Volume 128, Issue 22 e2023JD038759
Research Article
Open Access

Recent Tangible Natural Variability of Monsoonal Orographic Rainfall in the Eastern Himalayas

Pratik Kad

Corresponding Author

Pratik Kad

Department of Climate System, Pusan National University, Busan, South Korea

Correspondence to:

P. Kad and K.-J. Ha,

[email protected];

[email protected];

[email protected]

Contribution: Conceptualization, ​Investigation, Methodology, Writing - original draft, Writing - review & editing, Data curation, Formal analysis, Software, Visualization

Search for more papers by this author
Kyung-Ja Ha

Corresponding Author

Kyung-Ja Ha

Department of Climate System, Pusan National University, Busan, South Korea

Center for Climate Physics, Institute for Basic Science, Busan, South Korea

BK21 School of Earth and Environmental Systems, Pusan National University, Busan, South Korea

Correspondence to:

P. Kad and K.-J. Ha,

[email protected];

[email protected];

[email protected]

Contribution: Conceptualization, Validation, Funding acquisition, Project administration, Writing - review & editing

Search for more papers by this author
First published: 27 November 2023

This article was corrected on 20 DEC 2023. See the end of the full text for details.


Himalayas hydroclimate is a lifeline for South Asia's most densely populated region. Every year, flooding in the Himalayan rivers is usual during summer monsoon, which impacts millions of inhabitants of the Himalayas and downstream regions. Recent studies demonstrate the role of melting glaciers and snow in the context of global warming, along with monsoonal rain causing recurrent floods. Here, we highlight the natural variability in the eastern Himalayan hydroclimate over the last 43 years (1979–2021). We found extreme monsoonal rainy years with six dry years and seven wet years after removing the climate change signal. Monsoon rainfall is a significant contributor, and melting snow is not a potential contributor to these anomalous extreme years. The variability of Himalayan monsoonal rainfall is strongly regulated by local monsoonal Hadley circulation associated with tropical sea surface temperature. Our findings demonstrate mechanisms associated with Himalayan wet and dry monsoon. Atmospheric dynamics are attributed as the primary modulating factor, influencing local thermodynamics through moist processes. The insights provided in this study underscore the impact of natural variability-driven challenging events that could be predictable. Thus, this mechanism could improve the predictability of the Himalayas floods.

Key Points

  • The study highlights the natural variability in the eastern Himalayan hydroclimate over the past 43 years, emphasizing its significance as a recurring natural hazard that affects the region

  • The research identifies extreme monsoonal rainy years, with monsoon rainfall as a major contributor to river discharge. Notably, the study rules out the role of melting snow in these extreme events

  • This research underscores the dominant influence of atmospheric dynamics as the primary modulating factor in the Eastern Himalayan monsoon

Plain Language Summary

The monsoon rainfall in the Himalayan region is very important for the people living in South Asia. Every year, heavy rainfall during the monsoon season causes floods in the region. We found the strong Himalayan rainfall variability in the eastern side of the Himalayas during the monsoon season and examined the natural rainfall cycle over the past few decades. As a result, we have characterized that there are extreme rainy monsoons and dry monsoons. We also found that the main cause of the extreme flooding is the heavy rain during the monsoon as a part of natural variability rather than the melting snow. Our study supports atmospheric bridge as the primary driver for extreme monsoons over the Eastern Himalayan region. We looked at different factors that influence monsoon rainfall, such as the local landscape, the conditions in the atmosphere, and large-scale features. The eastern Himalayas region includes Bhutan, Nepal, India, and China. Understanding these rainfall patterns can help us predict floods in these countries and better prepare for their impact. Furthermore, as climate change continues to affect the climate system, there is growing concern about how it may influence the natural variability of the Himalayan monsoon.

1 Introduction

The Himalayas are essential to the global water cycle (Immerzeel et al., 2020). It also holds the most dominant biodiversity hotspots in the Hindu Kush Himalayan ranges, including natural heritage like Chitwan National Park, Kaziranga National Park, Khangchendzonga National Park, Manas Wildlife Sanctuary, etc. The Himalayas is one of the youngest mountain range on Earth, which emerged around 50 million years ago, resulting from a continental collision according to plate tectonics (Besse et al., 1984; Yin, 2006). These mountain ranges penetrate the atmosphere and regulate monsoonal circulation (Sandu et al., 2019), tropical easterly jets, and river systems. The Intergovernmental Panel on Climate Change (IPCC) special report about the cryosphere has pointed out that the Himalayan snow cover has reduced, and their glaciers underwent substantial ice loss during the last half-decade (Hock et al., 2019; Maurer et al., 2019; Shukla & Sen, 2021). Previous studies predominantly concentrated on the Tibetan Plateau and its warming (e.g., Guo et al., 20162021; Rangwala et al., 2009). On the other hand, studies on the Himalayan hydroclimate and its variability are limited, which is an essential aspect of the Indian subcontinent. The Himalayan range comes under the Indian Monsoon framework, which is intensively studied and the most eminent domain of the global monsoon system. Himalayan rainfall distribution mainly depends on moisture availability via southwest monsoon flow and earns massive rainfall during the summer monsoon season (June to September, JJAS). Traditionally, the land-sea thermal contrast is a primary physical mechanism that drives monsoon circulation (Geen et al., 2020; Jin & Wang, 2017; Roxy et al., 2015). When this moist wind is lifted over the Himalayas and mountain ranges, it cools and condensates in the form of orographic rainfall. Due to complex elevation topography, synoptic features are complicated on the spatial and temporal scale over Himalayan regions and are characterized by the steep gradient (Figures 1a and 1b). Northeast India experiences the highest rainfall during the Indian monsoon, surpassing other parts of the Indian subcontinent (Mahanta et al., 2013). This region is also home to renowned locations like Cherrapunji and Mawsynram, recognized as the wettest places on Earth (Kuttippurath et al., 2021). Himalayan rainfall distribution mainly depends on moisture availability via southwest monsoon flow and earns massive rainfall during the summer monsoon season.

