Hydrological Extremes in Western Himalayas-Trends and Their Physical Factors

Abstract

Recent exacerbation of extreme precipitation events (EPEs) and related massive disasters in western Himalayas (WH) underpins the influence of climate change. Such events introduce significant losses to life, infrastructure, agriculture, in turn the country’s economy. This chapter provides an assessment of long-term (1979–2020) as well as recent changes (2000–2020) in precipitation extremes over WH for summer (JJAS) and winter (DJF) seasons. Different high-resolution multi-source climate datasets have been utilized to compute the spatiotemporal trends in intensity and frequency of EPEs. The hotspots of rising extremes over the region have been quantified using the percentile approach where daily precipitation exceeds the 95th percentile threshold at a given grid. The findings reveal geographically heterogeneous trends among different datasets; however, precipitation intensity and frequency show enhancement both spatially and temporally (though insignificant). For both seasons, dynamic and thermodynamic parameters highlight the role of increased air temperatures and, as a result, available moisture in the atmosphere, signifying the consequences of global warming. Rising precipitation extremes in summer are sustained by enhanced moisture supply combined with increased instability and updraft, due to orography, in the atmosphere whereas winter atmosphere is observing an increase in baroclinicity, available kinetic energy, vertical shear and instability, contributing to a rise in precipitation extremes.

1. Introduction

The Western Himalayas (WH) are a strong modulator of weather and climate patterns over northern India and surrounding regions. WH are highly rich in biodiversity and covered with forests, agricultural landscapes, glaciers, wetlands and urbanized land, which underlines the significance of the region. The regional spatio-temporal distribution of precipitation over WH depicts high variability [1, 2] which can be partly attributed to influences from the atmosphere-land surface exchange processes over the region, keeping in mind the diverse land surface characteristics and large geographical variability [3, 4]. Additionally, complex interplay of regional topography with moist airflow and temperature gradient magnifies this variability further [5, 6, 7].

WH receives precipitation during both summer and winter monsoons [8]. The Indian summer monsoon (ISM), spanning through June-September (JJAS), contributes about 67–75% to the annual precipitation received over WH [2]. These mountainous environments have a significant impact on the spatiotemporal distribution of precipitation [1, 9]. Majorly, precipitation over the region during the ISM is contributed by convection followed by an orographically locked system, with Himalayas as barriers forcing moisture-laden southwest monsoon winds to dissipate moisture. Additionally, strong Tibetan high combined with the monsoon trough in northern India, creates a strong moist flow from Bay of Bengal and Arabian Sea into the Himalayas [10, 11]. Occasionally, interaction of tropical monsoon depressions and extratropical disturbances leads to the formation of heavy precipitation over the region during summer [12].

The region also receives a significant amount of precipitation during the winter season (December to February), primarily through extra-tropical cyclonic systems called Western Disturbances (WD; e.g. [1, 13]). WDs, embedded in upper tropospheric sub-tropical westerly jet stream, propagate eastward towards WH carrying moisture mainly from Mediterranean sea, Caspian sea, and Black sea (e.g. [13, 14, 15]). The interaction of these disturbances with elevated topography of WH results in their intensification and subsequent precipitation [16]. This precipitation holds key significance for sustenance of regional glaciers through snow accumulation and agricultural activities. The glacial mass balance is particularly crucial for regional river runoff and flows [17]. Any major precipitation variations in these glaciers can lead to adverse consequences on freshwater availability in downstream areas (e.g. [18]). Moreover, the vast river basin of WH acts as a watershed for a large population, assists in sustaining the regional biodiversity and provides various key ecosystem services to the surrounding north Indian plains. However, changing climate and expected hydroclimatic variability raises serious concerns related to impacts on this richly biodiverse and fragile mountainous landscape.

WH is highly prone to extreme precipitation events (EPEs) due to its intricate topography and altitude-dependent climate [10, 19, 20], which can give rise to surface runoff during such events, causing additional natural hazards such as landslides and floods (e.g. [21, 22]). Sharp regional weather fluctuations over the Himalayas makes this region unpredictable leading to sudden occurrences of heavy precipitation events. Various states in WH including Uttarakhand (UK), Jammu and Kashmir (J&K), Himachal Pradesh (HP), often face the problem of river flooding and landslides due to torrential downpours and localized occurrences of intense precipitation along the southern slopes of the Himalayas. Several case studies of EPEs over WH have highlighted massive losses through cloudbursts triggered by terrain-locked deep convective systems in valleys, as well as flash floods triggered by extratropical disturbances [23, 24]. Many scientific reports suggest an enhancement of precipitation in the order of 5–20% in the Himalayas in the 21^st century (e.g. [25]). Moreover, the changes in the intensity and frequency of EPEs may vary seasonally. The increased susceptibility of WH to heavy precipitation during ISM has been discussed in various studies (e.g., [22, 26, 27, 28, 29, 30, 31, 32]). The variability aspects of WH winter precipitation under climate change scenarios, including a possibility of enhancement in precipitation extremes has also been frequently highlighted [14, 33, 34]. An increase in avalanche activity over WH slopes related to enhanced frequency of wet-snow conditions during recent decades has also been reported by [35]. Being thickly populated, WH and surrounding regions are highly vulnerable to climate extremities. Such extreme precipitation events (EPEs) can affect both natural and anthropogenic ecosystems through damage to life, infrastructure, agriculture, energy sectors, etc.

