[ref59] 59. Li Z, Niu F, Fan J, Liu Y, Rosenfeld D, Ding Y. Long-term impacts of aerosols on the vertical development of clouds and precipitation. Nature Geosci. 2011;4(12):888–94.
View Article
Google Scholar

[164] View Article

[165] Google Scholar

[ref60] 60. (GMAO) GMaAO. MERRA-2 inst1_2d_asm_Nx: 2d,1-Hourly,Instantaneous,Single-Level,Assimilation,Single-Level Diagnostics V5.12.4. In: DISC) GESDaISCG, editor. 5.12.4 ed. Greenbelt, MD, USA; 2015.

[ref61] 61. Choudhry P, Misra A, Tripathi SN. Study of MODIS derived AOD at three different locations in the Indo Gangetic Plain: Kanpur, Gandhi College and Nainital. Ann Geophys. 2012;30(10):1479–93.
View Article
Google Scholar

[168] View Article

[169] Google Scholar

[ref62] 62. Misra A, Jayaraman A, Ganguly D. Validation of Version 5.1 MODIS Aerosol Optical Depth (Deep Blue Algorithm and Dark Target Approach) over a Semi-Arid Location in Western India. Aerosol Air Qual Res. 2015;15(1):252–62.
View Article
Google Scholar

[171] View Article

[172] Google Scholar

[ref63] 63. Sayer AM, Hsu NC, Bettenhausen C, Jeong M ‐J. Validation and uncertainty estimates for MODIS Collection 6 “Deep Blue” aerosol data. JGR Atmospheres. 2013;118(14):7864–72.
View Article
Google Scholar

[174] View Article

[175] Google Scholar

[ref64] 64. Tripathi SN, Dey S, Chandel A, Srivastava S, Singh RP, Holben BN. Comparison of MODIS and AERONET derived aerosol optical depth over the Ganga Basin, India. Ann Geophys. 2005;23(4):1093–101.
View Article
Google Scholar

[177] View Article

[178] Google Scholar

[ref65] 65. Breugem AJ, Wesseling JG, Oostindie K, Ritsema CJ. Meteorological aspects of heavy precipitation in relation to floods – An overview. Earth-Science Reviews. 2020;204:103171.
View Article
Google Scholar

[180] View Article

[181] Google Scholar

[ref66] 66. Lyngwa RV, Nayak MA. Atmospheric river linked to extreme rainfall events over Kerala in August 2018. Atmospheric Research. 2021;253:105488.
View Article
Google Scholar

[183] View Article

[184] Google Scholar

[ref67] 67. Xavier A, Kottayil A, Mohanakumar K, Xavier PK. The role of monsoon low‐level jet in modulating heavy rainfall events. Intl Journal of Climatology. 2018;38(S1).
View Article
Google Scholar

[186] View Article

[187] Google Scholar

[ref68] 68. Rajeevan M, Bhate J, Jaswal AK. Analysis of variability and trends of extreme rainfall events over India using 104 years of gridded daily rainfall data. Geophysical Research Letters. 2008;35(18).
View Article
Google Scholar

[189] View Article

[190] Google Scholar

[ref69] 69. Mohandas S, Francis T, Singh V, Jayakumar A, George JP, Sandeep A, et al. NWP perspective of the extreme precipitation and flood event in Kerala (India) during August 2018. Dynamics of Atmospheres and Oceans. 2020;91:101158.
View Article
Google Scholar

[192] View Article

[193] Google Scholar

[ref70] 70. Kaufman YJ, Tanré D, Boucher O. A satellite view of aerosols in the climate system. Nature. 2002;419(6903):215–23. pmid:12226676
View Article
PubMed/NCBI
Google Scholar

[195] View Article

[196] PubMed/NCBI

[197] Google Scholar

[ref71] 71. Ramanathan V, Crutzen PJ, Kiehl JT, Rosenfeld D. Aerosols, climate, and the hydrological cycle. Science. 2001;294(5549):2119–24. pmid:11739947
View Article
PubMed/NCBI
Google Scholar

[199] View Article

[200] PubMed/NCBI

[201] Google Scholar

[ref72] 72. Rosenfeld D. TRMM observed first direct evidence of smoke from forest fires inhibiting rainfall. Geophysical Research Letters. 1999;26(20):3105–8.
View Article
Google Scholar

