Analysing climatic variability and extremes events in the Himalayan regions focusing on mountainous urban agglomerations

Abstract The present study highlights the long-term variations in temperature and precipitation using time series datasets of CRU and ERA5-Land across the Himalayan region. The Mann-Kendall and Sen’s slope-based trend analysis exhibited an apparent warming trend in the region, with higher increase in Tmin (1.5 °C) and comparatively lower increase in Tmax (0.8 °C) during 1901–2018. The joint CCl/WCRP/JCOMM based 19 extreme indices elucidated an increasing frequency of warm days (16%) in Central Himalayan urban agglomerations (HUAs) with a declining number of cold days (−6%) and cold nights (−4%) during 2000–2019. The moderate decline in the wet days (8%) and consecutive wet days (−5 days/20 years) observed in all HUAs compared to an increasing frequency of the consecutive dry days (3 days/20 years). Study reported higher warming in Kathmandu UA, while increased precipitation in Srinagar UA, and contributes to framing climate change adaptation and mitigation strategies in the mountainous system.


Introduction
Natural ecosystem of the earth and its sustainability are being threatened by an enormous rise in human population, rapid urbanisation (Kumari et al. 2022) and economic development (Seto et al. 2013). The substantial rise in built-up land was primarily evident by replacing green cover and arable land that induced significant alterations in temperature (Diksha 2022;Gattia et al. 2022). It accelerated the anthropogenic contribution in global climate change (Pingale et al. 2014) as observed in the form of significant variability in annual and seasonal patterns of temperature and precipitation that combinedly affected the natural processes, life forms and environment to a large extent (Seneviratne et al. 2012). Significant warming has been observed due to land use alteration at Pan India Scale . Precipitation and temperature are frequently being used as major parameters in the discipline of climatology and hydrological studies to determine the magnitude and variability of climate change and its impact (IPCC 2007). The rise in temperature (by 0.8 C since 1880) influenced a varied range of phenomenon such as melting of blue towers (13.4% decade-1 reduction in Arctic), glaciers (287.0-billion-ton ice year-1), together with sea level rise (3 to 4 numbers of precipitation events year-1), and increased frequency of extreme events (Bhatla et al. 2019). The earth's average temperature is rising gradually and is expected to rise more due to the increasing concentrations of heat-trapping greenhouse gases (USGCRP 2017). The rise in average temperatures at the Earth's surface induces more evaporation that leads to increased precipitation (Konapala et al. 2020). Thus, in many regions a warming condition is attributed to extreme climatic events . Also, the variability in trends and pattern of precipitation would greatly influence hydro-meteorological extremes viz., flood events and droughts . However, climate change causes shifts in wind patterns and oceanic currents that influence the world's climate system, leading to more warming, cooling conditions at varied locations, and erratic precipitation patterns across the globe. Furthermore, the effects of climate change are exacerbated by several severe socio-environmental stresses causing risk and heading to conflicting perceptions (Ives 2004).
Mountainous areas are comparatively highly vulnerable to climate change than other physiographic regions having equal latitudinal extents. They are regarded as sensitive markers to climate change indicators (Barry 1990;Beniston 2016). The Hindukush Himalayas (HKH) have observed serious melting and permafrost degradation over the past few decades Yao et al. 2012bYao et al. , 2012a, resulting in substantial changes in the hydrological systems in the Himalayan Rivers and biodiversity hotspots (Shrestha et al. 1999;Immerzeel and Bierkens 2012;Yang et al. 2014). Climate change is influencing the temperatures and amount of snow and ice in the Himalayas, as well as precipitation patterns in Asia's densely populated downstream regions, all of which have significant implications on the people's livelihoods and well-being of the region (Xu et al. 2009;. Therefore, it is necessary to develop and adapt effective management policies and mitigation strategies for better interpretation and forecasting effects of climate change on the ecosystems of the Himalayas (Tewari et al. 2017). The Himalayan region is critically data-deficient and has logistic difficulties in terms of observational datasets of climate change impact studies (IPCC 2007), which may be to higher elevations and rugged topography (Yadav et al. 2004). This led to limited systematic studies and empirical observations of climate change in the Himalayas (Gautam et al. 2013).
