1 Introduction

The likelihood of significant climatic changes caused by increases in radiatively active gases in the atmosphere, a process usually referred to as ‘climate change’, is causing rising anxiety in both the scientific community and general public. According to the Intergovernmental Panel on Climate Change (IPCC), climate change is a shift in the state of the climate that can also be observed or identified over time (e.g., using statistical tests) – generally decades or more – by variations in the mean and/or changes in its parameters, whether brought on by natural variability or anthropogenic influences (IPCC 2007).

Global surface temperature is considered one of the crucial meteorological climate parameters that plays a vital role in altering the planetary climate system and is frequently employed to find the first indication of climate change. Other major global climate indicators, such as precipitation rate and patterns, temperature and precipitation extremes, ocean heat content, ocean acidification, concentration of atmospheric water vapour, glacier melt in continents, sea-level rise, and the frequency of intense cyclones, are all trending in the same direction as the predicted response from a warmer globe (Stocker et al. 2013).

In addition to the natural variability, human intervention has significantly modified the Earth’s climate, particularly since the Industrial Revolution (Stocker et al. 2013). The activities of a quickly growing and industrialised human population have significantly increased the quantity of heat-retaining gases in the atmosphere over the past couple of centuries. The current rate of climate change, which is unparalleled in the history of modern civilisation, is one of its distinguishing features when compared to natural changes alone. Increased emissions in greenhouse gases (GHGs), aerosols, and changes in land use and land cover (LULC) throughout the industrial period were the main contributors to today's climate change, and all these factors had a considerable impact on atmospheric composition and the planetary energy balance (IPCC 2014).

The phenomenon of global warming, characterised by a substantial rise in global average surface temperatures, has emerged as one of the most critical challenges of our time, profoundly impacting both the natural biosphere and human activities (Vinnikov et al. 1990). Over the course of the past century, both global and local mean surface air temperatures have exhibited a significant upward trend, a well-documented phenomenon (Jones and Moberg 2003). Notably, this temperature increase has become more pronounced in recent decades, as confirmed by multiple assessments, including those from the Intergovernmental Panel on Climate Change (IPCC-AR4 2007; IPCC-AR5 2014 and references therein).

The Indian subcontinent has witnessed significant changes in climate patterns, including increased heat extremes, shifting monsoons, and rising sea levels, primarily attributed to anthropogenic climate change (Singh et al. 2019). Understanding the underlying mechanisms driving these changes and their potential future trajectories is of paramount importance for safeguarding lives and livelihood of people in the region. However, at the regional scale, comprehending climate change remains a formidable challenge due to the paucity of localised observational data and the intricate nature of unique physical processes (Flato et al. 2014). To plan for disaster management and risk mitigation, as well as develop locally applicable adaptation measures, common people and policymakers should be aware of current and predicted changes in regional climate (Burkett et al. 2014). It has been extensively discussed the extent to which global temperature changed in many regions during the 20th century (Cane et al. 1997; Reilly et al. 2001; Brohan et al. 2006; Li et al. 2010; Santer et al. 2011). The first report on the temporal fluctuations over India came in the 1950s (Pramanik and Jagannathan 1954) using observational temperature data from 30 Indian stations between 1880 and 1950. They have observed no noticeable trend in maximum and lowest temperatures. However, the average annual temperature in India was found to rise by around 0.4°C/100 years between 1901 and 1982, according to one of the early studies on modern global warming by Hingane et al. (1985). Srivastava et al. (1992) examined the decadal changes in the parameters related to climate over India between 1901 and 1986, where northern India was found to cool largely while southern India was generally warming. According to Rupa Kumar et al. (1994), the increasing trend in annual mean temperatures over India between 1901 and 1987 was mostly caused by an increase in maximum temperatures, while minimum temperatures remained trendless. After eliminating the global influence of greenhouse gases and natural variability, Krishnan and Ramanathan (2002) suggest that India's whole surface air temperature during the drier months of the year (January–May) has declined by as much as 0.3°C, since 1971. Kothawale and Rupa Kumar's (2005) investigation revealed that both the yearly maximum and minimum temperatures over all of India increased considerably between 1971 and 2003. Kothawale et al. (2010) used temperature data from the years 1901 to 2007 to conclude that there had been a considerable increase in the mean, maximum, and minimum temperatures throughout all of India and that the rate of warming has accelerated recently. Kothawale et al. (2012) investigated the regional and temporal asymmetry of temperature changes across India and the possible impact of aerosols. Likewise, several studies have examined and emphasised temperature trends in India.