Many massive disasters related to unexpected heavy precipitation such as Leh flood (August 2010), Kedarnath disaster (June 2013), Chamoli river floods in Uttarakhand (July 2016), Nadum disaster (August 2018), Bilaspur and Shimla floods (August 2019), Jammu & Kashmir floods in Kishtwar district (July 2021) and Dharmsala floods (August 2022) have been reported in the recent times during the ISM, leading to immense life and economic losses. Increased frequency of summer EPEs (see [10, 36]) has been attributed to a number of possible causes including enhanced water vapor transport in the northern Hindu Kush Himalayan region [37], formation and movement of local deep convective systems along the orography [10, 23, 24], convergence of low-level monsoon westerly winds and northeasterly winds along the foothills combined with enhanced vertical wind shear and interaction of tropical systems with extratropical disturbances (see [9, 10, 12, 23, 26, 34]). Furthermore, elevated Tibetan plateau plays a key role during ISM by producing mesoscale precipitation through small-scale circulations and enhancing the synoptic weather conditions, leading to extreme precipitation in the WH [19, 38, 39]). Additionally, [12] found that EPEs over the Himalayas are associated with a southward extension of the western upper trough and a simultaneous northward migration of the lower monsoon trough towards the foothills of the Himalayas. These two systems amplify the low-level moisture flux from the Arabian Sea and Bay of Bengal towards the foothills of the Himalayas and contribute to cloud development and heavy precipitation.

Studies pertaining to the wintertime trends of extreme precipitation intensity and frequency are comparatively fewer. Shekhar et al. [40] reported a significantly increasing trend for heavy precipitation events (>70 mm) in the Pir-Panjal range of WH, however, other altitudes and ranges did not portray any clear trends. A significantly increasing trend for winter to early spring precipitation extremes intensity (exceeding 90th percentile threshold) has also been observed by [14] between 1900 and 2011, attributable to higher baroclinicity and in turn enhanced variability of WD activity during the recent decades. Krishnan et al. [33] also studied the impact of climate change on WD activity over WH and reported an increasing trend of winter precipitation extremes in the recent decades appertaining largely to anthropogenic forcings in conjunction with natural factors. Increasing trends of different extreme precipitation indices (exceeding 90^th and 95^th percentiles) using station-based records were further observed by [41] for all western Himalayan ranges during the month of February. Further, [42] demonstrated an increased frequency of atmospheric rivers (ARs) over the Himalayas during winter and underlined the association of intense ARs with EPEs over the Ganga and Indus basins.

Weather and climate extremes are generally a result of variations in different atmospheric dynamic and thermodynamic variables, as well as of some surface properties or states. Several studies, involving observations or climate model simulations, indicate that the frequency of these events would intensify with global warming due to an increment in the atmospheric moisture holding capacity as per Clausius Clapeyron relationship [43, 44, 45, 46]. Additionally, the influence of local thermodynamics and orographic forcing in the WH produces abrupt changes in synoptic circulation, which have the potential to produce EPEs that can last for a few days [12, 47].

However, certain limitations are associated with the study of precipitation extremes over complex topographic regimes of WH. The remoteness of the region and the sparse coverage of rain gauges and automatic weather stations in mountainous areas makes precipitation monitoring in this region quite difficult [10, 48, 49]. The lack of availability of data directly affects the research studies investigating EPEs to under-perform in the Himalayan region. As a result, studies of extreme precipitation patterns, trends and possible causes in the WH remain limited and insufficient. Furthermore, very few studies have been conducted on the spatial distribution of EPE trends over WH. The inadequate representation of regional orography in the available coarser resolution datasets adds to the uncertainties associated with assessment of EPEs over any region [50]. Most of the past studies for EPEs over WH utilize comparatively coarser resolution datasets and there is a lack of high-resolution data-based studies for these extremes. Thus, it becomes important to utilize high-resolution datasets for a finer and accurate understanding of extreme precipitation trends and their possible causes over WH. The latest advancements in meteorological satellites and reanalysis products at finer resolutions with improved precipitation estimation algorithms have further facilitated the research on extreme weather events.

Our study focuses on changes in the spatial and temporal distributions of extreme precipitation events over WH during 1979–2020 in various high-resolution multi-source climate datasets, including the potential of recently released high-resolution regional reanalysis, Indian Monsoon Data Assimilation and Analysis (IMDAA). We evaluate extreme precipitation distribution and associated changes in key atmospheric parameters of EPEs over WH. Such knowledge about the climatological features of precipitation extremes and their associated dynamical and thermodynamic changes is crucial to interpret how precipitation patterns are changing in a varying climate scenario, further giving way to carry out vulnerability impact assessment studies.