Long-term climatic data records are essential inputs for climate impact assessments and hydrological research (Cook et al. 2004;Lobell et al. 2011). Previous studies reported a decreasing summer precipitation in the Himalayan and the northwest Himalayas during the last six decades. The trends of winter precipitation, however, varied according to the datasets used (Climatic Research Unit data and Global Precipitation Climatology Centre), with each of the values surpassing the test of significance (Palazzi et al. 2013). Time-series long-term analysis of gridded temperature data from various open sources are useful and being widely used at spatio-temporal scales, improving the gaps in observational datasets (Cowtan and Way 2014). Climate data products from the Hadley Centre and Climatic Research Unit is the one of the most referred data, compiled using station data from various sources barring the Antarctic region (Harris et al. 2014). CRU data has been preferred over other climatic data viz., NDC (National Data Centre) or IMD (India Meteorological Department) attributed to its property of higher resolution (Rao et al. 2014). Long-term variations in surface air temperature for the entire world (Hansen and Lebedeff 1987) and the hemispheres have demonstrated an increasing trend in recent decades (Jones et al. 1986). Long-term temperature trend analysis is essential to study climate change and its impacts. Most of the observational and meteorological station data show a deficit of spatio-temporal continuity in the region of interest for a greater period (Van Wart et al. 2015). However, such measured datasets are often limited and restricted access (Hughes et al. 2009). Global or regional extreme climatic events variations induced by climate change in the recent decades are drawing more attention (Easterling et al. 2000;Frich et al. 2002). The spatio-temporal climate change and modelling studies and its observed changes suggested warming conditions have not been consistent (McAvaney et al. 2001). The changing nature of daily temperature and precipitation patterns is one of the most important attributes of climate change and especially, changes in the extremes of temperature and precipitation distribution (Liu et al. 2006;Kumar et al. 2020). Global climate change can clearly be observed through the characteristics of the extreme climatic events as these events are more sensitive to climatic changes than their mean values (Katz and Brown 1992). Extreme climatic events are rare occurrences that are related with the high/low extreme events of the meteorological variables. Evaluation of extreme climatic events are essential to anticipate the impact of changing climate and its variability (Sun et al. 2016). During the 21st century, some of these extreme climatic events attributed to anthropogenic activities will become more frequent and severe, causing widespread damage globally Kumar et al. 2020). Apparently, with variation in temperature extremes scale with changes in the mean temperature (simple alterations of the probability distribution), it is hypothesised that with respect to the climate of 1960-1990, warm extremes would increase and cold extremes would decrease (Seneviratne et al. 2012).
Consequently, there is an increasing demand for data on climate extremes which is critical for long-term risk management (Dubey et al. 2021). To assess the suitability and efficacy of climate change adapting strategies to enhance human well-being, a detailed assessment of climate change (mainly, temperature and precipitation variability) is necessary for evidence-based decision-making (IPCC 2019). Henceforth, to detect the climate change variability that occurred over the past decades, quantification is necessary which will be further aid to make predictions and for better preparedness. Global climate change should be identified at its initial stage by evaluating tendencies in climate variability and phenomenon of extreme events especially in mountainous regions (Messerli and Ives 1997). In the present study, the long-term variations and patterns of temperature and precipitation were examined using new climate datasets to develop a better understanding of climate change variability in the Himalayan region. The study also attempts to analyse the extreme climatic events and spatio-temporal climatic variability in the Himalayan region mainly focusing on the major Himalayan urban agglomerations (HUAs), which have significant implications on the people's livelihoods and well-being in the region. The study also focuses on providing more information related to regional climate change impacts on ecological and social systems for modelling and analysis.

Study area
The study area mainly covers the Indian Himalayan region, stretching from Jammu & Kashmir (J&K) in the west to Arunachal Pradesh in the east have a continuously stretch for approx. 2,500 km, with special emphasis on the major seven HUAs (Srinagar, Kathmandu, Dehradun, Shimla, Itanagar, Thimphu, Gangtok) of the region located in India, Nepal, and Bhutan ( Figure 1). These HUAs significantly contribute towards the socio-economic development of the region. The Himalayas, as a key climatic boundary impacting extensive systems of air and water circulation, contribute in influencing weather conditions in the Indian region and Tibet (Diksha 2017). The altitude and strategic location of the Himalayas prevents the passage of the cold winds coming towards the Indian subcontinent from the north thus providing a more moderate climate to India and largely influences the precipitation pattern. The Himalayas usually observes rainfall from June to September during the southwest monsoon season, however it decreases from east to west (Singh and Mal 2014). The combined effect of precipitation, latitude and altitude largely influences the forest belts in the Himalayan region. The region is home to one-tenth of the world's known higher-altitude plant and animal species, and half of India's native plant species (Padma 2014).

Data used and methodology
The present study utilized Climate Research Unit (CRU TS4.00) (http://data.ceda.ac.uk// badc/cru/data/cru_ts/cru_ts_4.00/data/pre/) to analyse the annual precipitation pattern and temperature for 118 years , while European Centre for Medium-Range Weather Forecasts (ECMWF) land component of the fifth generation of European ReAnalysis (ERA-5 land; 1980-2020) time series dataset were used to analyse extreme events. ERA5-Land is an integral and operational component of the Copernicus Climate Change Service (C3S), which produces a total of 50 variables describing the water and energy cycles over land, globally, hourly, and matching the ECMWF triangular-cubic-octahedral (TCo1279) operational grid (Malardel et al. 2016). ERA5-Land does not assimilate observations directly, because the observations is influence by atmospheric forcing. Forcings such as air temperature, humidity, and pressure are corrected using a daily lapse rate derived from ERA5. Later, the land surface model is integrated in 24 h cycles provides the evolution of the land surface state and associated water and energy fluxes. In addition to the hourly data, monthly means products are also derived (Muñoz-Sabater 2019). The temporal resolution of ERA5 is hourly and the horizontal resolution is 9 km (for both temperature and precipitation), making it unique and suitable for a large number of land surface applications. The quality of ERA5-Land fields was evaluated by direct comparison to a large number of in situ observations collected mainly for the period 2001-2018, as well as by comparison to additional model or satellite-based global reference datasets and reported its high level of matching/correlation barring minor exception (Muñoz-Sabater et al. 2021). CRU TS 4.1 with a high spatial resolution of 0.5 , contains month-wise gridded values of temperature and precipitation over land surface on the basis of daily values (Mitchell and Jones 2005;New et al. 1999;Shi et al. 2017). Evaluation studies have shown the reliability of CRU temperature and precipitation products both globally and regionally for both in the long and short-terms (Harris et al. 2014;Zhao and Fu 2006). Moreover, it has been used in previous studies for trend analysis, characterising the spatio-temporal distribution of temperature, precipitation and assessing their impact on other climatic variables (de Barros Soares et al. 2017;D'Orgeval et al. 2006;Peng et al. 2020).
