Spatio-temporal trends in precipitation and temperature means/extremes in the Himalayan states of India

With an increase in the emission of greenhouse gases, noticeable trends in climate patterns are reflected globally as climate change (Trenberth 2011; Allen et al. 2012). Climate change has led to the warming of the atmosphere, affected the hydrological cycle, and influenced all the critical parameters that determine the climate of a region (Trenberth 2011). There is an agreement among the climate community that the overall radiative forcing has increased significantly over the past decades, thus increasing the global temperatures and precipitation patterns worldwide (Palazzi et al. 2013). The persistence of climate change has also established a condition for the occurrence of extreme climatic events (Sharma & Goyal 2020). In recent decades, extreme climatic events linked to increasing global temperature and accelerated warming have been a significant concern (Allen et al. 2012). The potential hazards of extreme weather events are realized by understanding their close association with aspects like destructive floods and unusual droughts, which have led to substantial damage to society, economy, agriculture, and a colossal loss of life (Changnon et al. 1999). Besides other physical mechanisms, greenhouse gases exhibit a leading role in driving the climate extremes worldwide (Zhang et al. 2011). The changes in the intensity and frequency of climate extremes disturb the natural ecosystems and the life of humans on earth (Easterling et al. 2000). While the precipitation extremes leading to devastating floods are creating wreckage majorly in the mountain and coastal areas, a drop in the percentage of cold days and nights with a significant rise in hot days and nights are leading to droughts and heatwaves in many regions of importance (Allen et al. 2012). Rising temperatures have altered the snowfall pattern in higher altitudes, resulting in a shift from snow to greater precipitation events, leading to the retreat of glacier cover and a potential risk to the water resources downstream (Houghton et al. 2001; Chiphang et al. 2018).

As climatic events behave differently in different parts of the globe, it has become necessary for resource planners and system managers to assess the means and climate extremes (Yang et al. 2012). The effects of climate change vary regionally over the earth. Hence, it has become vital to study the regional climatic extremes in addition to climatological means to reveal the spatially inconsistent and region-specific trends. Home to half of the hotspots of global biodiversity, the mountainous regions of India are the most exposed to climate change, the distinct signs of which are glacier reduction, shrinkage, and changing streamflow (Kohler & Maselli 2009). The mountainous Himalayan Region, located in the northern part of the Indian subcontinent, is an abode to the major river basins critical for agriculture, tourism, and economy. The significant heterogeneity in topography and varied climatic gradients in the mountainous regions is associated with environmental conditions and complexities. The Himalayan region houses a wide variety of flora and fauna. The significance of the Himalayan Region on the environment and economic activities like agriculture makes the regional assessment of climatic extremes of great importance to planners at both regional and national levels (Dimri & Dash 2011).

For researching climate change, a requisite detection of the long-term historical hydrologic trends in the climate system is essential (Houghton et al. 2001), a challenging task in regions in and beyond high altitude regions is assessing and precisely representing climate dynamics because of the paucity of the network of meteorological gauges and long-term records of quality data (Kohler & Maselli 2009). While analyzing the role of extreme weather events in different Himalayan Regions is of great prominence, there is limited extensive research to understand the complexities of climate extremes exclusively in the Himalayas (Zhang et al. 2011; Bhardwaj et al. 2021). Furthermore, the studies conducted in the Himalayas have mainly focused on mean climatology and trend detection using limited station data. Recent investigations specified significant trends in precipitation and temperature in the northwest Himalayan region (Bhutiyani et al. 2010; Bharti et al. 2016; Malik & Kumar 2020) and the neighbouring Himalayan regions (Archer & Fowler 2004; Khatiwada et al. 2016). Mir et al. (2021) reviewed different studies on the impact of climate change on the water resources in the Himalayas, reporting a consistent rise in air temperature, precipitation, and extreme climatic events. With the limited research and higher occurrences of extreme events, extreme event analysis has gained popularity in recent times in India and the Himalayan region. Yaduvanshi et al. (2021) computed the climate extremes for the different climate zones of India and reported significant trends at all the warming levels, justifying an increase in the hot climate extremes and a decrease in the cold climate extremes. Kumar et al. (2015) analyzed that the dominant flood events correlated to varying precipitation intensity in the Himalayan Region. Goswami et al. (2018) and Sharma & Goyal (2020) examined and interpreted the extreme indices for spatiotemporal variations in Sikkim-eastern Himalayas, reporting an increase in the temperature and precipitation extremes. Bharti et al. (2016) conducted a state-level study on the precipitation extremes in the Northwest Himalayas and stated significant increasing trends in the frequency of extreme precipitation events in the plains and foothills of the region. Ren et al. (2017) analyzed the long-term trends in the mean and climatic extremes of temperature and precipitation and stated a statistically increasing trend in annual precipitation and precipitation days. Realizing the importance of the Himalayan Region in the Indian context and global climate change, the study has focused on a comprehensive analysis of the following objectives:

• 1.

  Annual and seasonal trends in precipitation and temperature means in the Himalayan states of India.

