ABSTRACT
Due to climate change, rising temperatures cause the atmosphere to hold more moisture, likely increasing the frequency and intensity of extreme precipitation events. These events, lasting several days, are the leading causes of floods. Recently, floods from multi-day precipitation extremes (MPEs) have risen significantly. The Brahmaputra River basin (BRB) in India is particularly vulnerable to flooding during the Indian summer monsoon due to its transboundary nature. It is crucial to characterize and rank MPEs to understand their risk, impact, and underlying drivers. This study ranked MPEs of different durations based on intensity and spatial extent during the Indian summer monsoon over the BRB using a high-resolution daily precipitation dataset from 1951 to 2022. In addition, it evaluates the association between atmospheric moisture transport and MPEs by quantifying integrated water vapor transport (IVT) during top-ranked MPEs. The analysis indicates intense IVT in the mid-tropospheric layer (850–500 hPa) over the MPE regions during top-ranked events and identifies significant low-pressure anomalies and shifts from ridge to trough patterns. Quantifying the connection between MPEs and IVT can aid in early prediction and risk reduction.
HIGHLIGHTS
Ranking multi-day precipitation extremes (MPEs) are crucial for capturing diverse spatiotemporal events.
Ranking MPEs are useful in identifying intense precipitation-driven floods in India.
Atmospheric moisture transport (IVT) in the mid-tropospheric layer (850–500 hPa) can be used as a precursor to MPEs.
Regional-scale atmospheric patterns identified can aid in early warning of MPEs.
1. INTRODUCTION
Extreme precipitation refers to weather events that produce a significant amount of precipitation in a relatively short period and exceed the normal range of precipitation for a given area. As the temperature rises due to climate change, the moisture-holding capacity of the atmosphere increases (Trenberth et al. 2003), which contributes to more frequent and intense extreme precipitation events. Over India, in the past 50 years, precipitation extremes have varied in heterogeneous ways (Joshi & Rajeevan 2006). According to Mukherjee et al. (2018), there has been a significant rise in the annual maximum precipitation during the past three decades in India. In addition, extreme precipitation frequency is projected to rise in India under a warming climate (Mukherjee et al. 2018). Extreme precipitation events persisting for more than 2 days are usually referred to as multi-day precipitation extremes (MPEs). According to Nanditha & Mishra (2022), MPEs are one of the leading causes of floods in India. In recent years, India has witnessed a significant rise in flooding caused by MPEs, and this trend is projected to continue under the influence of anthropogenic warming (Ali et al. 2019). The occurrence of floods in India is mostly confined to the Indian summer monsoon (ISM) season (Nanditha & Mishra 2021). ISM (June–September) accounts for the majority of rainfall (Turner & Annamalai 2012) over India (∼80% of the annual rainfall) and contributes as the critical source of water for agriculture in the region (Hrudya et al. 2021). However, despite significant year-to-year variability, the intensity and frequency of extreme rainfall events during ISM have increased significantly over the past 50 years (Goswami 2006) which is causing significant damage to infrastructure and loss of life.
A multitude of studies have been adopted to characterize and rank precipitation extremes in various parts of the world. Beguería et al. (2009) adopted the extreme value theory to evaluate the characteristics of extreme precipitation occurrences over the north-eastern Iberian Peninsula. An equivalent threshold-based objective method was adopted for identifying regional extreme events (REEs) while considering their impact area and duration (Ren et al. 2012). Ramos et al. (2014) proposed a novel approach for ranking daily extreme precipitation based on intensity and areal extent. This approach was refined further for MPEs by Ramos et al. (2017) using accumulated normalized departure from climatology at shorter time scales (e.g., 3 days) and longer time scales (e.g., 5, 7, and 10 days). Raj et al. (2021) also adopted a similar approach to rank and characterize MPEs in the Indian western Himalayas. However, it is worth mentioning that normalized anomalies of precipitation cannot ensure typical Gaussian distribution. Thus, by accounting for the 95th percentile criterion for computing extreme precipitation days. Ramos et al. (2018) improvised the proposed ranking approach for highly skewed precipitation distribution. Despite the advancements, a ranking procedure which accounts for areal extent, intensity, duration, and highly skewed distribution of precipitation has yet to be implemented in Indian river basins to assess its robustness.