In the study, various 19 extreme climatic indices (Table 1) were used to deduce the changing climatic (temperature and precipitation) patterns in the Himalayan region as well as in the proximity to major urban agglomerations using ERA-5 Land for the period 2000-2019 with base reference period 1981-2000. The major extreme climate indices were adopted from the joint CCl/WCRP/JCOMM Expert Team on Climate Change Detection and Indices (ETCCDI) (Zhang et al. 2011). The details of major extreme climate indices including 10 temperature indices and 9 precipitation indices were listed in Table 1.
The extreme temperature cold indices, referred to as cold days and cold nights, respectively, are the percentage of days when the maximum temperature (TX10p) or minimum temperature (TN10p) is below its 10th percentile during the observation period. The extreme temperature warm indices include warm days and warm nights, respectively, are the percentage of days when the maximum temperature (TX90p) or minimum temperature (TN90p) is below its 90th percentile during the observation period. The annual maximum and minimum values of daily minimum temperatures are referred as TNn (minimum value of daily Tmin) and TXn (minimum value of daily Tmax), and annual maximum and minimum values of daily maximum temperatures are referred as TNx (maximum value of daily Tmin) and TXx (maximum value of daily Tmax). The severe precipitation indices are classified according to precipitation amount (mm), such as the maximum 1-day precipitation in a year (Rx1day), wet days (R90p), very wet days (R95p), extremely wet days (R99p), and simple daily intensity index (SDII). The extreme indices can also be referred in terms of occurrence/frequency (days), i.e. the no. of days surpassing certain threshold (depth) of precipitation (no. of heavy precipitation days ! 10 mm (R10 mm), no. of very heavy precipitation days ! (R20 mm), consecutive dry days (CDD), and consecutive wet days (CWD).
The trend analysis of annual precipitation, annual temperature, annual mean temperature (Tmean), annual maximum temperature (Tmax) and annual minimum temperature (Tmin) were calculated for 118 years. The trend analysis of any time series dataset involves estimating the magnitude of the trend and its statistical significance (Jain and Kumar 2012). The statistical significance of trend was determined by Mann-Kendall (MK) test and the magnitude of trend was determined by Sen's slope approach for 118 years  in the Himalayan region. Both MK test and Sen's slope are non-parametric statistical techniques, used to detect trend and its magnitude in time series data (Pohlert 2016). The MK trend test is based on two hypotheses: (a) null (H0) which refers to no trend and (b) alternate (H1) which indicates significant increasing or decreasing trend (Ma et al. 2014). The no-trend points have been rejected. The MK test statistics 'S' can be equated as: The trend test is applied to X i data values (i ¼ 1, 2, … n À 1) and Xj (j ¼ i þ 1, 2, … n). The data values of each Xi is used as a reference point to compare with the data values of X j which is given as: where X i and X j are the values in period i and j. When the number of data series greater than or equal to ten (n ! 10), MK test is then characterised by normal distribution with the mean E(S) ¼ 0 and variance Var(S) is equated as (Ma et al. 2014): where m is the number of the tied groups in the time series, and t k is the number of ties in the kth tied group.
The test statistics Z is as follows: If Z > zero, it represents an increasing trend while a decreasing trend is represented if Z < zero.
The Sen's slope estimator (b) is represented as: The temperature and precipitation data were standardised to bring uniformity at varying altitudes by subtracting the mean and dividing it by their standard deviation (Pant and Kumar 1997). The standardised anomaly of mean annual precipitation and mean annual temperature was calculated using long-term observations (1901-2018) of precipitation pattern and temperature using CRU dataset for 118 years. An anomaly refers to the deviation from the mean at a given time series (Jain and Kumar 2012). It provides a better interpretation of trends in the data series (Singh et al. 2013). The difference between the average value (x) and the data value (xi) at the i th time period in the data series can be used to calculate an anomaly (x'i) as data normalisation is essential for computing two datasets. The standardised anomaly (z) is computed by dividing the anomaly by its respective standard deviation S x . The standardised anomaly removes locational influences which are induced from a sample data. Excursions from the mean in different batches of data are treated equally when they are divided by the standard deviation (Wilks 2011).