• 2.

  Trends in precipitation and temperature extremes.

To quantify the climate extremes, we used the extreme indices from the globally coordinated set of 27 climate indices designated by The Expert Team on Climate Change Detection and Indices (ETCCDI) in 1999 (Peterson et al. 2001). These indices quantifying the climatic extremes are widely used (Pitman et al. 2012). The authors believe that the comprehensive results presented in this study will facilitate resource planners to formulate strategies for mitigating the impact of climate change on the environment, water resources, and above all, human life.

In this work, the areas selected for the study are the states of Himachal Pradesh (Western Himalayas), Uttarakhand (Western and Central Himalayas), Sikkim (Eastern Himalayas), and Arunachal Pradesh (Eastern Himalayas) in the IHR. Figure 1 shows the geographical locations of the states. The mountainous and hilly state of Himachal Pradesh has an area of 55,673 km² with an elevation ranging between 300 to 6500 m a.s.l and average annual precipitation of 1180 mm/year. The climate varies from sub-tropical to semi-arctic conditions in the state, depending on the geologically distinct zones. The topographically diverse state of Uttarakhand, located on the southern slope of the Himalayan Range, has an elevation ranging from 180 to 7500 m a.s.l. Uttarakhand spans an area of 53,483 km², of which an area of 46,035 km² is mountainous. As per Köppen's climate classification scheme, the state of Himachal Pradesh falls in the major climatic group E, exhibiting a cold, polar-type climate. The state of Uttarakhand under the same classification shows a tundra-type (represented by the letter ‘T’) climate and falls in the climatic group ET. Sikkim covers an area of 7096 km² and has an average annual precipitation of 2400 mm. The maximum temperature in the high mountain ranges of these two states varies from 0 to 10 °C. The altitude in Sikkim ranges between 300 to 9000 m a.s.l, with the higher elevation zones in the northern, western, and eastern parts of the state. In the east of the Himalayas, Arunachal Pradesh spreads over 83,743 km². In Köppen's classification scheme, the eastern states of Sikkim and Arunachal Pradesh fall in the major group D, exhibiting a cold, humid continental type climate. These states are characterized by severe, cold, and humid winters with no dry season (represented by the letter ‘f’) followed by short and cool summer months (represented by the letter ‘c’) and fall in the Dfc climate group.

The Himalayas feed the major perennial rivers, with this region rightly addressed as the water tower of Asia, making it a location of great prominence. We chose the Himalayan states for our study, given the considerable variability in the different Himalayan divisions in precipitation and temperature patterns, altitude, and topography. The state-level vulnerability index assessed for the Himalayas helped select the geographically distinct states. The analyses of the four Himalayan states will be beneficial for visually interpreting the trends on maps from different Himalayan regions. The individual spatio-temporal patterns developed for each state give an idea of their vulnerability to hazards without referring to individual studies.

The ground-based, high-resolution, daily precipitation dataset of IMD (India Meteorological Department), Pune, available from 1901 to the present, was used in this study. The spatial resolution of this fourth dataset of IMD, referred to as IMD4, is 0.25×0.25° latitudinal/longitudinal resolution (Pai et al. 2021). This dataset has been developed from the existing rain gauges available throughout the country after stringent quality checks with the other existing gridded datasets of IMD. The dataset is reliable because of the quality data used for its development. The new quality dataset (IMD4) has approximately 900 more stations and the rain gauges used in the older dataset. Given the large study area, the other coarser datasets were inappropriate for the study. The other high-resolution dataset (0.25×0.25°) available from 1948 to 2016 used for the maximum and minimum temperature records has been made available by the Terrestrial Hydrology Research Group, Princeton University (Sheffield et al. 2006). Compared with the IMD dataset, this dataset shows agreeable results for the Indian mainland (Mishra et al. 2014).

The methodology followed in the study includes the change point detection of the time series data, the computation of annual and seasonal means of precipitation and temperature, their variability, and trend analysis. The test for change point analysis was carried out on the annual time series to identify a significant change point, after which the statistical properties of a time series dataset change. We calculated the estimated change in mean using the ‘changepoint’ package in R programming software (Killick et al. 2016). Changepoint calculates the optimal positioning and the number of changepoints for data. After change point detection and inputs from literature, we grouped the time series of 60 years (1955–2016) into two periods of 30 years each. For analysis of trends, we refer to the period from 1955 to 1985 as Period 1(P1) and 1986 to 2016 as Period 2 (P2). Various research works confirmed a faster rate of warming in the Himalayas from 1951 to 2014, with a rapid increase starting from the 1980s. Considering the other aspect of global warming, we observe that the extreme streamflow also doubled post the 1980s because of an increased frequency and intensity of extreme events in the recent past (Chug et al. 2020). The study period was divided based on these observations for a relative analysis (see Fig S (a) of the supplementary material). We classified the seasons according to the IMD classification. In IMD classification, the climate of India has been divided and designated into four seasons, namely: the winter season (January and February), the summer season (March, April, and May), the monsoon season (June to September), and the post-monsoon season (October to December). The coefficient of variation (CV) (Equation (1)) for rainfall and temperature was calculated. We calculated and analyzed the climate extremes and conducted the statistical tests for the calculated means and extremes of P1 and P2 at the end.