Details are in the caption following the image

Topographic elevation and demography of the Himalayan region. (a) Shaded surface relief (Topography surface of the Earth data at five-minute grid resolution; https://www.ngdc.noaa.gov/mgg/global/etopo5.HTML) with overlayed national borders in black color and surface relief contour intervals of 1 km in white color. (b) Shaded steepness was estimated using the slope raster method with surface relief contour intervals of 1 km in white color. (c) The shaded population density map highlights the exposed population in 2020 (gridded population of the world version 4 (GPWv4) population density; https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11) overlayed national borders in black color and river system in blue and cyan color.

In recent years, extreme monsoon in the Himalayan region has highlighted their societal significance. These remarkable extreme monsoons exhibit the substantial impact that changes in rainfall can have on various aspects of society, from agriculture sector to infrastructure and overall community resilience. Floods and droughts regularly occur in the monsoon regions, and eventually causes socio-economic consequences. Research has contributed to our understanding of the drivers of wet and dry monsoons in the South Asian region. These drivers encompass various factors, including large-scale circulation and oceanic patterns. While considerable progress has been made, discernible knowledge gaps merit further exploration. These gaps are particularly evident in comprehending the intricate interactions and forcing mechanisms that influence the monsoon dynamics that contribute to extreme monsoons. Studies have illuminated that the Indian monsoon is modulated by climate mode features such as El Niño–Southern Oscillation (ENSO) (Jian et al., 2009; Saini & Attada, 2023), Indian Ocean dipole (IOD) (Saji et al., 1999), Atlantic Niño (AN) (Sahoo & Yadav, 2021; Zebiak, 1993). The complex interplay among different natural variability modes and their combined influence on monsoon dynamics require further investigation, including their potential to heighten the intensity of monsoon rainfall. Himalayan monsoon rainfall has complexity due to its orographic features. This monsoon exhibits various temporal characteristics, including decadal variations (Halder et al., 2022; Kumar et al., 1999), interannual variability (Gadgil, 2003), annual patterns (Wang & Ding, 2008), seasonal variations (Goswami & Ajaya Mohan, 2001), and diurnal fluctuations (Goswami & Ajaya Mohan, 2001). Rainfall over the south-facing terrain of the Himalayas is distinguished by a diurnal cycle (Bhatt & Nakamura, 2006; Hunt et al., 2022). However, a recent study highlights drying due to amplification in the diurnal cycle (Norris et al., 2020). The analysis also confirmed the rising temperature over the Himalayas (Pepin et al., 2015; Sabin et al., 2020), along with decreasing trend in monsoon rainfall (Roxy et al., 2015). These all pointed to the minor role of monsoon and multi-year flooding that might be attributed to ice loss and snow melting. Himalaya is the source of Earth's major rivers, the Ganges and the Brahmaputra, essential water resources for the Indian subcontinent, which also provides irrigation and transportation in a densely populated region (Figure 1c). The Himalayan Rivers flood yearly during the monsoon season, especially the Brahmaputra. Thus, an advanced early warning system in the Himalayan ranges is necessary for policymakers and stakeholders.

In short, recent extreme monsoon rainfall underlines the societal importance of studying and comprehending such occurrences within the Himalayan region. While advances have been made in understanding monsoon drivers, knowledge gaps persist, necessitating a more nuanced exploration of the intricate factors shaping monsoon dynamics. Although the influence of IOD, ENSO, and AN on the Indian Monsoon has been well-documented, contributing dynamics to extreme monsoon variability over Himalaya still need to be explored. A comprehensive understanding of natural variability modes enables us to disentangle complexities associated with extreme monsoons. This can contribute to informed decision-making and preparedness efforts in South Asia. The analysis aims to highlight mechanisms by disentangling the natural variability linkage between extreme rainy seasons in the Eastern Himalayas. The structure of the current investigation is as follows. Section 2 outlines the data and methodology employed in the study. Section 3 focuses on the investigation of extreme monsoon seasons in the Eastern Himalayas, including an examination of the physical mechanisms underlying local processes during both wet and dry years. Additionally, we explore the influence of large-scale drivers on precipitation variability. Finally, in Section 4, we summarize the key findings and conclusions drawn from our study. In the final section, we deliberate on the ramifications of our discoveries and offer insights for potential avenues of research. We also discuss the implications of our discoveries and propose potential directions for future research.

2 Data and Methodology

Rainfall observation reliability across mountainous regions is limited (Hock et al., 2019; Zandler et al., 2019) due to less spatial coverage of situ measurement. On the other hand, the Summer monsoon representation employing global climate models over the Himalayas and downstream regions is still challenging (Palazzi et al., 2015; Pathak et al., 2019; Salunke et al., 2018). For this purpose, we utilize a variety of rainfall data sets available, including gauge-based observation, reanalysis products, and merged products from 1979 to 2021.

2.1 Data Availability and Coverage

Inspired by our research objectives, we thoroughly explored rainfall variability during the satellite era, aiming to gain comprehensive insights we focused on the period from 1979 to 2021, encompassing the timeframe relevant to our study. However, it is worth noting that due to long-term data limitations, we had access to a restricted set of data sets for our analysis. Despite this constraint, we tried to extract meaningful insights and draw reliable conclusions from the available data sources; CRU(Harris et al., 2020), ERA5 (Hersbach et al., 2020), MSWEP (Beck et al., 2019), and IMDAA (Indirarani et al., 2021) as given in Table 1. The topography elevation at a five-minute grid resolution (etopo5) is obtained from NASA. The global population density estimates in 2020 from the Gridded Population of the World (CIESIN, 2018) at a resolution of 15 arc-minute (approx. 30 km). We use atmospheric variables at 0.25° horizontal resolution from the European Center for Medium Range Weather Forecasting (ECMWF) reanalysis ERA5 (Hersbach et al., 2020). HadSST sea surface temperature (SST) obtained from the Met Office Hadley Center. Daily mean river discharge at 0.1° × 0.1° horizontal resolution reanalysis data downloaded from GloFAS-ERA5 (Harrigan et al., 2020) reanalysis.