Although extreme precipitation is the main cause of floods, the mechanisms underlying their intensification in response to a warmer climate are of great theoretical interest (Gimeno-Sotelo & Gimeno 2023). The acceleration of the hydrologic cycle brought on by climate change, which is mostly attributable to atmospheric moisture transport, is largely responsible for this intensification (Mukherjee & Mishra 2021; Zhao et al. 2022). The fundamental component of the atmospheric branch of the water cycle is atmospheric moisture transport, and variations in this mechanism have a considerable impact on precipitation extremes. During ISM, the majority of atmospheric moisture is transported through the surface to mid-tropospheric levels from the Arabian Sea (AS) and the Southern Indian Ocean across the equator (Patil et al. 2019). According to Cadet & Greco (1987), the majority ( ∼ 70%) of moisture transport reaches to the west coast of India from the Southern Indian Ocean through the Somali low-level jet (SLLJ), while the remaining moisture (∼30%) arises from the AS through evaporation. Numerous indices have been developed for quantifying atmospheric moisture transport. Due to the accessibility of water vapor images from satellites, many studies have previously employed column-integrated water vapor (IWV) data from microwave sensors (Ralph et al. 2004; Neiman et al. 2008). The column-integrated water vapor transport (IVT), which incorporates horizontal winds in its computation, is now the standard metric to quantify moisture transport because it was later demonstrated that IWV does not account for the flux component, i.e., wind. The ability to more precisely detect better associations with precipitation and are more adeptly anticipated by numerical weather prediction models are two further benefits of IVT over IWV (Waliser & Guan 2017; Dettinger et al. 2018). Several studies in past (Dhana Lakshmi & Satyanarayana 2019; Dhana Laskhmi & Satyanarayana 2020; Lyngwa & Nayak 2021; Gupta et al. 2023; Mahto et al. 2023; Singh Raghuvanshi & Agarwal 2023) have documented atmospheric moisture transport linkages to extreme precipitation and floods in India. However, the majority of these studies focused on understanding the atmospheric moisture transport associated with extreme precipitation events that lasted for shorter time periods (e.g., up to 3 days).
Understanding and forecasting weather in meteorology is strongly linked to regional atmospheric patterns, such as large-scale weather systems spanning roughly 1,000 km and lasting several days to a week, which are closely related to IVT (Singh Raghuvanshi & Agarwal 2023). Key meteorological variables are studied at various tropospheric levels using isopleths or anomaly maps, which aid in identifying critical patterns for surface cyclone and anticyclone development. Various tropospheric charts, such as surface charts (0 m) and upper-tropospheric charts (500 hPa), provide information on mean sea level pressure (MSLP), geopotential height (Z), and other factors. For instance, surface charts (at 0 m) use MSLP to identify surface high- and low-pressure systems, which have a significant impact on the weather (Milrad 2018a). Upper-tropospheric charts (at 500 hPa) often show geopotential height, temperature, and wind speed. The 500 hPa isobaric charts are used to locate upper-tropospheric troughs and ridges. At 500 hPa, troughs represent upper-tropospheric cold air, whereas ridges represent upper-tropospheric warm air (Milrad 2018b). Recent research emphasizes the regional impact of atmospheric anomalies caused by climatic patterns (Raghuvanshi & Agarwal 2024a), which influence moisture transport and result in extreme precipitation occurrences (Mukherjee & Mishra 2021; Singh Raghuvanshi & Agarwal 2023; Raghuvanshi & Agarwal 2024b). While previous research has evaluated regional atmospheric patterns during specific events (Houze et al. 2017) or focused on specific tropospheric levels (Mahto et al. 2023) across the Indian subcontinent, there is an urgent need for systematic quantification of the spatiotemporal evolution of regional climate patterns associated with IVT-linked MPEs in India. This is particularly significant given the increasing frequency of extreme precipitation events in the region (Roxy et al. 2017), and analyzing the progression of regional-scale atmospheric patterns and moisture transport prior to IVT-linked MPEs can help us better understand the processes that contribute to extreme weather over the Indian subcontinent.
According to flood data, the Brahmaputra River basin (BRB) typically experiences the worst floods of significant magnitude during the late ISM season of August and September (Vegad et al. 2023). As a result, the BRB is more vulnerable to floods than other regions in the country, which can be linked to their greater vulnerability, likelihood of hazard, and exposure as transboundary river basins (Vegad et al. 2023). Hence, ranking and characterizing MPEs in BRB will help us in identifying the impact causing extreme events and investigating underlying drivers that contribute to their occurrence and intensification. With this aim, the present study focuses on (1) evaluating the rank of MPEs (1, 3, 5, 7, and 10 days) over BRB during ISM by employing a suitable ranking procedure, (2) investigating the association between IVT and MPEs by assessing the spatiotemporal variability of IVT characteristics during identified top-ranked MPEs, and (3) unraveling the regional atmospheric patterns that serve as precursors for IVT-linked MPEs to understand the underlying dynamics of IVT-linked MPEs
This paper is structured as follows: Section 2 covers the study area and data used, Section 3 details the methodology employed, Section 4 displays the results and provides a brief discussion, and Section 5 provides a summary of the key findings and conclusions.