The degree of variability estimates the deviation of the values of the variables from the mean.

The non-parametric Mann–Kendall (MK) statistical test is performed for the monotonic trend analysis and statistical significance of data (Mann 1945; Kendall 1948). The MK test performed for data (not normally distributed) is recommended by the research community worldwide (World Meteorological Organization 1966). The MK equation for calculating MK test statistic ‘S’ (Equation (2)), (Equation (4)) and the standardized Z-statistic (Equation (5)) are as given below:

The Z statistic for MK is used to reveal specific trends in the data. Z-values greater than 1.96 confirm a significantly increasing (positive) trend, and values lesser than −1.96 confirm a significantly decreasing (negative) trend in the time series. As the Z statistic follows a normal distribution, the null hypothesis (⁠ for the test is explained as the non-existence of any trend in the time series data (distributed identically and is independent). A two-sided test of alternate hypothesis (⁠⁠, confirms an increasing or a decreasing option of a null hypothesis. In any case if the null hypothesis (H₀) stands rejected, confirming a statistically significant trend in the hydrologic time series.

Hamed & Ramachandra Rao (1998) pointed out the flaws in the MK test and stated that the presence of any type of autocorrelation in the series might be a reason that the null hypothesis is rejected. To prevent such cases, a modified MK test was projected. The modified MK was used in this study. The trends in this study were observed at a 95% significance level (α=0.05).

In the equation, med equals the slope between the data points x_p and x_q, the pth and qth observations at times p and q, respectively. The positive values of ‘med’ indicate an increasing trend, and the negative values show a downward decreasing trend.

We calculated the chosen climate indices and their trends after assessing the trends in annual and seasonal means of precipitation and temperature. The results were analyzed in R software and spatially mapped in Arc GIS 10.8. We selected a list of 14 important climate indices (see Table 1) from the range of 27 indices given by ETCCDI for all the states in the IHR for computation.

Table 2 shows the statistical analysis for the annual CV of precipitation and temperature. We categorized the CV into low, moderate, high, very high, and extremely high classes. When CV<20%, it is categorized as low, moderate when 20%<CV<30%, high when 30%<CV<40%, very high when 40%<CV<70%, and extremely high when CV>70%. The CV is high in the northern and eastern parts of the state of Arunachal Pradesh in both periods in the study area (see Fig S (b) of the supplementary material).

The changes observed in the precipitation and temperature time series in the IHR were evaluated using the non-parametric modified MK, and TSS tests and maps were produced. The current study has confirmed a significant variation in the annual patterns in different zones of the Himalayas, with the seasonal patterns substantially influenced by the regional atmospheric processes. The state-level results of the trends are discussed in the successive sections.

Figure 2(a) and 2(b) shows the spatio-temporal variation of AMP for P1 and P2, respectively.

Most Himachal Pradesh and Uttarakhand regions observed an AMP ranging from 1000 to 1500 mm in P1 and P2, with slight variations in both the states' western and eastern parts. A decrease in precipitation was observed in the west part of Himachal and the eastern part of Uttarakhand in P2. In Sikkim and Arunachal Pradesh, where the precipitation is greater than 2000 mm, the precipitation remained almost the same in both periods. The results show an increase in only the eastern part of Arunachal Pradesh in P2.

A significantly decreasing trend was found in the western, central, southern, and northern parts of the state of Himachal Pradesh in P1, with a slightly increasing trend in precipitation in many regions of the state in P2. There was a transition from a significantly decreasing trend in P1 in Uttarakhand to an increasing precipitation trend in P2, consistent with the study conducted by Malik et al. (2019). As shown in Figure 2(a) and 2(d), a decreasing trend was recorded in Sikkim in most regions compared to a significantly increasing trend in P1. The eastern and the western parts of Arunachal Pradesh showed a significantly decreasing trend in AMP in P2 compared to the trend in P1, similar to Srivastava et al. (2021). The increase and decrease in AMP in most parts of the study area and large spatial variability back the climate change observations reported in different Himalayan regions.

The magnitude and slope of the trend in the states of Himachal Pradesh and Uttarakhand increased in P2, ranging between 0 and 15 mm/year. The data shows a decline in the TSS in the state of Sikkim in P2. A dominant decrease with a slope less than −15 mm/year marked P2 in Arunachal Pradesh.