2. Data and methods

2.1 Data used

In this study, we analyze daily summer (JJAS) and winter (DJF) precipitation extremes using various fine resolution multi-source gridded datasets including a gauge-based, a satellite dataset, a recently released regional reanalysis as well as a global reanalysis over WH (29°N-37.5°N and 72.5°E-80.5°E, see Figure 1a) within the time period 1979–2020. We have used India Meteorological Department’s (IMD) daily rainfall data available at 0.25° × 0.25°, interpolated from 6955 rain gauge stations throughout the Indian subcontinent [51]. However, comparatively less stations are available over WH. We have also used, Integrated MultisatellitE Retrievals (V3) for Global Precipitation Measurement (GPM-IMERG), a merged high-resolution satellite product. Precipitation estimates are produced using Day-1 IMERG algorithm through intercalibration, merging, and interpolation of microwave and infrared records from GPM satellite constellation with gauge-based observations [52]. The regional reanalysis-IMDAA, is a recently released high resolution (12 km) product over the South Asian domain, generated by National Centre for Medium Range Weather Forecasting in collaboration with UK Met Office and IMD using a unified atmospheric model and the four-dimensional variational (4D-Var) data assimilation technique [53]. The dataset provides advantages in better representation of orographic features owing to its high spatial resolution [1, 2, 54]. Finally, we have also utilized the state-of-the-art global reanalysis ERA5, developed by European Centre for Medium-Range Weather Forecasts [55] available at a resolution of 0.25° × 0.25°. Further, daily values of different meteorological variables, including air temperature, specific humidity, vorticity, three-dimensional wind components, etc. at various pressure levels from ERA5 have also been considered.

Figure 1.

(a) Elevation (in meters) of Western Himalayan region. Subplots (b) represent annual distribution of liquid (blue line) and solid (red line) precipitation and (c) indicates same as (b) in terms of precipitation fraction over WH during 1979–2020.

2.2 Methodology

2.2.1 Identification of precipitation extremes

Keeping in mind the complexity of the study region and high spatio-temporal variability of precipitation over the regime, we have identified extreme precipitation thresholds at each grid point using percentile approach. This helps in describing intensity of extreme precipitation without having a stringent threshold for such varying terrains. Extremes are considered when the daily precipitation amount from the entire time series of precipitation exceeds the 95^th percentile threshold at a particular grid point. The accumulated precipitation exceeding these thresholds have the potential to expedite floods in the downstream regions and affect the agricultural crops sown in the region during summer and winter seasons. Thus, we have chosen the 95^th percentile as the threshold for our study for identifying EPEs.

2.2.2 Trends for intensity and frequency of EPEs

WH is known to be a complex and topographically heterogeneous regime. A wide discrepancy in precipitation patterns is observed among different datasets over this region [1]. Thus, we focus on analyzing precipitation extremes in various datasets to understand how different datasets depict precipitation extremes over the region. Datasets from different sources have been selected on the basis of the availability of long-term (∼20 years) daily precipitation records, and having a relatively finer resolution. Considering that the selected datasets are generated with different input data and dissimilar developmental methods, the presence of any similar signals are strong indicators of real situation [1, 56]. Collaterally, we also explore the fidelity of the newly developed high resolution IMDAA reanalysis in representing WH precipitation extremes during the summer and winter season. IMDAA’s high resolution offers significant potential in better resolution of orography [1, 2, 54], thus, a relatively better depiction of precipitation extremes is a key possibility. IMDAA’s potential in the representation of climatological winter and summer precipitation characteristics over WH has been described thoroughly by [1, 2], respectively. The present study provides additional characterization of precipitation extremes in IMDAA.

Here, we investigate the intensity and frequency of EPEs in the respective datasets for summer and winter seasons after performing a bilinear interpolation to a common spatial resolution of 0.25° × 0.25° for a fair comparison purpose. Further, a common period 2000–2020 has also been considered for extreme precipitation intensity and frequency trend analysis to understand how EPEs are changing in the recent decades. Spatial trends at decadal scale for intensity and frequency of EPEs are calculated at each grid point from the daily time series of precipitation. The magnitude of trend to the whole seasonal extreme precipitation at individual grid points is computed using a non-parametric Mann-Kendall Test [57, 58]. This is a rank-based method and widely used in hydrometeorological data studies [59]. Temporal trends for the frequency of EPEs are the seasonal average of the total number of grid points on each day that satisfies the EPE criteria specified earlier.

2.2.3 Trends for dynamical and thermodynamic variables

In addition, the climatological trends of key atmospheric thermodynamic variables such as air temperature and specific humidity at different pressure levels and hydrometeors in the total atmospheric column are investigated as these parameters strongly influence the extreme precipitation over any region. We further estimated climatological temporal trends for different derived variables such as moist static energy (MSE), vertically integrated moisture transport (VIMT), eady growth rate (EGR) and eddy kinetic energy (EKE) over WH to explore the impact of synoptic signatures associated with summer and winter precipitation extremes during the period 1979–2020.