Table 1. Summary of Data Sets Used for Rainfall Analysis (1979–2021)
Data (reference) Data type Horzional resolution Data source
CRU TS v4.06 (Harris et al., 2020) Global gauge-based product 0.5° × 0.5° Climatic Research Unit (CRU) https://crudata.uea.ac.uk/cru/data/hrg/
ERA5 (Hersbach et al., 2020) Global reanalysis 0.25° × 0.25° ECMWF https://cds.climate.copernicus.eu/cdsapp#!/home
MSWEP V2 (Beck et al., 2019) Global merge product (including gauge-based and satellite) 0.1° × 0.1° GloH2O website https://www.gloh2o.org/mswep/
IMDAA (Indirarani et al., 2021) Regional reanalysis 12 km × 12 km NCMRWF https://rds.ncmrwf.gov.in/datasets
  • Note. The table summarizes the various data sets utilized in the present study to comprehensively analyze rainfall patterns. The analysis covers the period from 1979 to 2021, enabling a thorough examination of long-term rainfall variability in the study region.

2.2 Rainfall Data Validation

We evaluated the performance of the above mentioned data sets (Table 1) in capturing the climatology (Figure 2) and variability of rainfall (Figure 3) over the Eastern Himalayas. In the manuscript, rainfall variability is defined as the standard deviation. The standard deviation is a statistical measure that quantifies the extent of variation in data set points from the mean value. In the context of rainfall variability, the standard deviation is used to assess how rainfall amounts deviate from the average or mean rainfall over a specific period of time. This approach allows for a straightforward and quantitative assessment of the variability in rainfall data and provides a clear understanding of how rainfall amounts vary across different regions at each grid points. Figure 3 illustrates the spatial variability of rainfall using reanalysis data from ERA5 and IMDAA, which were compared with observation data from CRU and the merged product data from MSWEP. The results demonstrate that these data sets exhibit similar climatological features (as given in Figure 2) and capture the spatial variability of rainfall in the region. Precipitation from ERA5, MSWEP, and IMDAA consistently indicates a notable variability in precipitation distribution across this region and is highly correlated (Table 2). Figure 3e illustrates the amplitude spectrum of Eastern Himalayan rainfall, highlighting the prominent interannual and decadal signals captured by all data sets. These signals reflect the inherent natural variability of the region's rainfall patterns. However, one can disentangle the individual contributions and assess the variability of these two components separately. However, The particularly CRU rainfall in Himalayan regions with complex terrain is still subject to uncertainties (Kanda et al., 2020; Thornton et al., 2022) and potential biases due to limited rain gauge coverage. Also, the Indian monsoon feature represented by both IMDAA and ERA5 reanalysis data is reliable with observations from IMD (Ashrit et al., 2020; Indirarani et al., 2021; Saini & Attada, 2023). Although the IMDAA regional reanalysis data set is known for its high resolution and reliability (Saini & Attada, 2023), we utilized ERA5 as a global reanalysis for consistency. In addition, ERA5 offers additional crucial variables relevant to our research objectives, including snow depth and river discharge.

Details are in the caption following the image

Rainfall climatology in the Himalayan region using different data sets. Panel (a) displays CRU data, panel (b) shows ERA data, panel (c) presents MSWEP data, and panel (d) showcases IMDAA data. Here black contour represents surface relief intervals, and the box represents the Eastern Himalayan region. The bottom panel (e) depicts detrended standardized anomalies for the respective data sets.

Details are in the caption following the image

Rainfall variability in the Himalayan region using different data sets. Panel (a) displays CRU data, panel (b) shows ERA data, panel (c) presents MSWEP data, and panel (d) showcases IMDAA data. Here black contour represents surface relief intervals, and box represent the Eastern Himalayan region. The bottom panel (e) depicts FFT power spectral density for the respective data sets over the Eastern Himalayan box, as indicated above.

Table 2. Correlation of Standardized Rainfall Anomalies (1979–2021)
CRU TS v4.06 1 −0.20 −0.19 −0.35
ERA5 −0.20 1 0.88 0.82
MSWEP V2 −0.19 0.88 1 0.65
IMDAA −0.35 0.82 0.65 1
  • Note. The table displays the correlation values of standardized rainfall anomalies for the period 1979 to 2021. These correlations are based on the analysis presented in Figure 2e, which depicts the detrended standardized anomalies of Eastern Himalayan rainfall.

2.3 Identification of Wet and Dry Monsoon Years

First, we considered detrended anomalies at each grid point in order to remove the influence of the annual climatological cycle, along with the linear global warming trend. To identify high variability regions within the Eastern Himalayan region, we specifically considered grid points exhibiting high variability exceeding an amplitude threshold of 2.8 mm per day, as illustrated in Figure 4b. Both rainfall and runoff in the Eastern Himalayan region display analogous hydroclimate patterns (Figures 4a and 4b), underscoring the dependency on precipitation and water hydrology in the Himalayan region. In order to gain a more comprehensive understanding of extreme monsoons, we conducted a composite analysis to explore underlying mechanisms. The Eastern Himalayas region holds high variability and is dominated by steep topographic elevation. The considerable rainfall variations over this region are highly associated with flooding and arid events. For the composite analysis (Figure 4c), we employed a threshold derived from the interannual standard deviation of the rainfall time series spanning the last 43 years. We created a time series of area-averaged detrended summer rainfall anomalies over the Eastern Himalayan region. If the detrended rainfall anomalies exceeded the threshold of +1.3 mm/day, the year was classified as a wet monsoon year. Conversely, if the anomalies fell below the threshold of −1.3 mm/day, the year was classified as a dry monsoon year. This analysis identified six wet monsoon years (1987, 1988, 1995, 1998, 2012, and 2020) and seven dry monsoon years (1981, 1992, 1994, 2001, 2005, 2006, and 2013). Since our sample sizes are small, we applied the bootstrap method (Aneesh & Sijikumar, 2020) on the composite to determine statistical significance at a 90% confidence level.