Figs. S (1–4) (see supplementary material for details) shows the results for seasonal precipitation. All the states exhibited a mixed trend for seasonal AMP. Fig. S1 (see supplementary material) shows the seasonal summer annual mean precipitation (SAMP). An increasing trend in SAMP was only majorly visible in Sikkim (Eastern Himalayas), with a significantly decreasing trend in Himachal Pradesh (Western Himalayas) in P2. Fig. S2 shows the seasonal winter annual-mean precipitation (WAMP). There is a decreasing trend in WAMP in all the states, increasing in some parts of Uttarakhand in P2. This result observed in WAMP is congruent with the analysis presented by Praveen et al. (2020) and Sabin et al. (2020), where the authors showed that the WAMP was lesser than the other seasons. The seasonal monsoon annual mean precipitation (MAMP) is shown in Fig. S3 and shows a significantly increasing trend except for Sikkim and Arunachal Pradesh in P2. In MAMP, the states in the Western Himalayas and the Central Himalayas followed the same pattern as AMP and agreed with the results presented by Malik & Kumar (2020) and Praveen et al. (2020) while the state of Uttarakhand in the Central Himalayas shows disagreement with the results of Srivastava et al. (2021). Studies relate to the varying precipitation trend with a decrease observed in the recent period to the land-use changes and increased urbanization in the Himalayas (Das et al. 2018). Fig. S4 shows the seasonal post-monsoon annual-mean precipitation (PMAMP). We concluded from the visible pattern that the PMAMP saw an increasing trend in Uttarakhand in P2 (Malik & Kumar 2020). Table 3 summarizes the state-level annual and seasonal means of precipitation in the IHR.

Trends in the arithmetic means of temperature indicate significant trends in both maximum and minimum temperature in P2, giving a clear indication that the regions are prone to greater warming rates than the globally averaged rate (Sabin et al. 2020; Srivastava et al. 2021; Yaduvanshi et al. 2021). Figure 3 shows the spatial variation of the annual mean maximum temperature (AMTmx) for the states and regions in the IHR.

The AMTmx for P1 and P2 was almost similar. The results show an AMTmx greater than 20 °C in the southeastern, southwestern, and southern regions of Uttarakhand and Himachal Pradesh and maximum areas of Arunachal Pradesh. In the state of Sikkim, a high value of ATMmx was observed down south. The entire state of Arunachal Pradesh had a significantly increasing trend in P2, where in P1, a significantly decreasing trend in the west, southwest, and southeast and a decreasing trend in all the other regions of AMTmx in P1 was dominant. The results show an increasing trend in both P1 and P2 in the state of Sikkim. The state of Himachal Pradesh shows an overall decreasing trend except for the regions in the northeast in P1. A significantly increasing trend is observed in the southern, southwestern, and southeastern regions in P2. The overall significantly decreasing trend of AMTmx in Uttarakhand in P1 was replaced by a significantly increasing trend in most of the regions except in the southern part, where an increasing trend was observed in P2. In P2, Sen's slope was higher than P1 with a rate of 0 °C/year in all the states of the study area. In P1, the lowest slope magnitude (less than −0.02 °C/year) was observed in Uttarakhand, the eastern part of Himachal Pradesh, and stretches north of Arunachal Pradesh. A Sen's slope greater than 0 °C/year was seen in the state of Sikkim in both periods. Various studies attributed the increase in the AMTmx in most parts of the IHR to anthropogenic climate change, topographic variations, weather dynamics, and a faster rate of warming from 1951 to 2014 (Dimri & Dash 2011; Sabin et al. 2020; Ahsan et al. 2021).

The spatial variation of annual mean minimum temperature (AMTmn) for the P1 can be seen in Figure 4(a), and for P2 in Figure 4(b). As observed in Figure 4(c) and 4(d), the eastern part of the state of Arunachal Pradesh had a significantly increasing trend in P2 and an increasing trend in the rest of the state, where in P1, a significantly increasing trend was dominant.

The results show an increasing trend in both P1 and P2 in Sikkim. The state of Himachal Pradesh had an increasing trend in many regions except the southern and eastern parts, where a significantly increasing trend was seen in P1, and this transitioned to a significantly increasing trend in P2 in the entire state. The overall significantly increasing trend of ATMmn in the state of Uttarakhand in P2 can be seen in Figure 4(d). In Himachal Pradesh and Uttarakhand, the magnitude of Sen's slope was higher than P1 in P2. In P2, a rate greater than 0.04 °C/year was observed in both the northern states of Uttarakhand and Himachal Pradesh. The mechanism for this change is the same as for AMTmx. (Zhao et al. 2019) reported the driving mechanisms as urbanization, shifts in the courses of lowland rivers, topographical variation, elevation-based warming, and climate change.