Here, z is standardised anomaly; x i À x is the anomaly (observed-long term mean); S x is the standard deviation.
This indicator displays annual anomalies, or deviations, from the average value of temperature and precipitation (e.g. an anomaly of þ2.0 C refers that the mean temperature was exceeding the long-term mean by 2 C). Daily temperature records in C were utilised to compute monthly anomalies, which were then averaged to get annual temperature anomalies for each of the years. Similarly, an anomaly for minimum and maximum annual temperature was calculated for each of the years. Anomalies for precipitation were calculated from total annual precipitation in mm.

Results and discussions
4.1. Analysing patterns of temperature in the Himalayan region  The CRU based long-term temperature pattern of Himalayan region was analysed for the 118 years  with reference to annual mean temperature (Tmean), maximum temperature (Tmax) and minimum temperature (Tmin) (Figure 2a). The study exhibited a latitudinal pattern of long-term mean temperature with high ($22 C) to low (2 C) from moving from south to north directions across the Himalayan region ( Figure 3a). This may be attributed to parallel geological formations of ridges and valleys along Himadri, Himachal and Shivalik Himalayas (Patel and Kumar 2003), which governs the mean temperature variation due to altitudinal differences. The considerable change in  temperature was observed in a narrower region of Himalayas, which is located between the significantly lower temperature (<1 C) in the north (Tibet plateau) and higher temperature (>25 C) in the south (Indo-Gangetic-Brahmaputra Plains). The previous studies reported an increase in annual temperatures and significant warming during the winter months primarily in western Himalayas (Mann et al. 1999;Kothawale and Rupa Kumar 2005). A high anomalous increase in annual Tmin (1 to 3.5) was recorded in the Himalayan region compared to annual Tmax (1.5 to 3.5) during 2001-2018.
The Sen's Slope based trend analysis of long-term annual temperature exhibited a moderate increase (0.006 C/year to 0.008 C/year) in temperature over a larger part of central, eastern and western in Himalayan region followed by a moderately high increase (0.008 C/year to 0.01 C/year), and moderately low increase (0.002 C/year to 0.004 C/ year) (with 99% of significance level) in remaining parts of central and northern Himalayan region over the past 118 years   (Figure 2b). However, a previous study at pan India scale reported maximum temperature ($>25 C) in south-eastern and northern India during 2002-2015, in contrast to the eastern and western Himalayan region having the lowest temperature ($4 C). Apparently, the north-eastern Himalayan region ($0.5 C) and eastern Jammu and Kashmir (J&K; $0.8 C) exhibited a decrease in temperature trends, whereas the northern parts of J&K ($0.2 C) observed an increase in temperature over the last two decades .
The annual mean (Tmean), maximum (Tmax), and minimum temperatures (Tmin) observed a statistically significant increasing trend with 99% of confidence (Figure 2c-e). A high rise in Tmin (>1.3 C/year) was observed in western Himalayas, while a moderate rise in Tmin (0.5 to 1.3 C/year) was recorded in the remaining Himalayan region. Similar observations indicated an rising trend in temperature in the entire HKH region during 1901-2014 (Mann et al. 1999;Ren et al. 2017), with a mean temperature rate of 0.104 C/decade (Bhutiyani et al. 2007), mean maximum temperature rate of 0.077 C/decade, and mean minimum temperature rate of 0.176 C/decade (Shrestha et al. 1999;Verma and Ghosh 2019). Tmin at varied topographic zones showed a relative increase of 0.02 C/year with the maximum rate of increase over the higher elevation zones (0.02 C/ year) during the period of 1980-2014. Also, the Tmax was observed comparatively lower in the higher mountainous region (0.02 C/year) as compared to other geomorphic features viz., basin (0.05 C/year), flood plains (0.04 C/year) or foothills (0.02 C/year) in the north-western Himalayan region (Shafiq et al. 2019). The large expanses of snow and glaciers, flora diversity, and high ecological stability in the elevated zones may account for the lower rate of growth in mountainous regions (Li and Sheng 2012). The increase in the annual temperature has mostly been associated with an increase in both the minimum temperature and maximum temperature in the Himalayan region.
The study exhibited a shift in the overall temperature range in the Himalayan region, with a high increase in Tmin (1.0 to 1.5 C) and a comparatively lower increase in Tmax (0.2 to 0.8 C) (Figure 2d and e), highlighting the significant warming trends in the Himalayan region, has had an impact on glacier environments and ecological sustainability in the region. Previous study reported that extreme cold events (number of cold days and cold nights) have reduced significantly whereas extreme warm events (number of warm days and warm nights) have increased greatly in the entire HKH region during 1961-2015 . Furthermore, the trends of extreme events concerned with Tmin have been observed mostly higher than trends of Tmax .