3. Results and discussion

3.1 Distribution of liquid and solid precipitation over WH

The study area includes highly rugged mountains and comparatively gentler foothills as well as surrounding plain regions of Punjab and Haryana (see Figure 1a). WH experiences precipitation in different seasons through different weather systems. Moreover, the effect of regional orography as well as seasonal variations is predominant in the distribution of liquid and solid precipitation over different sub-regimes of WH (e.g. [3]). The intra-annual cycle of climatological mean precipitation averaged over the WH region from ERA5 reanalysis (Figure 1b) shows dominant contributions of liquid form of precipitation (rainfall) during summer season (JJAS) whereas solid precipitation (snowfall) is observed to be the primary form of precipitation during winter months (DJF). This implies that a large fraction (about 61% or more) of the received total precipitation during the DJF months comes from frozen hydrometeors (Figure 1c). The maximum precipitation over WH is received during the month of July followed by August (Figure 1b), of which rainfall contributes almost 95% and 94% respectively (Figure 1c). During DJF months, highest amounts of total precipitation are observed during February followed by January and December, respectively. Similar results were reported for winter precipitation amounts at sub-seasonal scale by [1, 54].

Looking into the segregated precipitation fractions, December observes almost 68% of total monthly precipitation in solid form and the rest 32% as rainfall. Approximately 65% and 61% of the total monthly precipitation during January and February respectively are observed as snowfall. Although the individual fractions of monthly solid precipitation are lesser for February compared to December, it is significant to note that the total monthly precipitation observed during February is nearly more than double compared to December, implying that total snowfall amounts observed in February are higher.

3.2 Extreme precipitation intensity, frequency and trends

3.2.1 Summer season

We have examined spatiotemporal changes in the intensity and frequency of precipitation extremes over the WH using the daily gridded precipitation data (IMD), satellite-based data (IMERG), IMDAA regional reanalysis and global reanalysis ERA5. Figure 2 shows the precipitation intensity exceeding the 95^th percentile at each grid point for the summer monsoon and trends in the intensity and frequency of EPEs. The distribution of grid-wise precipitation intensity exhibits heterogeneity in precipitation amounts over the WH (Figure 2a–d). However, the Himalayan foothill belt shows the highest precipitation intensity of EPEs. All datasets show that the spatial distribution of precipitation intensity over the WH is characterized by low precipitation at higher altitudes (the northern part of the WH) and high precipitation over low-altitude regions, albeit with varying magnitudes. Satellite-based data (IMERG) and reanalysis dataset (IMDAA) overestimates the precipitation intensity as compared to the daily gridded precipitation (IMD), specifically over the Himalayan foothills. Although the datasets show heterogeneity in spatial distribution of long-term trends of extreme precipitation intensity, specific hotspots in the Himalayas show significant increasing trends (up to 12 mm day⁻¹ per decade; Figure 2e–h). In addition to the increasing intensity of EPEs, these same hotspots have been observed for the increasing frequency of EPEs in Himachal Pradesh, Uttarakhand, and Jammu and Kashmir (Figure 2i–l). However, relatively mixed trends for frequency of EPEs over different regions of WH is clearly visible. The northern part of the study region is an exception in terms of opposite grid wise trends of frequency and intensity of extremes. To have a more elaborate understanding, we have further investigated the temporal trends of precipitation extremes (both intensity and frequency) through an area-averaged time-series of EPEs at seasonal scale in different gridded datasets (Figure 2m–n). The time-series of intensity of daily precipitation extremes shows an increasing trend for all datasets in the long-term as well as recent decades (Figure 2m). In addition, all datasets agree on the long-term rise in frequency of EPEs over the region as well as during recent decades, exception being only IMD which shows negative trend in recent decades (Figure 2n).

Figure 2.

Precipitation intensity (> 95^th percentile) and trends in intensity and frequency of EPEs for the summer season. Set of plots (a-d) represents intensity climatology, (e-h) shows intensity trend (per decade), and (i-l) represents frequency trend (events per decade) of EPEs for IMD, IMERG, IMDAA, and ERA5 respectively. Subplots (m, n) represent the area averaged time series in intensity and frequency of EPEs for entire duration (1979–2020) in dotted line and recent decades (2000–2020) in solid line.

3.2.2 Winter season

The geographical distribution of multi-year seasonal winter precipitation extremes’ intensity and decadal trends in each dataset (after re-gridding) during 1979–2020 as per the duration of data availability for each dataset have been presented in Figure 3. Looking into the climatology of extreme precipitation intensity (> 95^th percentile), considerable heterogeneity in extreme precipitation amounts as well as patterns are observed among different datasets, highlighting the role played by complex regional topographical variations. However, the highest precipitation amounts are observed along the western Himalayan foothills in all datasets (Figure 3a–d). The spatial extent of precipitation is maximum with high intensity over J&K, followed by HP and UK, respectively. This is due to the fact that vigor of WDs decreases as they move from J&K along WH towards central Himalayas [60]. Daily extreme precipitation amounts are found to be reaching beyond 50 mm day⁻¹ in IMDAA. Although it becomes important to note here that, like any reanalysis product IMDAA too, has been found to overestimate regional precipitation amounts [1].