2. STUDY AREA AND DATA
Elevation map of the BRB in India. Color bar denotes elevation in meters.
For ranking MPEs, daily gridded rainfall data developed by the India Meteorological Department (Pai et al. 2014) at a spatial resolution of 0.25° × 0.25° is used for ISM months, i.e., June–September (JJAS) for the 1951–2022 period. These datasets were prepared with the use of daily rainfall records obtained from 6,955 rain gauge stations over India (Pai et al. 2014). Pai et al. (2014) provided the specifics on how the precipitation data was prepared. Further, we used recently developed European Centre for Medium-Range Weather Forecasts (ECMWF) version 5 (ERA5) Reanalysis data (Hersbach et al. 2020) due to its high spatiotemporal resolution on a global scale. Compared with other reanalysis products, recent studies (Mahto & Mishra 2019; Dullaart et al. 2020; Mahto et al. 2023) have determined ERA-5 to be suitable for hydrological and meteorological assessments in the Indian subcontinent. The latitude–longitude grid of the ERA5 has a spatial resolution of 0.25° × 0.25°. To measure the amount of moisture transported, specific humidity and horizontal wind fields (zonal and meridional wind speed) at various vertical tropospheric levels (1,000–300 hPa; 20 total pressure levels) are retrieved at 6-hourly temporal resolution. In addition, the mean sea pressure level (MSLP) and geopotential height fields at 500 hPa (Z500) are obtained at 6-hourly temporal resolution to examine regional-scale atmospheric patterns.
3. METHOD AND METHODOLOGY
3.1. Ranking MPE events
Spatial plots depicting skewness of daily precipitation distribution in the BRB during the Indian summer monsoon for the period 1951–2022.
Spatial plots depicting skewness of daily precipitation distribution in the BRB during the Indian summer monsoon for the period 1951–2022.







We computed the accumulated extreme precipitation anomalies for 1-, 3-, 5-, 7-, and 10-day periods. For example, the accumulated precipitation anomalies for 5 days on 13 August 2006 corresponds to the sum of the extreme precipitation anomaly relative to 9–13 August 2006.



3.2. Quantifying atmospheric moisture transport
A schematic illustration depicting the calculation of IVT in three distinct layers of the troposphere, represented by different shades of blue: the lower (IVT_L; 1,000–850 hPa), middle (IVT_M; 800–500 hPa), and upper (IVT_U; 500–300 hPa) troposphere. ‘Z’ represents the pressure levels at which meteorological variables (q, u, v) are available to compute IVT across these different tropospheric layers.
A schematic illustration depicting the calculation of IVT in three distinct layers of the troposphere, represented by different shades of blue: the lower (IVT_L; 1,000–850 hPa), middle (IVT_M; 800–500 hPa), and upper (IVT_U; 500–300 hPa) troposphere. ‘Z’ represents the pressure levels at which meteorological variables (q, u, v) are available to compute IVT across these different tropospheric layers.
3.3. Methodology
Initially, the MPEs for varying durations (1, 3, 5, 7, and 10 days) were ranked utilizing the approach discussed in Section 3.1. Subsequently, IVT was calculated using ERA5 datasets at a resolution of 0.25° × 0.25°. The 6-hourly IVT values were averaged to a daily temporal scale for effective comparison with daily precipitation data. Unique top-ranked MPEs for various time periods (1, 3, 5, 7, and 10 days) were selected to investigate their relationship with IVT, IVT_L, IVT_M, and IVT_U. In addition, the underlying regional-scale dynamics, which act as precursors for IVT-associated MPEs, were examined. Within the BRB region, daily anomalies in atmospheric variables were computed at 0, 2, 4, and 6-day lags and a 2-day lead for the selected top-ranked MPEs. The day of maximum precipitation within the MPEs was considered as day 0. Following an approach similar to Mukherjee & Mishra (2021) and Ridder et al. (2018), anomalies in daily MSLP and 500 hPa geopotential height (Z500) were computed at each grid location, highlighting lower tropospheric and upper-tropospheric atmospheric dynamics, respectively. This approach facilitates the identification of regional-scale atmospheric patterns, such as the development of surface cyclones–anticyclones and troughs–ridges patterns, crucial for cloud formation, respectively. To unravel information related to atmospheric patterns and dynamics underlying IVT-linked MPEs, the hourly atmospheric data were averaged on a daily temporal scale.