The standardised anomaly (StdA) of annual mean temperature (Tmean) based on long-term mean exhibited a varying pattern in the Himalayan region, ranging between 9 C/year to 22 C/year during 1901-2018 (Supplementary Figure S1a). The study exhibited an increasing anomalous pattern in mean annual temperature (1 to 3.5) during the recent decades (1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018). However, there was no significant change recorded in standardised anomaly of Tmean during 1901Tmean during -1940Tmean during , barring the years 1902Tmean during , 1915Tmean during , and 1921 wherein a moderate rise in standardised anomaly of Tmean (0.5 to 1.5) recorded in the western Himalayan region. Although the standardised anomaly of Tmean increased (1 to 2.5) during 1941, 1946, 1947, 1953 and 1958 in the western and central Himalayan regions. Moreover, a significant rise in the standardised anomaly of annual mean temperature has been observed during the last 21 years    Figure S1c). Previous studies reported a significant increase in the annual mean surface air temperature in the HKH (0.10 C/decade) during 1901 to 2014, while the warming rate over the last 50 years was 0.2 C/decade (p > 0.05) (Krishnan et al. 2019). A high warming trend with 0.5 C was observed in the Himalayan region during the recent period  in contrast to the preceding period   (Brohan et al. 2006;Diodato et al. 2012). Significant growth of tree rings in high-elevation tree-ring chronologies in the western Himalaya regions as evidence of increased temperature (Borgaonkar et al. 2009). This current warming trend in the region corresponds to increasing temperatures in the northern hemisphere and follows the worldwide warming pattern of the twentieth century (Mann et al. 1999).

Analysing patterns of precipitation in the Himalayan region (1901-2018)
The CRU based long-term cumulative mean annual precipitation pattern in Himalayan region was analysed for the 118 years   (Figure 3a). The study exhibited a parabolic pattern of long-term cumulative mean annual precipitation with high ($>3000 mm/year) to low (<200 mm/year) from eastern to western directions across the Himalayan region (Figure 3a). The large parts of the Himalayan regions were observed under moderate high cumulative mean annual precipitation (1200 to 1800 mm/year) comprising Central, and Eastern Himalayas, while the western Himalayan region was primarily observed under the low to very low cumulative mean annual precipitation (200 to 600 mm/year).
The long-term precipitation trend analysis revealed a slightly decreasing trend (À2 to À1.5 mm/year) in annual precipitation (with 95% significance level) in the Himalayan region over the past 118 years . Most of the HUAs exhibited a negative trend while the western HUAs (Srinagar and Shimla) exhibited a positive trend. The central Himalayan region primarily observed a decreasing trend (À2.5 to À1 mm/year) followed by eastern Himalayan region (0 to À1.5 mm/year) (Figure 3b). Similar findings have also been reported by highlighting a slight decrease in precipitation trend in the whole HKH region over the past 100 years, however during 1961-2013, a significant increase, with a rate of 5.28% per decade was observed ) and specifically in the central Himalayan region (Thakural et al. 2018). Previous studies reported inconclusive trends of precipitation in the entire HKH region over the past century. However, the long term  trend of annual precipitation in the whole HKH region did not indicate a rising trend. While the monsoon precipitation is projected to increase (4-12%) in the coming times and in the long run (4-25%). Winter precipitation is forecasted to rise (7-15%) in the Karakoram, but to decline slightly in the Central Himalaya (Krishnan et al. 2019).
The standardised anomaly of annual cumulative precipitation based on its long-term assessment of mean for 118 years  represented a varying anomalous pattern with an which has mostly kept on increasing pattern over the past 118 years. The study indicated a low anomalous precipitation in the western Himalayan region during 1901-1920 barring 1917 that exhibited a high anomalous precipitation (2.5 to 3.5). Moreover, western Himalayan region exhibited much higher anomalous precipitation during 1933, 1957, 1996 and 2015. In contrast, the central and eastern Himalayan region observed an episodic variation with a high deficit in precipitation (À2 to À3.5) in 1957 and 2015, while slightly increasing anomalous precipitation (1.5 to 2.5) during 1934during , 1936during , 1984during , 1998during and 2000).
In the present study, extreme events (extreme wetness or dryness) have been studied over the past 118 years in the Himalayan region (Figure 4). The analysis of high standard anomaly of precipitation (!2.3) referred as wet years exhibited a relatively low or negative anomalous precipitation in the western Himalayas during 1901Himalayas during -1920Himalayas during barring 1908Himalayas during , 1917, while the central and eastern HUAs recorded a low/moderate anomalous precipitation (À0.5 to 1.5). However, in the recent 20 years (1999-2018), a gradual decline (À0.5 to À1.5) was observed in the western HUAs except during 2015 (Supplementary Figure S2b).