The warming trends have been observed during all the seasons in the Himalayas (Archer & Fowler 2004). The spatial variation of seasonal mean maximum and minimum temperature for the period from 1955 to 1985 can be seen in Fig. S (5–11) of the supplementary material. The spatial variation of seasonal summer annual mean maximum temperature (SAMTmx) is shown in Fig. S5. The SAMTmx showed a significantly increasing trend almost everywhere except Sikkim in P2. In P2, the winter annual mean maximum temperature (WAMTmx) had an increasing trend everywhere except in Sikkim, and the trends are visible in Fig. S7. Shekhar et al. (2017) warned against the negative impacts of increasing winter temperature in the northwest Himalayas, leading to early glacier melt. The monsoon annual mean maximum temperature (MAMTmx) for P1 and P2 observed mixed increasing and significantly increasing trends in most parts of the study area, with a rise observed in Himachal Pradesh and Uttarakhand, as shown in Fig. S9. There was a significantly increasing trend in P2 in Himachal Pradesh and Uttarakhand and an increasing trend in Arunachal Pradesh in post-monsoon annual mean maximum temperature (PMAMTmx), as can be interpreted in Fig. S11. In Fig. S6, a seasonal summer annual mean minimum temperature (SAMTmn) of less than 10 °C was seen in the upper central, northern, northwestern, and northeastern parts of Himachal Pradesh and Uttarakhand, and a slightly higher SAMTmn in the southwest and southeast in P2 (see Fig S6 (b)). The SAMTmn in the study area almost followed the same significantly increasing trend in P2 in most states. The seasonal winter annual mean minimum temperature (WAMTmn) in P2 remained the same, showing an increasing trend. Still, a decreasing trend can be seen in Sikkim and parts of Arunachal Pradesh in Fig. S8. In Fig. S10, we observe that the seasonal monsoon annual mean minimum temperature (MAMTmn) in P2 decreased in most of the parts of the study area. In P2, a seasonal post-monsoon annual mean minimum temperature (PMAMTmn) lesser than 10 °C was recorded in most of the states located in the IHR, with a significantly increasing trend in Himachal Pradesh and Uttarakhand, as can be seen in Fig. S12. The spatially varying mixed trends in the different geographical locations show how differently the hydrometeorological variables respond to the variations in topography and climate dynamics (Chevuturi et al. 2018). For details on the seasonal analysis of temperature, see supplementary material. A summary of the state-level annual and seasonal means of maximum and minimum temperature in the IHR is presented in Tables 4 and 5, respectively.

Figure 5 shows the results for the consecutive wet days (CWD) or the days when the precipitation remained greater than 1 mm.

There was a transition from a decreasing trend in P1 to a dominant increasing trend in most parts of Sikkim. The CWD shows a decrease only in the northwest in P2. The state of Himachal Pradesh had a dominant decreasing trend in both periods even though a significantly decreasing trend was seen in some regions in the west, mid-north, northeast, and central parts in P2. A significantly increasing trend in CWD was found in some parts in the east and west of Uttarakhand with a dominant increasing trend in other parts of the state in P2 in contrast to a dominant decreasing trend in P1 in the state consistent with the findings of Malik et al. (2019) and Malik & Kumar (2020). As shown in Figure 5(c) and 5(d), a significantly increasing trend in CWD in many of the regions of Arunachal Pradesh in P1 transitioned to a decreasing and significantly decreasing trend in the state in P2. Similar results were reported by Singh et al. (2021). The magnitude of the slope increased in the state of Sikkim in P2, with values ranging between 0 to 0.5 days/year. The Sen's slope also decreased slightly in some parts of Sikkim's northeastern and mid-west regions in P2. The state of Himachal Pradesh and Uttarakhand had a Sen's slope ranging between −0.5 to 0 days/year widely in P1, and the same pattern was seen in P2 in Himachal Pradesh with an increase in slope in most parts of Uttarakhand.

The annual total wet day precipitation (PRCPTOT) had a similar spatial pattern as AMP and is shown in Fig. S13 (see supplementary material for details). The number of heavy precipitation days (R10) or the number of days on an annual basis when the precipitation is greater than or equal to 10 mm is one of the threshold-based indices recommended by ETCCDI. The spatial pattern of R10 on a grid scale in the states located in the IHR can be seen in Fig. S14 (see supplementary material). Figure 6 shows the precipitation results for days when the precipitation is greater than the 95th percentile (R95p).

The precipitation occurring during very wet days is measured annually. The spatial pattern of R95p shows that significant trends were seen in Sikkim and Arunachal Pradesh in P2 compared to P1. The R95p had a significantly decreasing trend in many parts of Uttarakhand and Himachal Pradesh in P1, with a significantly increasing trend in the northeastern part of Uttarakhand, which transitioned to a decreasing trend in Himachal Pradesh and an increasing trend in P2 in most regions of Uttarakhand. A decreasing trend in P2 replaced a significantly increasing trend in P1 in the state of Sikkim. The results show an increasing trend in the southwestern and southeastern parts of Sikkim in P2

Figure 7 shows the annual frequency of the very heavy precipitation days (R20) or the annual number of days when the precipitation is greater than or equal to 20 mm.

We observe a significantly decreasing trend in the western part of Arunachal Pradesh in P2. A decreasing trend of R20 was prominent in both periods in the state. A significantly increasing trend can be seen in a small region in the north of the state of Arunachal Pradesh in P2. The results show an increasing trend in Sikkim in P2. There was an increase in slope in the state of Uttarakhand in P2. The values range between 0 and 1 days /year in most parts of the state. A Sen's slope ranging between −0.5 and 0 days/year was dominant in both periods in Himachal Pradesh. The Sen's slope ranged between 0 and 1 days/year in Sikkim in both periods. An increase in slope with most regions having a Sen's slope between −0.5 and 0 days/year is observed in the state of Arunachal Pradesh in P2. Table 6 summarizes the state-level annual mean of precipitation extremes in the IHR.