Details are in the caption following the image

Variability of the Himalayan hydroclimate. Panel (a) illustrates the shaded variability of runoff anomalies, and panel (b) displays rainfall anomalies over the last 43 years (1979–2021) using the ERA5 reanalysis data set. White contours represent steep topography. The red box in panel (b) highlights the Eastern Himalayan region. Panel (c) presents the time series of linearly detrended monsoonal rainfall anomalies over the Eastern Himalayas. Shaded green (light brown) indicates extreme wet (dry) monsoon years. For further composite analysis, we have considered years with rainfall anomalies exceeding the interannual standard deviation (dashed line). Black dots in Panels (a) and (b) indicate statistically significant regions at a 90% confidence level calculated using a bootstrap method.

2.4 Moist Static Energy

The moist static energy is used as a thermodynamic variable, which represents the addition of dry static energy and latent energy as:
h = C p T + g z + L v q $h={C}_{p}T+gz+{L}_{v}q$
where Cp is the specific heat at constant pressure, g is the gravitational constant, z is the height above the surface, and Lv is the latent heat of vapourization.

2.5 Buoyancy Diagnostics

The moist air parcel buoyancy approach has been taken from previous work (Pascale et al., 2017). To evaluate changes in the atmospheric convective instability, we calculate the buoyancy index (Fu et al., 2021; Pascale et al., 2017; Randall, 2015) at each horizontal grid point with a vertical level.
b = h 10 m h env 2 $b=\frac{\left({h}_{10\mathrm{m}}-{h}_{\text{env}}\right)}{2}$
where h 10 m = C p T 10 m + g z 10 m + L v q 10 m ${h}_{10\mathrm{m}}={C}_{p}{T}_{10\mathrm{m}}+g{z}_{10\mathrm{m}}+{L}_{v}{q}_{10\mathrm{m}}$ is moist static energy at a surface 10 m and henv is the environmental saturation moist static energy.

The anomalous buoyancy index Δb is taken with respect to mean state climatology. Positive values indicate upward, and negative values indicate downward acceleration.

2.6 Moisture Flux Convergence

We computed the three-dimensional Moisture flux convergence (MFC) as it can tell more about topographic features. The horizontal MFC can be expressed as follow:
MFC = · q V h $\text{MFC}=-\nabla \cdot \left({qV}_{h}\right)$
where, Vh(u, υ) is horizontal wind velocity; u and v are the zonal and meridional components of the wind.
Furthermore, Anomalous MFC can be decomposed into dynamical MFC and thermodynamical MFC. Delta indicates the anomaly with reference to mean state climatology.
· q V h = · q V h · q V h ${\increment}\left(-\nabla \cdot \left({qV}_{h}\right)\right)=-\nabla \cdot \left(\overline{q}{\increment}{V}_{h}\right)-\nabla \cdot \left({\increment}q\overline{{V}_{h}}\right)$

2.7 Dynamical and Thermodynamical Parts of Rainfall

We performed vertically-integration of MFC to get its dynamical and thermodynamical parts (Oueslati et al., 2019) of rainfall,
dynamical part = · q V h $\text{dynamical}\,\text{part}=\left[-\nabla \cdot \left(\overline{q}{\increment}{V}_{h}\right)\right]$
thermodynamical part = · q V h $\text{thermodynamical}\,\text{part}=\left[\nabla \cdot \left({\increment}q\overline{{V}_{h}}\right)\right]$

The use of brackets in this context indicates a mass-weighted vertical integration from the surface to 100 hPa at the upper of the atmosphere. It involves calculating the integral of a quantity A, represented as [ A ] = 1 g S u r f a c e 100 h P a A d p g $[A]=\frac{1}{g}\,\underset{Surface{}}{\overset{100hPa}{\int }}A\frac{dp}{g}$ where g denotes the acceleration due to gravity.

2.8 Local Hadley Circulation

We consider the mass stream function (Peixto & Oort, 1984) to understand the mean local meridional circulation. The local meridional mass stream function is expressed as follows:
Ψ = 2 π a cos ϕ g p p s V d p ${\Psi }=\frac{2\pi \mathrm{a}\,\cos \,\phi }{g}\underset{p{}}{\overset{{p}_{s}}{\int }}\overline{V}dp$
where a is the Earth's radius, ϕ is latitude, g is the acceleration due to gravity, V $\overline{V}$ is the zonal mean meridional velocity, p is the pressure, and Ps is the surface pressure.

2.9 ITCZ Location

Intertropical Convergence Zone (ITCZ) location (Byrne & Schneider, 2016) is defined as the latitude closest to the equator where the zonal mean streamfunction vertically averaged between 700 and 300 hPa is zero.
ϕ ITCZ = ϕ Ψ = 0 ${\phi }_{\text{ITCZ}}={\phi }_{\int {\Psi }}=0$

2.10 ITCZ Width

The ITCZ width (Byrne & Schneider, 2016) is defined as the latitude distance between the maxima and minima points using the zonal mean streamfunction vertically averaged between 700 and 300.
ϕ width = ϕ max ϕ min ${\phi }_{\text{width}}=\left({\phi }_{\max }-{\phi }_{\min }\right)$