Percentile-based temperature indices (2000-2019)
A decreasing trend was apparent for cold indices (TX10p and TN10p) highlighting a decrease in the number of cold days (2 to À6% of days) and cold nights (À2 to À4% of days) in the entire Himalayan region (Figure 5a and b). The number of cold days (TX10p) declined especially in the central and eastern Himalayan region while the number of cold nights (TN10p) observed a significant decline in the western Himalayan region followed by eastern and central Himalayan regions. An increasing trend of warm indices (TX90p and TN90p) indicated an increase in the number of warm days (4 to 10% of days) and warm nights (5 to 20% of days) as well. However, the frequency of warm nights (12 to 18% of days) was higher than warm days (8 to 14% of days) in the eastern Himalayan region. The spatial observation showed a higher number of warm days (8 to 16% of days) in the central Himalayan region than the warm nights (8 to 14% of days). Whereas, the western Himalayan region exhibited almost similar trends (8 to 14% of days) in the frequency of warm days and warm nights (Figure 5c and d). In addition, the magnitude of trends is statistically significant at 95% confidence level. Previous studies in the HKH region reported that extreme temperature indices have changed significantly over the period 1901-2014 with a declining frequency of extreme cold days and cold nights (daytime: 0.85 days/decade, night time: 2.40 days/decade), whilst increasing frequency of extreme warm days and warm nights (daytime: 1.26 days/decade, night time: 2.54 days/decade) (Krishnan et al. 2019). Moreover, alterations in extreme cold events in the HKH region in the previous decades seemed to be more sensitive to altitude, with declining cold nights and cold days with elevation, while alterations in warm extremes did not show any identifiable relation with altitude ).

Annual maximum and minimum values of daily temperatures (2000-2019)
The minimum value of the annual daily Tmin (TNn) showed an increasing trend (1 C to 2.2 C) especially in the western Himalayan region and some parts of central and eastern most Himalayan regions over the last 20 years (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019). While most of the central and eastern region exhibited a declining trend (À0.8 C to À2 C) in the annual daily Tmin (Figure 5e). The minimum annual daily Tmax (TXn) has been increased in the eastern and central Himalayan region (0.6 C to 2 C) in contrast to the western Himalayan region which observed insignificant change. However, the trend of minimum value of daily temperature was not very significant (Figure 5f). An increasing trend has been observed in the maximum value of the annual daily temperature during the last two decades in the entire Himalayan region. Both TNx (0.8 C to 1.6 C) and TXx (0.8 C to 1.8 C) have increased significantly over the years, indicating a substantial rise in the daily maximum temperature which was higher in the central Himalayan region followed by the eastern and western Himalayan region (Figure 5g and h). Past studies on temperature extremes reported the maximum value of daily Tmax/minimum value of daily Tmin (TXx/TNn) showed rising trend over the entire HKH region (Krishnan et al. 2019).
The emission of greenhouse gas (GHG) is considered as one of the major anthropogenic causes of climate warming in the entire HKH, and its consequences are apparently larger than the rest of the world (You et al. 2010). Moreover, the HKH covers the largest section of the cryosphere apart from the polar zones ). The recent decades have exhibited declining trends in snowfall and retreating glaciers as well as reduction in both perimeter and area coverage of the glaciers (Yao et al. 2012a), exhibiting rapid warming in the region. In contrast, the Karakoram ranges of Himalayas received more winter precipitation in association with intense amplitude changes of synoptic western disturbances (Sabin et al. 2020).
Similarly, the simple daily intensity index (SDII) observed a significant decreasing trend (< À3 mm/20 year) in the central and eastern Himalayan region in contrast to western Himalayan region that observed a moderately increasing trend (>2 mm/20 year) (Figure 7b). The trend of heavy (R10 mm) and very heavy precipitation days (R20 mm) exhibited a high decreasing trend primarily in the central Himalayan (R10mm/R20 mm: À11 to À13 days) followed by eastern Himalayan region (R10 mm/R20 mm: À6 to À10 days) (Figure 7c and d). The coverage and intensity of high decreased R10 mm was comparatively higher than R20 mm exhibiting the lesser number of very heavy precipitation (! 20 mm).
There is an increasing trend in the number of consecutive dry days (CDD; À2 to 3 days/20 years) except in the northernmost parts of the Himalayas during the periods 2000-2019 ( Figure 7e). Meanwhile, the consecutive wet days (CWD) had a decreasing trend (À1.5 to À5 days/20 years) in the entire Himalayan region. The increasing number of dry days over wet days indicated drought risk in the region, especially in the central Himalayan region (Figure 7f). There are diverse trends in the frequency of wet days (R90p), very wet days (R95p) and extremely wet days (R99p) (Figure 7g-i). There is a decreasing trend in the frequency of wet days and very wet days in the central and eastern Himalayan region whereas the western Himalayan region observed an increasing trend. However, extremely wet days did not observe any trend except the northernmost Himalayan region which observed an increasing trend in the last two decades.