The hot extremes to understand the warming trends in a region are the TXX and TNX. The TXX is one of the indices used to calculate the maximum values of daily maximum temperature on an annual timescale in terms of the frequency and occurrence of extreme events in the IHR. The spatial pattern followed by the TXX can be seen in Figure 8 for both periods.

The TXX in both periods did not show notable trends in the spatial pattern. The TXX represents the highest or hottest day temperature of a region and can be interpreted from the figures briefly. The TXX had a significantly decreasing trend in the northwestern, central, and southwestern parts of Himachal Pradesh in P2. The results show a similar significantly decreasing trend in P2 in the northeastern and mid-eastern parts of Uttarakhand and the southeastern parts of Arunachal Pradesh. Figure 8(d) shows a significantly increasing trend in the northwestern part of the state of Uttarakhand, the western and central parts of Arunachal Pradesh, the northwest part of Sikkim, and the northeastern part of Himachal Pradesh in P2. The results show an increasing trend in P2 in most parts of all the states except Himachal Pradesh. A significantly decreasing trend in P2 replaced a significantly increasing trend in Himachal Pradesh in P1. A significantly increasing trend can be seen in a small region in the southeast of Arunachal Pradesh, with a decreasing trend of TXX dominating the area in P2. In Sikkim, the significantly increasing trend of TXX in P1 was replaced by an increasing trend almost everywhere in P2. The spatial pattern of TXX suggests an overall increasing trend in P2. The state of Sikkim had a Sen's slope greater than 0.04 °C/year in both periods. There was an increase in the magnitude of slope in Arunachal Pradesh. The results show a Sen's slope greater than 0.04 °C/year in P2 in most parts of the state from a dominant slope of less than −0.012 °C/year in P1. The highest value of Sen's slope is seen in the central, western, and northwestern parts of Himachal Pradesh in P1, and this decreased to a slope less than −0.012 °C/year in the same region in P2. There was an increase in slope magnitude with the highest rate greater than 0.04 °C/year in the eastern stretch from top north to down south in the state of Himachal Pradesh in P2. The increase in the magnitude of the slope was also observed in the state of Uttarakhand in P2. The results show a Sen's slope greater than 0.04 °C/year in the western and central parts from the northern to southern regions in P2. In P2, there is a decrease in the state of Uttarakhand on the eastern side. The Sen's slope decreased to less than −0.012 °C/year. The TNX, another hot extreme, is one of the indices used to calculate the maximum values of daily minimum temperature on an annual timescale in terms of the frequency and occurrence of extreme events in the IHR.

The TXN, a cold extreme, is one of the indices used to calculate the minimum values of daily maximum temperature on an annual timescale in terms of the frequency and occurrence of extreme events in the IHR. The spatial pattern followed by the TNX can be seen in Fig. S16 for both periods (see supplementary material for details). The TNN, another cold extreme, is one of the indices used to calculate the minimum values of daily minimum temperature on an annual timescale in terms of the frequency and occurrence of extreme events in the IHR. The spatial pattern of the TNN can be interpreted in Figure 9 for both periods.

The TNN is hence a representation of the lowest or coldest night temperature of a region. Himachal Pradesh and Uttarakhand had a dominant decreasing trend of TNN in P1. The results show an increasing trend in most of the Himachal Pradesh, a significantly increasing trend in some central, southern, eastern, and northeastern parts of Uttarakhand, and an increasing trend elsewhere in the state in P2. The state of Sikkim had a dominant decreasing trend in P1 with a significantly decreasing trend only in the northeastern regions, which transitioned to an increasing trend in the state in P2. The state of Arunachal Pradesh had a significantly decreasing trend in P1 except in the northeastern, eastern, and southeastern parts. The results show an increasing trend almost everywhere except for a significantly increasing trend in the mid-south and southeastern parts in P2. There was a considerable increase in Sen's slope in all the states in P2, with the highest values greater than 0.12 °C /year observed in the eastern parts of Uttarakhand and some parts in the mid-south and northeast of Arunachal Pradesh. An increase in Sen's slope in the east and northern parts of Sikkim can also be seen, with values ranging between 0.04 to 0.08 °C /year. The rate of Sen's Slope in P1 was less than 0 °C/year in all the states in P1.

One of the two indices that help us understand the cold events in a region is TN10p, which gives us the percentage of days when the minimum temperature is less than the10th percentile. The cold nights (TN10p) are one of the important indices provided by ETCCDI. In P1, all the states had a TN10p ranging between 10 and 12% of days. In P2, notable trends were seen in almost all the states, as shown in Fig. S 17 (see supplementary material). TX10p shown in Fig. S18, which represents the cool days in a region, is the other of the two indices that help us understand the cold events taking place. It gives us the percentage of days when the maximum temperature is less than the 10th percentile. A decreasing trend in P2 replaced the increasing trend of TX10p in the state of Sikkim in P1. A dominant significantly increasing trend of TX10p in the states of Himachal Pradesh, Uttarakhand, and Arunachal Pradesh transitioned to a dominant significantly decreasing trend in all these states in P2.