4. RESULTS AND DISCUSSIONS
4.1. Ranking MPE events
Ranking of 1-day MPEs
Date . | % Area (A) . | Mean (M) . | R = A × M . | Rank . |
---|---|---|---|---|
19 June 1974 | 19.72 | 222.47 | 4,386.75 | 1 |
07 July 1967 | 23.59 | 183.47 | 4,328.33 | 2 |
21 August 1963 | 30.99 | 126.90 | 3,932.09 | 3 |
04 August 1974 | 26.41 | 146.13 | 3,859.06 | 4 |
14 September 1960 | 54.23 | 64.86 | 3,516.81 | 5 |
10 August 1958 | 40.85 | 80.95 | 3,306.37 | 6 |
07 July 1959 | 24.65 | 132.95 | 3,276.95 | 7 |
11 August 1987 | 58.10 | 53.00 | 3,079.01 | 8 |
14 July 1961 | 33.45 | 88.18 | 2,949.68 | 9 |
11 August 2017 | 53.17 | 54.02 | 2,872.06 | 10 |
Date . | % Area (A) . | Mean (M) . | R = A × M . | Rank . |
---|---|---|---|---|
19 June 1974 | 19.72 | 222.47 | 4,386.75 | 1 |
07 July 1967 | 23.59 | 183.47 | 4,328.33 | 2 |
21 August 1963 | 30.99 | 126.90 | 3,932.09 | 3 |
04 August 1974 | 26.41 | 146.13 | 3,859.06 | 4 |
14 September 1960 | 54.23 | 64.86 | 3,516.81 | 5 |
10 August 1958 | 40.85 | 80.95 | 3,306.37 | 6 |
07 July 1959 | 24.65 | 132.95 | 3,276.95 | 7 |
11 August 1987 | 58.10 | 53.00 | 3,079.01 | 8 |
14 July 1961 | 33.45 | 88.18 | 2,949.68 | 9 |
11 August 2017 | 53.17 | 54.02 | 2,872.06 | 10 |
Schematic representation of identified top 10 MPE events at different durations (1, 3, 5, 7, and 10 days). The color bar denotes month and year during which the top 10 MPEs occurred.
Schematic representation of identified top 10 MPE events at different durations (1, 3, 5, 7, and 10 days). The color bar denotes month and year during which the top 10 MPEs occurred.
Further, it can be noticed that there are 10 unique events out of the total 50 (Figure 4). In addition, it is clear from Supplementary Tables S1–S4 that at different durations top ranks are dominated by two events that occurred in September 1960 and June 1974, respectively. Since we analyzed successive accumulated precipitation days over relatively long periods of time, it is expected that this dominance by just a few MPEs will occur at longer durations.

Spatiotemporal variation of precipitation during June 1974 event. The color bar denotes the intensity of daily precipitation in mm.
Spatiotemporal variation of precipitation during June 1974 event. The color bar denotes the intensity of daily precipitation in mm.
Spatiotemporal variation of precipitation during September 1960 event. The color bar denotes the intensity of daily precipitation in mm.
Spatiotemporal variation of precipitation during September 1960 event. The color bar denotes the intensity of daily precipitation in mm.
4.2. Moisture transport associated with MPEs
Top-ranked common MPEs across different time periods (1, 3, 5, 7, and 10 days) were selected (September 1960, August 1963, July 1967, June 1974, July 1974, August 1974, and August 2017) to explore the association between IVT and MPEs. In this section, we only focused on investigating the spatiotemporal variability of atmospheric moisture transport, quantified as IVT, and its association with MPEs during June 1974 and September 1960 due to their dominance across different durations based on spatial extent and intensity. According to the international disaster database (EM-DAT), the June 1974 MPE event had substantial socio-economic consequences, flooding Tinsukia, Dhemaji, Nowgong, Majuli, Jorhat, and Dhubri districts in upper Assam and certain parts of Arunachal Pradesh which fall in the upper BRB region.