The spatial variations of long-term trends per in extreme precipitation intensity although exhibit considerable heterogeneity, however, significant (confidence level = 0.95) increasing trends (up to 3.5 mm day⁻¹ per decade) over some parts of HP, UK, JK and Ladakh can be observed (Figure 3e–h). Moreover, the trends in IMERG are very intense and highly significant, implying the strengthening of extreme precipitation intensities in the recent decades. Along with extreme precipitation intensity, the frequency of occurrence of such extreme events at different grid locations also seems to be on the rise over some parts of JK, northern part of HP and some parts of Punjab (Figure 3i–l). Again, IMERG shows highly significant rising trends of extreme precipitation intensity over most regions of WH.

Figure 3.

Precipitation intensity (> 95^th percentile) and trends in intensity and frequency of EPEs for the winter season. Set of plots (a-d) represents intensity climatology, (e-h) shows intensity trend (per decade), and (i-l) represents frequency trend (events per decade) of EPEs for IMD, IMERG, IMDAA, and ERA5 respectively. Subplots (m, n) represent the area averaged time series in intensity and frequency of EPEs for entire duration (1979–2020) in dotted line and recent decades (2000–2020) in solid line.

Further, we have investigated the trends for area-averaged precipitation extremes at seasonal scale in different re-gridded datasets over the entire study domain (Figure 3m–n). All datasets including IMDAA clearly indicate that precipitation extremes are not only becoming frequent but are also intensifying over time, exception being IMD which shows a decreasing trend. Although it is important to note that none of the trends pass the significance test.

IMDAA and ERA5 show that precipitation intensity of extremes has increased by almost 16% and 23% respectively from 1979 to 2020 (Figure 3m). The magnitude of increase in intensity observed in IMERG from 2000 to 2020 is about 23.5%. In terms of rise in frequency of EPEs, IMDAA shows an increase of about 14 events (grid wise) per day seasonally over the domain since 1979 (Figure 3n). ERA5 agrees well on the enhancing frequency too but shows a slightly lesser increase of about 8 events per day. However, the rise is much sharper in the recent decades observed in IMERG satellite data which shows an increase over approximately 32 events per day in the region. The findings highlight the fact that extreme precipitation conditions are strengthening recently. The obtained results are in compliance with individual station-based trends over WH reported in some studies [40, 41] as well as other studies based on gridded observations and satellite datasets [14, 33], thus indicating a rise in precipitation extremes. Considering that the precipitation extremes are on a rise over the region, it becomes critical to understand the possible causes for these enhancements through trends for various dynamics and regional atmospheric conditions contributing to rise in precipitation extremes over WH.

3.3 Trends for different dynamic and thermodynamic controls related to EPEs

Any changes in weather and climate extremes are generally related to local exchanges in heat, moisture, and other thermodynamic quantities as well as dynamic changes. Although dynamic and thermodynamic processes in the atmosphere are interlinked, it is important to separately investigate their roles for variations in climate extremes (e.g. [61]). The thermodynamic controls of precipitation extremes are associated with an enhancement in the atmospheric moisture content, the most basic assumption being that precipitation extremes portray a tendency to rise in a warming climate, as per Clausius–Clapeyron relationship [62, 63]. Several studies propose the direct link for amplification of extremes over the WH region with increasing temperatures as well as atmospheric moisture content during both summer and winter seasons (e.g. [14, 64, 65]). Changes in the thermodynamic signatures significantly contribute to variations in precipitation patterns.

3.3.1 Summer season

In order to understand the variability of precipitation and the underlying factors might be contributing to the rise of extreme precipitation over the WH during summer monsoon season, seasonal trends of various atmospheric parameters, including hydrometeors, have been investigated. Interannual variations of tropospheric air temperatures at different levels and their trends for long-term as well as recent decades are shown in Figure 4a–b. The results show that the upper level (200 hPa) and mid-tropospheric (500 hPa) temperatures are significantly increasing, and in recent decades this increase is much sharper, which indicates the warming in the upper levels of troposphere over the WH (Figure 4a).

Figure 4.

Time series of the temperature (a-b) and specific humidity (c-d). Total column variables (e) snow water, (f) ice water, (g) liquid water, and (h) rainwater. Subplot (i) represents vertical velocity at 500 hPa, and (j) represents moisture flux convergence, (k) represents moist static energy, and (l) represents vertically integrated moisture transport using ERA5 reanalysis data during 1979–2020 for summer (JJAS) season. Dotted line represents the trend in entire duration and solid line indicate recent decades.

However, the seasonal trends at the lower level (850 hPa) and near-surface (1000 hPa) show a relative cooling trend in the long term (1979–2020), whereas the recent decades show warming at these near surface levels (Figure 4b). It is well known that global warming and related changes in the atmosphere above the WH generate EPEs, flash floods, cloud bursts, river flooding, landslides etc. [23, 66]. The overall result suggests that the increased precipitation intensity and frequency over the WH are directly associated with warming. Additionally, we have investigated the climatology of specific humidity at upper-level and low levels over the study region (Figure 4c–d). It has been observed that the specific humidity shows an increasing trend during the summer monsoon even though it is not significant at upper level but mid-tropospheric and lower-tropospheric specific humidity exhibits significantly increasing trends that is indicative of possibility of increased evaporation which can consequently contribute to increased precipitation, thus supporting the rise in precipitation intensity and frequency trends over the WH.