3 Results and Discussions

3.1 Natural Variability of Himalayan Hydroclimate

Monsoon rainfall variability (Figure 4b) dominates in the steep topography of Himalaya, extending from west to east. The unique orientation of the Himalayan mountain range results in a wider distribution of rainfall across the Eastern Himalayas. To further understand the variability of Himalayan rainfall, we look at composite maps for six wet and seven dry monsoons. Anomalous rainfall patterns are almost identical in the eastern Himalayas, showing consistent signs in wet and dry years (Figures 5a and 5b). A similar feature is replicated in runoff anomalies (Figures 5c and 5d); river discharge is water streaming through a river routing. Rainfall anomalies are responsible for this runoff, which can further aggravate river flood hazards (Jian et al., 2009). Brahmaputra river flooding years match a previous study (see ref. (Jian et al., 2009; Rao et al., 2020)) reflecting the role of natural climate variability. However, this phenomenon is not restricted to the Brahmaputra basin alone (Figures 6a and 6b); it extends across the entire Himalayan drainage system. The distinctive pattern of anomalous surface runoff aligns with the steep southeastern region of the great Himalayan range (Refer to Figures S1a and S1b in Supporting Information S1), highlighting the integral role of elevation topography in influencing dry or wet years. However, snow melting (Figures 5e and 5f) mostly shows a reduction upstream of the Himalayas. Its contribution to natural variability seems less than monsoon rainfall in dry and wet monsoons. Additionally, the observed pattern of snow melting (Figures 5e and 5f) does not correspond with the rainfall composite map (Figures 5a and 5b). This misalignment suggests that the process of snow melting might be influenced by factors related to global warming. In other words, the way snow melts doesn't seem to be solely explained by rainfall patterns, and the influence of global warming is likely playing a role in this process.

Details are in the caption following the image

Composite analysis for monsoonal hydroclimate during 1979–2021. (a) Composite map of rainfall anomalies for wet monsoon and (b) dry monsoon years using ERA5 (Hersbach et al., 2020) data set. (c) Composite map of runoff anomalies for wet and (d) dry years. (e) Composite map of snowmelt anomalies for wet and (f) dry years, multiply by 5 factor. Here black contour represents surface relief intervals from 1,000 to 4,000 m.

Details are in the caption following the image

The composite map displays anomalies in river discharge during wet and dry monsoons. In (a), the shaded area represents wet monsoons, while (b) represents dry monsoons. The river discharge data utilized in this analysis is obtained from GloFAS-ERA5, a widely used global river discharge data set. Here black contour represents surface relief intervals. Figure (c) depicts detrended anomalies of the water budget, with the blue line representing rainfall, the orange line representing snowmelt (multiplied by a factor of 5), and the maroon line representing river discharge. Additionally, correlation coefficients are provided with river discharge.

The composite map presented in Figures 6a and 6b shows the anomalies in river discharge during wet and dry monsoons—the river discharge variations in river water flow during these different monsoon conditions. In this region, the Ganges and Brahmaputra rivers exhibit a strong connection with the variability of the Himalayas. Analyzing the anomalies in river discharge strongly suggests the changes in water availability during wet and dry monsoons, providing insights into the hydrological response to monsoonal variability. In order to validate the results obtained from the composite analysis of river discharge anomalies, Figure 6c presents the detrended anomalies of the water budget, including river discharge, rainfall, and snowmelt (scaled by a factor of 5). Correlation coefficients are also provided to evaluate the relationship between these variables and river discharge. The obtained high correlation coefficient between rainfall and river discharge (Figure 6c) signifies the major contribution of rainfall to the overall river discharge. This finding aligns with the results derived from the composite map, which also indicated a strong association between rainfall and river discharge. On the other hand, the lower correlation between snowmelt and river discharge confirms the findings from the composite map analysis of snowmelt, suggesting a comparatively weaker influence of snowmelt on river discharge. By examining the detrended anomalies and assessing the correlation coefficients, these results help validate the outcomes of the composite analysis and provide additional evidence for the dominant role of rainfall in influencing river discharge.

3.2 Local Atmospheric Moist Features

Local atmospheric conditions and topography significantly modulate rainfall events (Kad et al., 2023; Zhang & Liang, 2020). To understand the characteristics and atmospheric key features over the East Himalayan region, we have taken a cross-section at 94°E from the valley to the upslope terrain of the Himalayas. We further provide a comprehensive explanation for these spatial patterns, considering the influence of multiple factors that contribute to the observed distribution, including topography and moist processes. The vertical profile climatology is illustrated in Figures 7a–7d, indicating how the elevation of the Himalayas impacts atmospheric variables. Typically, the temperature decreases with increasing height in the troposphere and declines vertically in pressure, as illustrated in the Thermal structure shown in Figure 7a. The primary driver of precipitation is the atmospheric humidity over the mountainous terrain (Kad et al., 2023; Smith, 2018; Tao et al., 2020), with the large-scale monsoonal circulation also playing a significant role. Figure 7b illustrates that relative humidity is notably higher over steep elevations, ranging from 80% to 90% in the lower atmosphere. Figure 7c displays the vertical distribution of MSE, which illustrates regions with lower energy levels concentrated over mountain peaks and areas with higher energy levels observed in valley regions (Kad et al., 2023). The composite analysis further unveils that warming temperature anomalies accompanied by elevated relative humidity and MSE contribute to wet years (as illustrated in Figures 7e–7g). Conversely, cooling temperature anomalies with declined relative humidity and MSE are responsible for dry years (Figures 7i–7k). Usually, environmental moist static energy enhances buoyancy, and Figure 6d shows the least buoyancy over valley regions, followed by terrain. Also, anomalies in moist buoyancy representation show that positive values indicate upward buoyancy over the steep topography during wet years, and negative indicates downward buoyancy during dry years (Figures 7h and 7l). This suggests the need to employ the moist dynamics analysis to interpret correctly.

Details are in the caption following the image

The thermal structure of climatological mean state and composite anomalies. The cross-section is along the 94°E, with shaded quantities representing temperature, relative humidity, moist static energy, and buoyancy. Panels (a)–(d) correspond to the mean state climatology of these parameters, with their respective contour intervals and units highlighted in green. Panels (e)–(h) display the composite anomalies during wet monsoons, while panels (i)–(l) represent the composite anomalies during dry monsoons. The black color indicates the masked region of elevation topography.