[203] View Article

[204] Google Scholar

[ref73] 73. Khain A, Rosenfeld D, Pokrovsky A. Aerosol impact on the dynamics and microphysics of deep convective clouds. Quart J Royal Meteoro Soc. 2005;131(611):2639–63.
View Article
Google Scholar

[206] View Article

[207] Google Scholar

[ref74] 74. Koren I, Dagan G, Altaratz O. From aerosol-limited to invigoration of warm convective clouds. Science. 2014;344(6188):1143–6. pmid:24904161
View Article
PubMed/NCBI
Google Scholar

[209] View Article

[210] PubMed/NCBI

[211] Google Scholar

[ref75] 75. Choudhury G, Tyagi B, Vissa NK, Singh J, Sarangi C, Tripathi SN, et al. Aerosol-enhanced high precipitation events near the Himalayan foothills. Atmos Chem Phys. 2020;20(23):15389–99.
View Article
Google Scholar

[213] View Article

[214] Google Scholar

[ref76] 76. Sarangi C, Tripathi SN, Kanawade VP, Koren I, Pai DS. Investigation of the aerosol–cloud–rainfall association over the Indian summer monsoon region. Atmos Chem Phys. 2017;17(8):5185–204.
View Article
Google Scholar

[216] View Article

[217] Google Scholar

[ref77] 77. Koren I, Altaratz O, Remer LA, Feingold G, Martins JV, Heiblum RH. Aerosol-induced intensification of rain from the tropics to the mid-latitudes. Nature Geosci. 2012;5(2):118–22.
View Article
Google Scholar

[219] View Article

[220] Google Scholar

[ref78] 78. Ramanathan V, Chung C, Kim D, Bettge T, Buja L, Kiehl JT, et al. Atmospheric brown clouds: impacts on South Asian climate and hydrological cycle. Proc Natl Acad Sci U S A. 2005;102(15):5326–33. pmid:15749818
View Article
PubMed/NCBI
Google Scholar

[222] View Article

[223] PubMed/NCBI

[224] Google Scholar

[ref79] 79. Schwartz SE. The whitehouse effect—Shortwave radiative forcing of climate by anthropogenic aerosols: an overview. Journal of Aerosol Science. 1996;27(3):359–82.
View Article
Google Scholar

[226] View Article

[227] Google Scholar

[ref80] 80. Steinfeld JI. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. Environment: Science and Policy for Sustainable Development. 1998;40(7):26–26.
View Article
Google Scholar

[229] View Article

[230] Google Scholar

[ref81] 81. Twomey S. The Influence of Pollution on the Shortwave Albedo of Clouds. J Atmos Sci. 1977;34(7):1149–52.
View Article
Google Scholar

[232] View Article

[233] Google Scholar

[ref82] 82. IPCC. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK: Intergovernmental Panel on Climate Change, 2007.

[ref83] 83. Trenberth KE, Fasullo J, Smith L. Trends and variability in column-integrated atmospheric water vapor. Climate Dynamics. 2005;24(7–8):741–58.
View Article
Google Scholar

[236] View Article

[237] Google Scholar

Figures

Abstract

1. Introduction

2. Material and methods

2.1. Characterization of EREs

2.2. Assessment of T-re profile of rainfall events

2.3. Nowcasting of EREs

3. Results and discussion

3.1. Characterization of EREs over Kerala (2001–2018)

3.2. Variation of the T-re profiles between normal, heavy, and extreme rainfall events

3.3. Nowcasting of EREs

3.3.1. Model evaluation.

3.4. Further evidence of the influence of aerosols and moisture availability on EREs

4. Summary and conclusions

Supporting information

S1 Fig. Image of the flood-affected areas over Kerala derived from the Sentinel-1 SAR data during August 2018.

S2 Fig. Daily rainfall over a grid centered at 9.5°N and 76.75°E during August 2018.

S3 Fig. Wind trajectory clusters from a central location (10.5°N, 76.3°E) over Kerala during 2001.

S4 Fig. Nowcasting the heavy/extreme rainfall events of 14 October 2023 in Thiruvananthapuram, Kerala.

S5 Fig. Temporal variation of AOD over Kerala from 2001 to 2018.

References

2.2. Assessment of T-r_e profile of rainfall events

3.2. Variation of the T-r_e profiles between normal, heavy, and extreme rainfall events