The interannual variability of cloud hydrometeors, total column of snow water, total column ice water, total column liquid water, and total column rain water in the long-term (1979–2020; dotted pink line) and recent decades (2000–2020; black line) are shown in Figure 4e–h. Several studies have found that changes in cloud microphysical properties can have an effect on the simulated mesoscale dynamics of extreme events (e.g. [31, 67]). All the hydrometeors exhibit increasing yet insignificant long-term trends except for total column liquid water, which shows a significantly increasing trend (Figure 4g). It is worth noting that all four hydrometeors reveal comparatively sharper increasing trends in recent decades when compared to the entire study period. Studies report that the presence of atmospheric aerosols potentially assists an alteration of these cloud properties leading to more precipitation under favorable atmospheric conditions (e.g. [68]). This could be a possible explanation for the role of natural as well as anthropogenic forcings to increasing levels of precipitation extremes. Although a clear justification of these possibilities is beyond the scope of this study.

We further investigate extreme precipitation events by examining the underlying changes in other related dynamical and thermodynamic parameters whose characteristics provide crucial information about atmospheric conditions, which is important in the case of extreme precipitation events. Generally, the variations in dynamical components are caused by changes in vertical motion, whereas variations in atmospheric water vapor lead to changes in thermodynamic components [69]. Figure 4i displays a decreasing trend of 500 hPa vertical velocity (Pa s⁻¹) over the study domain, indicating a rise in convection over the WH which can favor cloud formation. This rising motion causes supersaturation, which is the primary cause of cloud droplet nucleation, condensation of water vapor into liquid water droplets and eventually precipitation [70, 71]. Therefore, increasing strength of vertical velocity over time is closely related and has far-reaching implications for vertical water, mass transport, and extreme precipitation.

Our study further investigates the roles of variability in atmospheric moisture transport over the topographic regimes of WH in case they show any contribution towards the rise of summer extreme precipitation. To accomplish this, we have investigated the trends for vertically integrated moisture flux convergence (VIMFC) and vertically integrated moisture transport (VIMT) over WH during the ISM which are given by,

$$\mathit{\text{VIMFC}}=-\frac{1}{g}\underset{\mathit{\text{psurf}}}{\overset{\mathit{\text{ptop}}}{\int }}\left(\frac{\mathit{duq}}{\mathit{dx}}+\frac{\mathit{dvq}}{\mathit{dy}}\right)\mathit{dP}$$E1

$$\mathit{\text{VIMT}}=\frac{1}{g}\underset{300\mathit{hPa}}{\overset{1000\mathit{hPa}}{\int }}\mathit{qV}\mathit{dP}$$E2

where, $q$ is specific humidity, $V$ is the horizontal velocity, and $\mathit{dP}$ denotes the vertical incremental change in pressure.

Observed increasing seasonal trends in VIMFC (Figure 4j) directly characterize the behavior of EPEs and provide a favorable condition for increasing trends of EPEs in the WH. The region receives moisture directly from the Arabian Sea through south-westerlies and Bay of Bengal from north-easterlies [11] during southwest monsoon. Increasing levels of VIMT observed in Figure 4l explains that more moisture is getting transported to the region over the time. Further, our study notes that VIMT shows a significantly increasing and comparatively sharper trend over the WH during the recent decades, which constitutes enhanced seasonal moisture transport and, thus more precipitation. MSE is the one of the most important thermodynamic parameters, defined as the total sum of an air parcel’s internal and gravitational potential energy.

$$\mathit{MSE}=C_{P}T+\mathit{gz}+L_{v}q$$E3

where, $C_P$ is the specific heat capacity at constant pressure, $T$ is the air temperature, $g$ is the gravitational acceleration, $z$ is the geopotential height, $L_v$ is the latent heat of vaporization, and $q$ is the specific humidity. The MSE at 500 hPa shows a significantly increasing trend, which reveals a higher atmospheric instability during the summer monsoon season Figure 4k. When the MSE is imported from the surrounding environment, it destabilizes the atmosphere by heating and humidifying it, resulting in deep convective precipitation [72, 73].

3.3.2 Winter season

The examination of area-averaged wintertime trends for mean air temperatures and specific humidity at different tropospheric levels (200 hPa, 500 hPa, 850 hPa and 1000 hPa) over the study region in ERA5 reanalysis are presented in Figure 5a–d. The trends indicate an increase in the air temperatures at all considered levels in the troposphere indicating that both lower and upper atmospheric temperatures are on a rise in western Himalayas, a clear indication of global warming effects. At the same time, atmospheric water vapor concentrations seem to be on a significant rise specifically in the upper and lower tropospheric levels. Conclusively, we can infer that changes observed in these regional thermodynamic variables are crucial and contributing to rising trends for precipitation extremes over the region. However, no discernible trend in middle troposphere vertical velocity has been observed over the region.

Figure 5.

Time series of the temperature (a-d), specific humidity (e-h), frequency of vortices (i), vertical velocity (500 hPa, j), eddy kinetic energy (k), brunt–Väisälä frequency (l), meridional moisture flux convergence (m), vertically integrated moisture transport (n), eady growth rate (o), and vertical shear (p) during 1979–2020 for winter (DJF) season in ERA5 reanalysis.