We found a distinctive anomalous MFC pattern in a cross-section at 94°E at a lower level, as results shown in Figures 8a and 8d. The MFC in mountainous terrain has important implications for weather and climate patterns. Regions with positive MFC lead to increased rainfall and the formation of clouds, while regions with low MFC lead to dry conditions and drought. This anomalous MFC shows an increased positive anomaly in wet years (Figure 8a) while a reduction in dry years (Figure 8d). Furthermore, these anomalous MFC patterns follow steep Himalayan terrain, indicating that areas with steeper terrain tend to have distinct orographic features (Figure S2 in Supporting Information S1). The beauty of this 3-dimensional MFC is that we can diagnose a cross-section to comprehend the vertical distribution and can see distinctively control of the elevation configuration. Furthermore, we decompose MFC in the dynamical and thermodynamical parts, which relate to the circulation effect and the moisture effect, respectively. Most of these changes are contributed by dynamical MFC (Figures 8b and 8e), which suggests an essential role of circulation. Likewise, Thermodynamical MFC (Figures 8c and 8f) shows a similar agreement dominated over slope terrain. Although the Thermodynamic MFC shows the same pattern as the MFC, its magnitude is relatively small and about 10 times less than that of the dynamical MFC. The anomalous MFC clearly indicates an enhancement of processes between the surface to 500 hPa, underscoring the role of moisture. These results reveal that dynamical MFC modulates extreme rainfall anomalies in the steep terrain of the Himalayas. The observed changes in MFC pointing toward the strong influence of monsoonal circulation during extreme wet and dry years. The results from Figure 8g indicate the rainfall anomalies observed during the study period, which also provides the decomposition (Oueslati et al., 2019) of rainfall into two components: dynamical and thermodynamical. The high correlation coefficient of 0.97 between dynamic rainfall and overall rainfall suggests a strong and positive relationship between these two variables. This indicates that changes in the dynamical part, such as atmospheric circulation patterns and weather systems, have a significant influence on the observed rainfall anomalies. Dynamic processes play a dominant role in driving rainfall variability. On the other hand, the weak correlation coefficient of 0.10 between the thermodynamical part and overall rainfall suggests a weak and possibly non-linear relationship between these two variables. The thermodynamical part, which is related to factors like moisture availability and energy balance, might have a less pronounced influence on the observed rainfall anomalies compared to the dynamic component. Overall, the results highlight the dominant role of dynamic processes in driving rainfall anomalies.

Details are in the caption following the image

Anomalous moisture flux convergence (MFC) for composite wet and dry monsoons. The cross-section was taken along 94°E, with shaded quantities representing MFC, dynamical MFC, and thermodynamical MFC. Panels (a)–(c) depict composite wet years, while panels (d)–(f) display composite dry years. Note that the thermodynamical MFC is multiplied by 10. The black color indicates the masked regions of elevation topography. Panel (g) shows rainfall anomalies represented by the black line, with its dynamical part in the green line and thermodynamical part in magenta.

3.3 Monsoonal Hadley Cell

We illustrated local monsoonal Hadley circulation to understand circulation linkage with dynamical MFC during both dry and wet monsoons. Our analysis focused on the zonal mean meridional mass stream function, which offers valuable information about the localized monsoonal circulation. To accomplish this, we averaged the stream function over the longitudinal range of 70°E to 102°E, which corresponds to the study area. During the monsoon season, the monsoonal Hadley cell exhibits a counterclockwise rotation pattern in the Northern Hemisphere. Figures 9a and 9b provide visual evidence of this rotation extending from the equatorial Indian Ocean region to the Himalayan mountain ranges. In contrast to this counterclockwise rotation, about 30° North over the Tibetan Plateau, there is a clockwise rotation phenomenon. This clockwise rotation is a consequence of the elevated topography of the region, creating a high-pressure system that forces air to move downward and outward, resulting in the observed clockwise rotation. The counterclockwise rotation is more pronounced during wet years, as clearly displayed in Figure 9a with a distinct core. Conversely, the counterclockwise rotation appears relatively subdued during dry years, as shown in Figure 9b. Overall, the analysis of the zonal mean meridional mass stream function in this study reveals the distinctive circulation patterns associated with wet and dry monsoons within the study region.

Details are in the caption following the image

Mean local Hadley circulation in wet and dry monsoon years. The zonal mean meridional mass stream function averaged over the range of 70°E to 102°E indicates the local monsoonal circulation. Panel (a) shows the stream function for composite wet years, while panel (b) shows it for composite dry years. Negative values indicate counterclockwise circulation, while positive values indicate clockwise circulation. The gray color represents masked regions of elevation topography (averaged over 70°E to 102°E). In panel (c), the vertically averaged zonal mean meridional stream function is shown from 700 to 300 hPa. The dashed line represents the location of the ITCZ, while the upper and lower empty triangles indicate the latitudinal points of maxima and minima, respectively.

Consequently, the mean local Hadley cell exhibits narrow characteristics during wet years and extends wider during dry years. This meridional mean overturning circulation consists of an ascending branch of warm moist air, commonly recognized as a tropical rain belt or ITCZ. The displacement of ITCZ influences Himalayan rainfall variability and also defines the amount of precipitation within the Himalayan region. In the pursuit of comprehending the Monsoonal Hadley Cell (Geen et al., 2020), our study delved into investigating the position of the ITCZ and its associated width (Byrne & Schneider, 2016), as depicted in Figure 9c and Table 3. Climatologically, the ITCZ is situated at a latitude 30.51° North of the equator with an average width of 756.2 km. The wet monsoon composite analysis indicates ITCZ location shifted by 0.49° latitude to the north (Hari et al., 2020) with a narrow width. A narrower ITCZ can increase rainfall amounts in the Himalayan foothills, potentially leading to localized flooding. Conversely, composite analysis for dry monsoons portrays ITCZ location remains comparable to the 43-year climatological mean. This suggests that the ITCZ remains relatively stationary during dry years, resulting in a wider width. A wider ITCZ can result in reduced rainfall, potentially leading to drier conditions and possible droughts. However, the atmospheric tropical bridge is more important, which establishes a connection between background tropical ocean SST anomalies. This elucidation enhances our understanding of the intricate interplay between the ITCZ, the Hadley cell, and the monsoonal dynamics that impact the Himalayan region. It also offers valuable insights into the underlying mechanisms driving wet and dry monsoons, contributing to a more comprehensive comprehension of the region's climate dynamics.