Spatiotemporal variation of (a) IVT and (b) IVT at different tropospheric layers during June 1974 precipitation event. The color bar denotes the intensity of daily atmospheric moisture flux in kg/m/s. The black color arrow denotes the direction of moisture flux.
Spatiotemporal variation of (a) IVT and (b) IVT at different tropospheric layers during June 1974 precipitation event. The color bar denotes the intensity of daily atmospheric moisture flux in kg/m/s. The black color arrow denotes the direction of moisture flux.
Spatiotemporal variation of (a) IVT and (b) IVT at different tropospheric layers during September 1960 precipitation event. The color bar denotes the intensity of daily atmospheric moisture flux in kg/m/s. The black color arrow denotes the direction of moisture flux.
Spatiotemporal variation of (a) IVT and (b) IVT at different tropospheric layers during September 1960 precipitation event. The color bar denotes the intensity of daily atmospheric moisture flux in kg/m/s. The black color arrow denotes the direction of moisture flux.
4.3. Atmospheric dynamics associated with IVT-linked MPEs
Spatial map showing (a) anomalies in MSLP (shading; units hPa), (b) anomalies in Z500 (contours; units: dam), (c) IVT (shading; units: kg/m/s) and (d) precipitation (shading; units: mm) over the BRB for 6, 4, 2, 0-day lag and 2-day lead during the September 1960 event. The red rectangular box delineates the region identified based on the maximum precipitation occurrence at 0-day over the BRB.
Spatial map showing (a) anomalies in MSLP (shading; units hPa), (b) anomalies in Z500 (contours; units: dam), (c) IVT (shading; units: kg/m/s) and (d) precipitation (shading; units: mm) over the BRB for 6, 4, 2, 0-day lag and 2-day lead during the September 1960 event. The red rectangular box delineates the region identified based on the maximum precipitation occurrence at 0-day over the BRB.
Spatial map showing (a) anomalies in MSLP (shading; units hPa), (b) anomalies in Z500 (contours; units: dam), (c) IVT (shading; units: kg/m/s) and (d) precipitation (shading; units: mm) over the BRB for 6, 4, 2, 0-day lag and 2-day lead during the June 1974 event. The red rectangular box delineates the region identified based on the maximum precipitation occurrence at 0-day over the BRB.
Spatial map showing (a) anomalies in MSLP (shading; units hPa), (b) anomalies in Z500 (contours; units: dam), (c) IVT (shading; units: kg/m/s) and (d) precipitation (shading; units: mm) over the BRB for 6, 4, 2, 0-day lag and 2-day lead during the June 1974 event. The red rectangular box delineates the region identified based on the maximum precipitation occurrence at 0-day over the BRB.
Figures 9(b) and 10(b) and Supplementary Figures S6(b)–S10(b) depict the spatiotemporal progression of anomalies in Z500 (shading) for lag times of 6, 4, 2, and 0 days and a lead time of 2 days (indicated by ‘ − 6 days’, ‘ −4 days’, ‘ − 2 days’, ‘0 days’, and ‘ + 2 days’, respectively). The Z500 anomaly contours corresponding to all MPEs (Figures 9(d) and 10(d) and Supplementary Figures S6(d)–S10(d)) illustrate a trough (negative anomalies)–ridge (positive anomalies) pattern from 6 to 2 days before the extreme event occurrences. Over the BRB region, a consistent trough–ridge pattern emerges, with the trough approaching the area of extreme precipitation (highlighted in the red box in Figures 9(b) and 10(b) and Supplementary Figures S6(b)–S10(b)), while the ridge over the extreme precipitation region gradually shifts away in the days leading up to the event ( − 6 to + 2 days). In addition, the trough–ridge pattern anomalies vary for each MPE. Similar geopotential height patterns at 500 hPa have been observed in the study of 14-day extreme precipitation events across the United States (Jennrich et al. 2020). These trough–ridge patterns, as described by Bluestein (1992), create favorable conditions for precipitation due to positive differential vorticity advection, warm air advection promoting rising motion, and positive moisture advection across the region. Thus, all MPEs exhibit a conducive regional geopotential height pattern preceding IVT-linked extreme precipitation events.