Further, we aimed to understand if there is any role of dynamical signatures of the atmosphere in influencing the rise in precipitation extremes over WH. We have focused on the variability associated with transient activity of westerly troughs (vortices) over WH. Our analysis considers total counts of cyclonic (relative) vorticity in each winter season to measure the activity of troughs over WH. The occurrences of vorticity (500 hPa) exceeding 1.5 standard deviation at individual grid points over WH for each winter season has been counted and the trends have been observed (Figure 5i). An increase in the frequency count of vortices over the region is found, which indicates that the formation of these troughs is becoming more frequent in the recent decades. Such conditions can lead to development of deep convection in the presence of enough moisture and thus can create favorable conditions for heavy precipitation. Moreover, moisture flux convergence plays an important role in inducing heavy precipitation through the deepening of such vortices, thus, we also investigated the trends for vertically integrated meridional moisture flux convergence over the region. Figure 5m shows the time-series for moisture convergence associated with meridional $\left(-\frac{\partial \mathit{vq}}{\partial y}\right)$ winds, area-weighted over WH.

The results reveal a slight increasing trend, though insignificant, in moisture flux convergence associated with meridional winds. Moisture transport from the Arabian Sea into WH during winter has been designated as a crucial moisture input source during extremes (see [34]). As our findings suggest an enhancement in both transient activity of westerly troughs and moisture flux convergence, it can be concluded that variability in dynamic responses of the atmosphere in conjunction with thermodynamic variations might be a major contributing factor for increasing trends of extreme precipitation over the region. Mediterranean, Caspian, Red and Arabian seas are primary contributors for eastward moisture advection towards WH leading to moisture availability for precipitation [74, 75]. Here, we have tried to investigate the trends for winter seasonal moisture supply over WH (Figure 5n) through time series for VIMT which reveals a clear increasing trend in the moisture transport over WH through the years. This implies that with more moisture available over the region, conducive conditions for development and sustenance of heavy precipitation events can be created.

WDs, termed as immature baroclinic waves [76], develop and intensify primarily through atmospheric baroclinic instability, known to be generated by the meridional gradient of temperature and vertical shear of the background subtropical westerly flow [14]. Strong upper-atmospheric baroclinicity is generally observed in the locality of the subtropical jet during winter season [77]. The baroclinic processes partially influence the vertical velocities in the region majorly through a coupling between the background westerly flow (jet), WDs, and the orography and, thus leads to precipitation over WH [56]. The baroclinic instability in the atmosphere can be measured through maximum Eady growth rate (EGR) which follows the maximum growth rates to configure the Eady problem [78, 79] and is given as:

$${\sigma }_E=0.3098\frac{\left|f\right|\left|\frac{\partial U\left(z\right)}{\partial z}\right|}{N}$$E4

where, $f$ is the Coriolis parameter, $U\left(z\right)$ is the vertical profile for zonal component of wind, $z$ is the vertical coordinate and, $N$ refers to the Brunt–Väisälä/buoyancy frequency defined by,

$$N^2=\frac{g}{\theta }\frac{\partial \theta }{\partial z}$$E5

where, $g$ is the acceleration due to gravity, and $\theta$ is the potential temperature. The buoyancy frequency represents atmospheric static stability. We have calculated EGR between two levels (200 minus 850 hPa) and further looked into its long-term trends (Figure 5o).

It is evident that the trends for EGR are increasing with high statistical significance. This highlights the fact that baroclinic instability is on a rise in the atmosphere over WH. A more baroclinically unstable atmosphere favors an intensification of WDs and can lead to heavy precipitation [13, 14, 41, 76, 80] have reported enhanced baroclinicity over WH during the recent decades and suggested the potential role of zonally asymmetric changes in the wintertime circulation caused by the elevation dependent climate warming signal. Further, we tried to investigate whether this enhancement in the regional baroclinicity, and consequently EPEs, is contributed through changes in the vertical shear of zonal wind (Figure 5p) or static stability (buoyancy frequency; Figure 5l). The results revealed that there has been an increase in the vertical shear over the period in WH and at the same time static stability of the region is observing a decreasing trend, meaning a more unstable atmosphere to intensify WDs. Further, it is clear that changes in both vertical shear and static stability are responsible for enhancing the regional baroclinicity and thus increase precipitation extremes over the region.

Various studies suggest that higher kinetic energy in the atmosphere in response to jet helps in the growth and intensification of WDs due to their baroclinic nature [13, 81]. The dynamical variations over WH are largely characterized by high frequency transient eddies in the atmosphere, which result from the conversion of the available potential energy into kinetic energy through baroclinic instability [82, 83, 84]. Therefore, we have investigated the trends for upper tropospheric (200 hPa) eddy kinetic energy (EKE) in ERA5 to understand the role of localized impacts of westerly flow and consequent energy exchange processes in the atmosphere for fueling the rise of precipitation extremes. EKE is defined as the kinetic energy associated with the time-varying component of the horizontal velocity field.

$$\mathit{EKE}=\frac{1}{2}\left(u^{\prime 2}+v^{\prime 2}\right)$$E6

$$u=\overline{u}+u^{\prime }$$E7

$$v=\overline{v}+v^{\prime }$$E8

here, $u$ and $v$ are horizontal velocity components, $u^{\prime }$ and $v^{\prime }$ denote time-varying velocity components whereas, $\overline{u}$ and $\overline{v}$ represent the time mean velocity components. A bandpass filter of 2–10 days has been applied to the anomalies of horizontal velocity fields to filter out the mesoscale transient eddies and their trends have been looked into. The findings reveal that EKE over the region has been increasing significantly (Figure 5k) which provides preferable conditions for intensifying WDs over WH and thus can contribute to heavy precipitation over the region.