Table 3. ITCZ Location and Width
Temporal scale ITCZ location (°N) ϕ I T C Z ${\phi }_{ITCZ}$ ITCZ width (km) ϕ width ${\phi }_{\text{width}}$
Climatological mean 30.51 756.2
Extreme Wet Monsoon (Δ) 31.00 (0.49) 513.4 (−242.8)
Extreme Dry Monsoon (Δ) 30.71 (0.2) 797.8 (41.60)
  • Note. This table provides information on the mean location and width of the Intertropical Convergence Zone (ITCZ) based on the analysis conducted in the study (For more information, please refer to Figure 9c).

3.4 Background Tropical Ocean SST

Our study examined the relationship between rainfall and the spatial pattern of SST anomalies in the Eastern Himalayan region. The location of SST anomalies in the tropical Ocean plays a significant role in driving changes in orographic rainfall (Halder et al., 2022). Figure 10a presents the correlation analysis between East Himalayan rainfall and tropical SST, revealing combined patterns in the Indian Ocean, Atlantic Ocean, and Pacific Ocean associated with the natural fluctuation in rainfall. This correlation analysis helps quantify the relationship and identify potential patterns or trends between rainfall and SST anomalies. The correlation coefficient in estimates the strength of the relationship between rainfall and SST in the Eastern Himalayan region, the stippling in the Figure 10a indicates statistically significant correlations at the 90% confidence level. We observed that areas with positive SST anomalies are associated with positive correlations. In contrast, areas with negative SST anomalies are linked to negative correlations. However, The magnitude and spatial extent of these relationships vary across the region for individual monsoons. However, our results indicate that the Indian Ocean and South China Sea predominantly influences the natural variability of rainfall in the Eastern Himalayan region. Tropical SST condition has been examined for composite wet and dry monsoons (in Figures 10b and 10c). The warm SSTs in the Atlantic, West Pacific, and Indian oceans are favorable for Himalayan wet monsoons. A recent teleconnection study also found that Atlantic Nino enhances the MFC over northeast India (Sahoo & Yadav, 2021). Atlantic and Indian Ocean cooling and weak La Niño like conditions seem favorable for Himalayan dry monsoons. The amplified SST anomalies observed in the tropical ocean basin underscore the potential significance of the Indian, Atlantic, and Pacific Oceans in influencing these rainfall responses. While significant tropical SST anomalies can have an impact on atmospheric circulation patterns, our study emphasizes the SSTs anomalies in the Atlantic and Indian Oceans are likely to influence the strength of ITCZ and local Hadley cell. However, when assessing the significance of the relationship, we found that the majority of the significant correlations were observed with Indian Ocean SST anomalies (Figure 10a), coinciding with the presence of our local Hadley cell over the Indian Ocean region. This suggests that the Indian Ocean (See Figure S3 in Supporting Information S1) plays a crucial role in driving the variability of rainfall in the East Himalayan region. This underscores the need for a localized perspective when examining the drivers of monsoonal circulation and their relationship with tropical SST anomalies. The composite analysis results indicate that the anomalous SST patterns contribute to the local monsoonal Hadley circulation, which plays a crucial role in regulating precipitation variability in the Eastern Himalayan region.

Details are in the caption following the image

Relationship between East Himalayan rainfall and spatial pattern of sea surface temperature (SST) anomalies. Panel (a) the correlation between rainfall and tropical SST using the HadSST data set. Stippling is used to indicate correlations that are significant at a 90% confidence level. SST anomalies in panel (b) for wet and (c) dry monsoon years.

4 Conclusions

The insights gained from our study are pivotal in unraveling the intricate dynamics of the monsoon system. In summary, the study focused on understanding the natural variability of orographic monsoon rainfall in the Eastern Himalayan region and identifying its potential drivers for wet and dry monsoons. The key conclusions derived from the analysis are as follows:
  1. The study emphasized the importance of considering multiple data sets to ensure comprehensive rainfall analysis, with data sets like CRU, ERA5, IMDAA, and MSWEP covering the same periods and providing valuable insights into rainfall variability.

  2. The Eastern Himalayan region exhibits significant natural variability in orographic monsoon rainfall over steep relief and dominated over south-facing slopes, with distinct wet and dry rainy years identified through composite analysis.

  3. Analysis revealed that rainfall is the major contributor to river discharge in the region, indicating the dominant influence of rainfall on the hydrological system. Correlation analysis further supported this finding, indicating a strong relationship between rainfall and river discharge. In contrast, the role of snowmelt in contributing to river discharge was found to be minor.

  4. Our study underscores the pivotal role of dynamical MFC in driving the variability of Himalayan monsoon rainfall. This influence is substantiated by the observed relative shift in the position of the ITCZ. Importantly, the width of the ITCZ also plays a crucial role: it is narrower during wet years due to the influence of a strong Hadley circulation and wider during dry years as a result of a weaker Hadley circulation. These findings emphasize the crucial role of large-scale dynamics in shaping the observed local atmospheric thermodynamic conditions.

  5. Spatial patterns of SST anomalies within the tropical Ocean have unveiled momentous connections with rainfall variability in the Eastern Himalayas. Through composite analysis, we have uncovered that these tropical SST anomalies play a pivotal role in influencing the strength of the local monsoonal Hadley circulation and exert a regulating effect on the precipitation variability of the Eastern Himalayas. These findings offer significant insights into the interplay between oceanic and atmospheric factors that contribute to the complex monsoonal dynamics of the region.