One of the two indices that help us understand the hot events in a region is TN90p, which gives us the percentage of days when the minimum temperature is greater than the 90th percentile. The hot nights (TN90p) is one of the important indices for hot events given by ETCCDI, and the results can be seen in Figure 10.

In both periods, the state of Sikkim had a significantly increasing trend of TN90p except for the southeast in P2. A significantly increasing trend in P2 replaced the dominant significantly decreasing trend of TN90p in Himachal Pradesh and Uttarakhand in P2. An increasing trend in P2 replaced the significantly increasing trend in a stretch northwest of Arunachal Pradesh in P1. The state of Arunachal Pradesh had a significantly increasing trend of TN90p starting from the center towards the eastern side of the state, while the rest of the state had an increasing trend in P2.

The other indices that help us understand the hot events in a region is TX90p which gives us the percentage of days when the maximum temperature is greater than the 90th percentile. The hot days (TX90p) are also important indices for hot indices given by ETCCDI and can be interpreted for the study area in Figure 11.

We observe a dominant increasing trend in Sikkim, Himachal Pradesh, and Uttarakhand in P2, with a dominant decreasing trend in Arunachal Pradesh. In P1, there was a significantly increasing trend in Sikkim and the northwestern part of Himachal Pradesh. An increasing trend in Arunachal Pradesh and Himachal Pradesh was dominant in most parts of these states in P1. The results show a decreasing trend in Uttarakhand in P1. There was a higher magnitude of slope in P1 ranging from 0.2 to 0.3 percent days/year in most parts of Sikkim compared to a lower Sen's Slope in the range of 0.1 to 0.2 percent/year in the entire state in P2. Most of the regions in the state of Uttarakhand had a lower Sen's slope of less than 0 percent days/year in P1. This trend transitioned to an increased value ranging between 0.1 to 0.2 percent days/year in most parts of the state except in the eastern, northeastern, and southeastern regions where the Sen's Slope was less than 0 percent/year in P2. We observe the lowest value of Sen's slope in the western and some central parts of Himachal Pradesh and many regions of Arunachal Pradesh in P2. The highest Sen's slope ranging between 0.2 to 0.3 percent days/year was only seen in the western part of Uttarakhand from the northern to the southern side and the northwestern parts of Uttarakhand in P2. Table 7 summarises the state-level annual mean of temperature extremes in the IHR.

An increase in the global mean temperature has led to intensifying environmental hazards because of its association with moisture, humidity, and other critical hydrometeorological variables. The temperature projections indicating an intensification of climate extremes have been reported over India and the Himalayas, as observed in our study (Shekhar et al. 2017; Chevuturi et al. 2018; Bhardwaj et al. 2021; Dash & Maity 2021). An exponential relationship between temperature and moisture availability results in the scaling of precipitation extremes explained by the Clausius-Clapeyron equation (Bao et al. 2017; Roca 2019). The enhanced moisture availability in a warmer environment brings about regional precipitation trends in doubled CO₂ scenarios (Turner & Slingo 2009). Various studies attribute the significant increase in temperature to black carbon resulting in localized warming trends and glacier retreats (Ménégoz et al. 2014). The mechanisms behind the statistically significant precipitation intensities in different divisions of the Himalayas behaving differently to extremes have been identified in research extensively. One such study (Bhardwaj et al. 2021) linked the extreme precipitation events in the Indian Himalayas to the Arctic Ocean climate system backed by the Arctic Amplification. The positive trends were also linked to land use/land cover and not found to be affected by elevation. These mechanisms need to be studied in detail, and exploring them is beyond the scope of this study. We conclude that multiple processes go hand in hand in different geographical settings in the Himalayan Region as the increasing positive spatio-temporal trends give an insight into the future ramifications of climate change.

We observed notable spatio- temporal trends have been observed in all the states of the IHR. The results show significant differences in trends because of the geographical location of the states. Many studies recognized and confirmed a dominating positive precipitation trend around the Himalayan Range and North India by linking the trend to the existence of lofty mountainous ranges, complex topography, and linkages to the changing patterns of saturated water vapor and temperature (Kazemzadeh et al. 2021). The Arctic Ocean climatic circulation, with its alternating positive and negative phases, has been stated to have the strongest influence on the extremes in the Northern Hemisphere (Thompson & Wallace 2000). Many studies have linked the substantially increasing monsoonal patterns of extreme events in the Himalayan Region and subsequent increase in the chosen states of our study to teleconnections (Chevuturi et al. 2018; Ahsan et al. 2021; Bhardwaj et al. 2021). Our study shows a considerable decrease in the cold extremes and a rise in the hot extremes in most parts of the Himalayan states. The findings are consistent with similar studies conducted on extreme events locally and globally. The results compared with relevant studies for trends in means and extremes are presented in Table 8. The statistical tests showed significant positive trends in annual and seasonal spatial precipitation and temperature means and indices, highlighting the importance of the statistical tests briefly discussed in section 3. An increase in the extreme precipitation indices was observed in the study area. The heavy precipitation events are to be monitored well to avoid calamities of any sort as they cause environmental hazards like the 2014 floods in the Western Himalayas, the 1970 floods in the Central Himalayas, and the 2010 floods in the Indus River basin (Roy & Balling 2004; Bhardwaj et al. 2021). On interpreting the results, we have also concluded that Uttarakhand is highly subjected to extreme events (Bharti et al. 2016; Bhardwaj et al. 2021).