Meteorological variability in Himalayan urban agglomerations
The HUAs experienced an increasing trend of moderate rise in temperature during 2000-2019, wherein the central HUA (Kathmandu) exhibited a moderately high increase in temperature (0.008 C/year to 0.01 C/year) followed by eastern HUAs (Thimphu and Itanagar) that experience a moderate rise in annual temperature (0.006 C/year to 0.008 C/year). The city-based study indicated an increasing anomalous pattern in both Tmin and Tmax in western HUAs in the 20 years in contrast to the central and eastern HUAs having slightly decreasing values. The standardised anomaly of minimum temperature has mostly been affected (Figure 2c). The cumulative annual precipitation  for the selected HUAs lies between 600 and 2200 mm/year during 1901-2018 ( Figure 3a). Among the major HUAs, Dehradun, Kathmandu, Thimphu and Itanagar exhibited declining trends barring Srinagar, which exhibited an increasing trend (1 to 1.5 mm/year) in annual precipitation. However, the other HUAs viz., Shimla and Gangtok observed negligible changes. Extreme condition of wetness was observed in Srinagar in 1957(2.34) and 2015(2.52), Shimal in 1909(2.66) and 1917(3.45), Dehradun (3.11) in 1917, Kathmandu (2.55) in 1913, Gangtok in 1984(2.63) and Thimphu in 1911. Extreme conditions of dryness were observed in Gangtok and Thimphu in 1957 and in the central Himalayan region (Kathmandu) in 2014 (Figure 4).
The city-wise climate extreme studies using various temperature and precipitation indices during the period 2000-2019 exhibited moderate decrease in number of both cold days (TX10p: 0 to À8% days) and cold nights (TN10p: 0 to À6%) in the major UAs of the Himalayan region. The number of cold days were observed moderately higher (0 to À2% of days) in the western HUAs (Srinagar, Shimla and Dehradun; Figure 8a-c) compared to the central (Kathmandu: À4 to À8% of days; Figure 8d) and eastern HUAs (Gangtok: À2 to À4% of days, Thimphu: À4 to À6% of days, Itanagar: À6 to À8% of days; Figure 8e-g). While the frequency of the number of cold nights were moderately lower (À2 to À8% of days) in all the selected urban agglomerations barring Thimphu, which observed no change (Figure 8f). In contrast, the frequency of warm indices viz., warm days (TX90p) and warm nights (TN90p) were observed moderately high to very high (4 to 18% of days) in Himalayan urban agglomerations during 2000-2019. The number of warm nights were comparatively highly increased (8 to 12% of days) than number of warm nights (4 to 8% of days) as prominently observed in the central (Figure 8d) to eastern HUAs (Figure 8e-g), while the Srinagar and Shimla recorded less increase in number of warm nights and warm days, respectively (Figure 8a and b).
The minimum annual daily Tmin (TNn) was observed increase TNn in eastern (0.6 to 0.8; Figure 8e-g) and western (0.8 to 1.2; Figure 8a and c) HUAs barring Shimla that recorded decrease in TNn (À0.8 to À0.9; Figure 8b), while the central HUA recorded moderate TNn (À0.6 to 1; Figure 8d).The minimum annual daily Tmax (TXn) was observed increasing TX (1.2 C to 2.2 C; Figure 8d-g) in Central to Eastern HUAs (Kathmandu to Thimphu) while western HUAs recorded moderate decrease in TXn. The maximum annual daily Tmin (TNx) and maximum annual daily Tmax (TXx) was observed high to very high increase (1 C to 2 C) in all the selected HUAs barring Shimla that observed moderate TXx (À0.2 C to 0.6 C; Figure 8b) indicating a significant warming implication in all the HUAs during 2000-2019. Moreover, the number of warm spell days (WSDI) was observed increasing (1 to 3 days/yr) while number of cold spell days (CSDI) was slightly reduced (À0.9 to 0.4 day/yr) in all the major HUAs barring Dehradun, which observed less increase in WSDI and comparatively higher increase in CSDI during 2000-2019 (Figure 8).
The extreme events for precipitation for the major HUAs indicated an increase in maximum 1-day precipitation (Rx1day) in western (Srinagar: 10 to 15 mm; Shimla: 15 to 25 mm; Figure 8a and b) and eastern HUAs (Thimphu: 10 mm to 20 mm; Itanagar: 0 to 10 mm; Figure 8f and g) barring Dehradun (À5 mm to À15 mm; Figure 8c), while decrease in central HUA (Kathmandu: À50 mm to À60 mm; Gangtok: À10 mm to À15 mm; Figure 8d and e). The simple daily intensity index (SDII) was observed to increase in Srinagar (1 to 2 mm/yr; Figure 8a), Dehradun (1 to 2 mm/yr; Figure 8c) and Itanagar (1 to 2 mm/yr; Figure 8g) while decreasing in remaining HUAs (À1 to À3 mm/ yr). Whereas the frequency of heavy (R10mm) and very heavy precipitation days (R20mm) were observed decreasing in western and extreme eastern HUAs while increasing in central (Kathmandu: À9 to À11 days/20yrs; Figure 8d) and remaining eastern HUAs. The percentage of wet days (R90p), very wet days (R95p) and extremely wet days (R99p) were moderately decreased (À8 to 8%/20 yrs) in all the HUAs, primarily central, eastern followed by western HUAs. However, the western HUAs exhibited moderately higher (2 to 8%/20 yrs; Figure 8a-c) percentage of wet days and very wet days than eastern HUAs except Itanagar which also exhibited moderately increasing percentage of wet days. Moreover, the frequency of very wet days in Srinagar (4 to 6%/20 yrs; Figure 8a) was moderately increased than the rest of the UAs. The scenario of consecutive wet days (CWD) and consecutive dry days (CDD) exhibited that the frequency of CWD was moderate (À3 to 1 day/20 yrs) in all the HUAs except for Gangtok which observed decreased (À3.5 to À4.5 day/20 yrs; Figure 8e) frequency of wet days. Thimphu (4 to 5 days/20 yrs; Figure 8f) observed an increased CDD than the rest of the HUAs which observed moderate decrease (À1.5 to 2.5 days/20 yrs) frequency of dry days (Figure 8).