This study can be helpful in the predictability of Himalayan rainfall variability and extremes. This understanding not only sheds light on the complexity of the orographic monsoon processes but also holds the key to a deeper comprehension of the broader climatic interactions. By delving into these complexities, we pave the way for a more comprehensive understanding of the monsoon dynamics and implications. The findings emphasize the need for further research and modeling efforts to quantify better and understand the complex interactions between local and large-scale processes driving orographic monsoon variability in the region. Even though our study emphasizes natural variability, there is a need to explore in detail the role of climate change as a further study considering the observed intensification (See Figure S4 in Supporting Information S1) of the monsoon after 2002 (Hari et al., 2020; Jin & Wang, 2017). The Eastern Himalaya are anticipated to experience increased precipitation in the future due to greenhouse warming (Kad et al., 2023). Additionally, it is important to note that uniform Indian Ocean warming (Dhame et al., 2020) is expected in response to global warming. This warming could potentially contribute to an increased extreme monsoon and associated flooding events in Himalayan rivers. The Brahmaputra seems more significantly impacted during amplified monsoonal wet years in the Himalayas are favorable for flood risk downstream. The flooding in this region can be the ultimate red alert for Humanity and our ecosystem (Elsen et al., 2020), and threats like degradation of soil (Borrelli et al., 2020) and biodiversity loss (Peters et al., 2019). Thus, an advanced early warning system in the Himalayan ranges is necessary for policymakers and stakeholders. The Brahmaputra and Ganges merge in Bangladesh and flow into the Bay of Bengal. This natural variability would have its freshwater signature into the Bay of Bengal, which might be fascinating to explore further.


This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (Grant 2020R1A2C2006860). KJH was supported by the Institute for Basic Science (IBS-R028-D1). PK acknowledges the Korea National Park Service (KNPS) for igniting inspiring ideas through mountain hiking explorations. Both authors wish to acknowledge the valuable contributions of anonymous reviewers and the editor, whose constructive and critical feedback greatly enhanced the quality of this paper. Authors gratefully acknowledge NCMRWF, Ministry of Earth Sciences, Government of India, for IMDAA reanalysis. IMDAA reanalysis was produced under the collaboration between the UK Met Office, NCMRWF, and IMD with financial support from the Ministry of Earth Sciences, under the National Monsoon Mission programme. The authors wish to acknowledge the use of the PyFerret program for analysis and graphics in this paper. PyFerret is a product of NOAA's Pacific Marine Environmental Laboratory (Information is available at https://ferret.pmel.noaa.gov/Ferret/). The authors acknowledge the use of Climate Data Operators (CDO), a powerful software package for processing and computing climate data.

    Data Availability Statement

    Earth topography five-minute grid etopo5 (Maus et al., 2007) is publicly available at National Geophysical Data Center, https://www.ngdc.noaa.gov/mgg/global/etopo5.HTML. The global population density estimates in 2020 (CIESIN, 2018) from the Gridded Population of the World version 4 (GPWv4), https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11. CRU rainfall (Harris et al., 2020) data is publicly available on Climatic Research Unit (CRU) are available at https://crudata.uea.ac.uk/cru/data/hrg/. IMDAA (Indirarani et al., 2021) reanalysis data is publicly available on NCMRWF portal: https://rds.ncmrwf.gov.in/. ERA5 (Hersbach et al., 2020) reanalysis data is publicly available from the ECMWF on their Climate Data Store (CDS), https://cds.climate.copernicus.eu/cdsapp#!/home. Multi-Source Weighted-Ensemble Precipitation (MSWEP) rainfall product (Beck et al., 2019) data from GloH2O is publicly available, http://www.gloh2o.org/mswep/. The GloFAS-ERA5 river discharge (Harrigan et al., 2020) reanalysis product is publicly available on the CDS, https://cds.climate.copernicus.eu/cdsapp#!/dataset/cems-glofas-historical?tab=overview. HadSST (Kennedy et al., 2019) data are available at the Met Office Hadley Centre website, https://www.metoffice.gov.uk/hadobs/hadisst/.


    The originally published version of this article contained a few typographical errors. The following changes should be made to the mentions of the panels in Figure 7: “(a)–(c)” should be “(a)–(d),” “(d)–(f)” should be “(e)–(h),” and “(g)–(i)” should be “(i)–(l).” In the caption for Figure 3, “The bottom panel depicts” should be “The bottom panel (e) depicts.” In the caption for Figure 5, “and (c) dry monsoon years using ERA5” should be “and (b) dry monsoon years using ERA5.” In the caption for Figure 8, panel “(b)–(f)” should be panels “(d)–(f)”. In the caption for Figure 10, panel “(c)” should be “(b)” and panel “(d)” should be “(c).” In the note to Table 2, “depicts the amplitude spectrum of Eastern” should be “depicts the detrended standardized anomalies of Eastern.” In the last sentence of the first paragraph of Section 3.2, the sentence “dry years (Figures 7h–7l)” should be “dry years (Figures 7h and 7l).” In the thirteenth sentence of the second paragraph of Section 3.2, the sentence “contributed by dynamical MFC (Figures 8c and 8d)” should be “contributed by dynamical MFC (Figures 8b and 8e).” In the fourteenth sentence of the same paragraph, the sentence “Thermodynamical MFC (Figures 8e and 8f)” should be “Thermodynamical MFC (Figures 8c and 8f).” In the fourteenth sentence of the first paragraph of Section 3.3, the sentence “during dry years, as shown in Figure 9a” should be “during dry years, as shown in Figure 9b.” In the sixteenth sentence of the first paragraph of Section 3.4, the sentence “dry monsoons (in Figures 10a and 10b)” should be “dry monsoons (in Figures 10b and 10c).” The errors have been corrected, and this may be considered the authoritative version of record.