The present study revealed significant annual and seasonal trends in precipitation and temperature means, and climate extremes in the Himalayan states of India. The results showed a significant increase in AMP in Uttarakhand, followed by Himachal Pradesh in P2. The changes in the AMP were pronounced in the western Himalayan region in P2. In P2, an AMP of 1989.29 mm was observed compared to an AMP of 1492.14 mm in the state of Uttarakhand in P1. After Uttarakhand, an increased AMP of 1304.42 mm was observed in Himachal Pradesh in P2 compared to an AMP of 1492.14 mm in P1. A significant decrease in the states of Himachal Pradesh and Arunachal Pradesh in SAMP is an alarming trend suggesting changes in the pattern of precipitation in P2. An increase in SAMP was evident in Uttarakhand and Sikkim. The SAMP in Sikkim increased from 398.23 mm in P1 to 610.6 mm in P2. An overall reduction in WAMP with a significant decrease in Arunachal Pradesh and Sikkim introduces the need to understand the physical basis causing a change in the precipitation events in these states in the recent period. A significant increase in the seasonal analysis of MAMP in parts of Himachal Pradesh, Uttarakhand, and Sikkim was noticeable and profound in P2. The MAMP in Uttarakhand was the highest compared to the increase in Himachal Pradesh and Sikkim. We observed an increased MAMP of 1510.92 mm in P2 from 1104.06 mm in P1 in Uttarakhand. A significant decrease in PMAMP in greater parts of the state of Arunachal Pradesh suggests a change in climate in the region. The AMTmx showed an increase with a greater rise in seasonal SAMTmx and PMAMTmx in P2, confirming the studies related to the warming of the planet and that of the Himalayan region. In the state of Uttarakhand, we observed a rise of 0.07 °C/year in P2 compared to a rate of −0.03 °C/year in P1.

The state of Uttarakhand showed a significantly increasing trend in CWD in P2. The annual mean of CWD increased from 25 days in P1 to 54 days in P2. A significantly increasing trend in the east and west of Arunachal Pradesh in CDD is evident in P2 from the maps produced. We observed an increasing trend in the precipitation extremes of R10 and R20 in the western and central Himalayan regions, with a decreasing trend in the eastern Himalayan regions. The R95p in Uttarakhand showed an increasing trend in the state. The annual mean of R95p increased from 357 to 385 mm in P2. The R10, R20, and R95p showed a positive trend in P2 in the western and central Himalayas. The results are evident in our study and consistent with other studies.

We observed an increasing trend in TXX in all the states of the IHR in P2. TXN and TXX increased significantly, mainly in parts of Arunachal Pradesh. An increasing trend in TX90p in most parts of Uttarakhand and Himachal Pradesh was evident in P2. The significantly increasing trend of TX90p in some northern parts of Uttarakhand can lead to accelerated glacier melt in the state. The TN90p also showed a significantly increasing trend in the western Himalayas. The results showed a significant decrease in TN10p, TX10p, and a significant increase in TN90p in the study area in P2. TNX in Arunachal Pradesh, Sikkim, and some parts of Uttarakhand and Himachal Pradesh, increased significantly in P2. TNN showed a significant rise in more than 50% of the regions of Uttarakhand and parts of Arunachal Pradesh.

The long-term comparative analysis of temperature and precipitation trends in the study estimated the warming climate reported around the globe. Climate change awareness has attracted the scientific community towards understanding the water resource potential of a country for the sustenance of agriculture and livelihood. The prevalent knowledge of the trends in the hydrological cycle and its behavior in various regions of the Himalayan Region will help realize the need to take the changes taking place in a warming environment into consideration for appropriate decision-making. The illustrated results update the research community's knowledge by providing insight into the micro-scale-regional patterns in light of global warming to benefit climate change adaptation and mitigation strategies in the state action plans. This study serves as a guide in planning water resources for flood management in vulnerable areas by giving prior knowledge of the exacerbated regional trends to prepare for hazards caused by extreme events. For future studies, an emphasis can be laid on the impact on hydrology and related processes due to the varying temperature and precipitation trends and extremes observed in a region at a micro-level individually. Future work will use high-resolution regional climate models (RCMs) to examine the temperature and precipitation trends under future projections (e.g., RCP4.5 and RCP8.5).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.