4. Summary and conclusion

The projected rise in precipitation extremes under the warming climate is a key matter of concern. This study investigates the spatiotemporal variations and trends for intensity and frequency of precipitation extremes over the western Himalayan region during summer and winter monsoon seasons. Furthermore, the potential dynamical and thermodynamic controls of precipitation have been explored to understand their possible roles in the increase of such extreme precipitation events over the region. Since the drivers and contributing atmospheric factors for summer and winter monsoons over WH are completely different, an effort has been made to explain the possible causes for increasing extremes for individual seasons separately. Based on the key findings from our study, the following conclusions can be drawn.

• WH receives precipitation mostly in liquid form during summer monsoon whereas solid precipitation is found to be the dominant form of precipitation in winter season.

• The trends for intensity and frequency of precipitation extremes over the region indicates that extremes are intensifying and becoming more frequent over the years in the western Himalayas during both seasons, specifically during recent decades. Although the trends show a somewhat heterogeneous and mixed pattern, intriguingly several pockets in Himachal Pradesh, Uttarakhand, and Jammu & Kashmir are exhibiting an increase in the frequency as well as the intensity.

• This increase seems to be directly associated with rising trends for air temperatures and atmospheric water content during both seasons, which signifies the role played by warming climate, providing the feedback of moisture for evaporation and enhanced cloud formation.

• During the summer season, increasing trends for most of the total column cloud hydrometeors, with significance in case of liquid water signals to the influence of cloud microphysical properties in the rise of EPEs.

• Moreover, the increase in the strength of vertical velocity over time during summer monsoon combined with an upward trend in 500 hPa MSE can make the overall atmosphere unstable and prone to more precipitation. What is more important to note is that this instability is further extended in the moisture transport where we observe increasing trends in VIMT and the VIMFC. This ever-growing availability of moisture can have serious impacts on the frequency and the intensity of EPEs over the Himalayas specifically along the Himalayan foothills.

• The possible dynamics related to increase of EPEs in winter season relates to a rise in formation of transient westerly troughs/vortices over the study region combined with an enhancement of regional moisture flux convergence associated with meridional winds as well as moisture transport. This explains the contribution of increased moisture supply possibly from Arabian sea in intensifying the vortices and deepening the convection.

• The enhancing baroclinicity and zonal vertical shear and reduced levels of static stability in the atmosphere is highly supportive of rising EPEs in the WH. Additionally, the winter atmosphere over WH is observing an increase in the amount of kinetic energy over the years which implies the availability of higher energy to feed and intensify WDs over the region.

• Lastly, the newly developed high-resolution reanalysis, IMDAA, is realistic in representing the regional extreme precipitation distribution and trends for intensity and frequency, in comparison with other utilized datasets.

In summary, we would like to acknowledge that the substantial variations and the spatial heterogeneity in magnitude exist between different data products over WH mainly due to lack of in-situ data and complex orography. Such uncertainties can be reduced by increasing the coverage of ground-based in-situ observational networks over the region. The climate warming signal with its positive feedback mechanism in terms of enhancing the available moisture content in the atmosphere is a crucial factor influencing the rise of precipitation extremes over WH. Significant variations of various dynamics supported by local thermodynamic processes have also been observed, helping the growth and intensification of EPEs. However, future work is required to understand the underlying physical processes and their interactions with regional orography during such events to have a more elaborate understanding of the feeding mechanisms.

Acknowledgments

This research work was supported by the Science and Engineering Research Board, Department of Science and Technology, Government of India under the “Start-up Research Grant (SRG) scheme” (Grant SRG/2020/001857) and Prime Minister’s Research Fellowship (PMRF), Ministry of Education, Government of India. Authors Nischal and Rohtash gratefully acknowledge the financial assistance from PMRF.

Conflict of interest

The authors declare no conflict of interest.

Nomenclature

WH	Western Himalayas
ISM	Indian summer monsoon
EPEs	Extreme precipitation events
WD	Western disturbances
ARs	Atmospheric rivers
IPCC	Intergovernmental Panel on Climate Change
IMDAA	Indian Monsoon Data Assimilation and Analysis
IMD	India Meteorological Department
GPM-IMERG	Integrated MultisatellitE Retrievals for Global Precipitation Measurement
MSE	Moist static energy
EGR	Eady growth rate
EKE	Eddy kinetic energy
J&K	Jammu and Kashmir
HP	Himachal Pradesh
UK	Uttarakhand
VIMFC	Vertically integrated moisture flux convergence
VIMT	Vertically integrated moisture transport