Mountain communities are highly vulnerable to extreme weather events as it exhibited with increasing warming conditions in the Himalayan the major Himalayan Urban Agglomerations over the last two decades . The city scale study exhibited comparatively higher warming conditions in Kathmandu and Dehradun UAs, while an apparent reduction in precipitation recorded in Kathmandu and insignificant change in precipitation in Dehradun. Moreover, the UAs, Gangtok, Thimphu and Itanagar exhibited increased warming with a declining precipitation trend in contrast to moderate warming with increasing precipitation in Srinagar and Shimla. Given the effects of climate change, it's probable that further warming is linked with anthropogenic effects such as increased emissions and aerosol deposition, could hasten the melting of alpine glaciers. The warming in conjugation with increased heavy precipitation will certainly exaggerate the flood risk in the lower Himalayan regions. Building and augmenting disaster resilience is very pertinent at regional scale through community participation, awareness, inclusive approach, adopting nature-based solutions, green space provisioning, to mitigate the adverse impacts of climatic variability (Mishra et al. 2019). Promoting green-blue urban as well as in the fragile ecosystem of Himalayan region is one of the very effective strategies to mitigate the drastic rising temperature, erratic precipitation and extreme climatic events. Strategies comprising, properly planned development activities in a sustainable manner, avoiding large scale land use alterations, retaining slope through afforestation, augmenting livelihood security to least impact the forest ecosystem, are some of the key variables to aid in mitigating climate change extremes particularly due to anthropogenically induced (Epple et al. 2016). It will help in reducing the vulnerability and enhancing the resilience of cities in respect of climatic change in the mountains. The fine resolution data for understanding the physical processes contributing to extreme events are very pertinent (Kim et al. 2010) and will be very useful for city scale for climate studies. More extensive and long-term monitoring of glacier ice volumes, assessment of alpine flora and fauna, landscape and transboundary measures to biodiversity conservation, open data interchange, and cooperation across all nations are suggested to overcome data gaps in the Himalayan region (Sarkar, 2007). The complex WRF modelling approach will provide the additional dimensions of changing patterns of climatic conditions due to large scale anthropogenic induced land use alterations.

Conclusions
The long-term study on climatic variability in the Himalayan region focusing on major HUAs exhibited a declining annual precipitation and rising annual temperature over the last 118 years . The long-term trend analysis based on Mann-Kendall test revealed an increasing trend in the annual temperature (0.002 C/year to 0.01 C/year), while a slightly decreasing trend in the annual precipitation (À2 to À1.5 mm/year) in the entire stretch of Himalayan region. The spatial observation exhibited an overall shift in the temperature range indicating a warming trend in the Himalayan region especially in the central Himalayas, with a high increase in Tmin (1.0 to 1.5 C) and a comparatively lower increase in Tmax (0.2 to 0.8 C) during 1901-2018. The standardised anomaly of mean annual precipitation and mean annual temperature has also been observed increasing with significant changes during 1999-2018. Though the precipitation seemed to be varying spatially over the years, the temperature has gradually been increasing over the past 118 years. The study of the extreme events indicated a higher number of warm days (8 to 16% of days) in the central HUA than the warm nights (8 to 14% of days). Also, the number of cold days declined especially in the central and eastern HUA, while the number of cold nights observed a significant decline in the western HUA followed by eastern and central HUA. The percentage of wet days, very wet days, and extremely wet days declined moderately (À8 to 8%/20 years) in all the Himalayan UAs, primarily central HUA. The consecutive dry days observed an increasing trend (À2 to 3 days/20 years), while consecutive wet days had a decreasing trend (À1.5 to À5 days/20 years) in the entire Himalayan region during 2000-2019. Among the selected HUA, Kathmandu observed comparatively higher warming and reduction in precipitation pattern while Srinagar observed an increasing trend of precipitation. The present study demonstrated precipitation and temperature variations in the Himalayan region especially in UAs during 1901-2018. The overall increase in the annual temperature observed in the study area is primarily attributed to an increase in the minimum and maximum temperature. Despite the uncertainties, the research will contribute to a better understanding of the intricate connections between the climatic components of the mountainous system, as well as improve the possible implications of climate change and variability on natural and human systems in critical places throughout the world.