The development of an intense IVT field over the region experiencing extreme precipitation, as indicated by the red box in the BRB (depicted in Figures 9(c) and 10(c) and Supplementary Figures S6(c)–S10(c)), is observed 2 days before the onset of IVT-linked extreme precipitation events (illustrated in Figures 9(d) and 10(d) and Supplementary Figures S6(d)–S10(d)). In addition, there is a substantial increase in the magnitude of atmospheric moisture transport, with IVT intensity exceeding 700 kg/m/s, precisely on the days when these events are witnessed in specific BRB regions. The moisture transport to these areas is predominantly influenced by the movement of moisture-laden monsoon winds originating from the Bay of Bengal, intensifying 2 days prior to the occurrence of extreme events. Furthermore, the anomalies of MSLP exhibited in Figures 9(a) and 10(a) and Supplementary Figures S6(a)–S10(a) align consistently with the temporal evolution of atmospheric moisture transport. Further, the dominance of a trough feature at 500 hPa during IVT-linked extreme precipitation events in the specified region (Figures 9(b) and 10(b) and Supplementary Figures S6(b)–S10(b)) confirms a cascading effect of regional-scale atmospheric patterns on the transportation of moisture by IVTs into the area.
Overall, the evolution of atmospheric moisture transport associated with MPEs indicates a cascading influence of regional-scale atmospheric patterns. Our findings provide fresh insights into the gradual unfolding of atmospheric moisture transport, culminating in the occurrence of devastating MPEs in the BRB region. The MPEs linked to IVT exhibit enhanced dynamical systems, evident through increased IVT intensity, the presence of negative geopotential height anomalies (trough features), and pronounced low MSLP anomalies. This signifies a robust and intensified atmospheric setup contributing to the manifestation of severe MPEs in the BRB region.
5. CONCLUSION
In this study, MPEs were ranked during ISM over the BRB region using the methodology proposed by Ramos et al. (2018). Further, the association between IVT, IVT_L, IVT_M, IVT_U, and MPEs across different durations were investigated by evaluating the characteristics of IVT and precipitation during top-ranked MPEs. The stronger association between intense IVT and MPEs in the BRB region suggests that intense IVT (particularly at mid-tropospheric level) can be utilized as a proxy for predicting MPEs. The spatiotemporal evolution of regional-scale atmospheric patterns, such as geopotential height (Z500) and MSLP, has a significant impact on atmospheric moisture transport during IVT-linked MPEs. The findings of this study will offer sufficient context and insight to aid in forecasting upcoming IVT–MPE events in the Indian subcontinent region. Furthermore, the ranking of MPEs emphasized in this research can be useful in identifying intense precipitation-driven floods in India. Furthermore, the regional-scale atmospheric patterns emphasized in this research can be useful for early warning systems for MPE-driven floods in India. Possible future work would be to investigate the IVT–MPE linkages by accounting for additional aspects such as low-level convergence, or to understand how MPEs interact with IVT under major moisture transport mechanisms (Gimeno et al. 2016; Nischal et al. 2024).
AUTHOR CONTRIBUTION STATEMENT
H.G.: Data curation, Resources, Investigation, Methodology, Software, Writing – original draft, review & editing. A.S.R.: Conceptualization, Data curation, Resources, Investigation, Methodology, Software, Supervision, Writing – original draft, review & editing. A.A.: Conceptualization, Project administration, Funding acquisition, Supervision, Writing – review & editing.
ACKNOWLEDGEMENTS
H.G. acknowledges the financial support from an MTech fellowship by the Ministry of Education, Government of India at IIT Roorkee. A.S.R. acknowledges the Prime Minister Research Fellowship (ID-2803578) provided by the Ministry of Human Resource Development (MHRD), Government of India. A.A. acknowledges the joint funding support from the University Grant Commission (UGC) and DAAD under the Indo-German Partnership in Higher Education (IGP) framework 2020–24 (Co-PREPARE project) at the IIT Roorkee. Data from the India Meteorological Department (IMD) is highly acknowledged. The authors gratefully thank Prof. Ricardo Trigo and Dr Alexandre M. Ramos with whom they had several brainstorming sessions under the Indo-Portugal call (DST/INT/Portugal/P-06/2021(G)) and the Indian Institute of Technology Roorkee for providing facilities to conduct MTech and PhD work.
DATA AVAILABILITY STATEMENT
All data used for this study are freely available. Precipitation data from IMD are available from https://www.imdpune.gov.in/lrfindex.php. ERA5 reanalysis data are obtained from the website: https://cds.climate.copernicus.eu/#!/search?text=ERA5&type=dataset.
CONFLICT OF INTEREST
The authors declare there is no conflict.
REFERENCES
Author notes
Share first authorship equally.