The present study is aimed at examining climate change and variability, particularly in temperature trends over North-East India (NEI) and surrounding territories for the past century, with special emphasis on the trends from the recent past (1979–2020), including the multidecadal variabilities till 2020.

2 Data and methodology

2.1 Study domain and prevailing meteorology

The study region encompasses the NEI and its adjoining regions (20º‒30ºN, 85º‒100ºE). The region has primarily four distinct seasons, viz., winter season (December–February; DJF), pre-monsoon (March–May; MAM), summer-monsoon season (June–September; JJAS), and post-monsoon season (October–November; ON). The summers are tropical, with temperatures ranging from 23.3° to 28.3°C. Acute rainfall during the monsoon season, which spans from June to September, can bring people’s lives and activity to a halt, with July experiencing the highest rainfall (565 mm). Moreover, chilly Himalayan winds blowing directly into it bring the region's temperature down to 16º–18ºC during winter. The mean yearly temperature of the area around the study locations is ~24.4°C and mean yearly precipitations are about 2188 mm (Mitchell and Jones 2005; Harris et al. 2014).

To evaluate the variability in surface temperature at the local scale within the study domain and their impact on the regional climate, eleven different study locations namely, Thimphu (THM), Tawang (TWN), Shillong (SHN), Aizawl (AZL), Imphal (IPH), Banmauk (BNK), Dibrugarh (DBR), Guwahati (GHY), Dhubri (DHB), Agartala (AGA) and Dhaka (DAC) with varying altitudes, topography and vegetation have been selected (figure 1).

Figure 1
figure 1

A topographic view of the study domain showing 11 different locations used in the study.

Full size image

2.2 Data source

For temperature, data from CRU and ECMWF Reanalysis v5 (ERA5) have been utilised. CRU provided near-surface temperature at spatial resolution of 0.5º × 0.5º and 1-month temporal resolution (https://catalogue.ceda.ac.uk/uuid/c26a65020a5e4b80b20018f148556681) where units were in °C. On the other hand, the most recent high-resolution ERA5 reanalysis provided 2m-temperature (the air temperature 2 meters above the Earth’s surface) data at 0.25º × 0.25º horizontal resolution (approx. 30 km) and monthly time intervals for the period 1979 to 2020 (https://cds.climate.copernicus.eu), where units were converted from Kelvin to °C.

2.2.1 CRU

The Climatic Research Unit data (v4.05) of near-surface temperature at 0.5º (~55 km) spatial resolution on monthly temporal resolution during 1901–2020 was utilised in this study. These datasets are available over only land areas of the world (except Antarctica) and can be accessed online at https://crudata.uea.ac.uk. Harris et al. (2020) may be consulted for further information on the procedure involved in obtaining these monthly climate observational datasets.

An interpolation technique called angular-distance weighting (ADW) was used to develop the CRU Time-Series (TS) 4.05 data. IDL triangulation routines were utilised in all versions until 4.00.

2.2.2 ERA5

Data from the fifth generation of atmospheric reanalysis of the global climate (ERA5) from the European Centre for Medium-Range Weather Forecasts (ECMWF) and produced by Copernicus Climate Change Service (C3S) was also utilised. From 1979 to the present, ERA5 climatic data cover the entire land and ocean domain of the world and are accessible in a 0.25º×0.25º grid (Hersbach et al. 2018). For the duration of 1979–2020, the monthly temperature data was accessed from https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels-monthly-means?tab=form.

The ERA5 datasets are the most recent version of the ECMWF's 5-day real-time global atmospheric reanalysis from 1979 (Hersbach and Dee 2016) and replaces the ERA-Interim reanalysis (Dee et al. 2011). The ECMWF provides estimates for a wide variety of atmospheric, land-based, and oceanic weather and climate variables every hour and every six hours. The data resolve the atmosphere using 137 levels from the Earth's surface up to an altitude of 80 km, covering the planet on a grid of approximately 30 km. ERA5 incorporates data at decreased spatial and temporal resolutions (https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5).

2.3 Methodology

2.3.1 Mann–Kendall trend analysis

Since a non-parametric test may avoid the issue caused by data skew, it is preferable over a parametric test (Smith 2000). When several stations are assessed in a single study, the Mann–Kendall (MK) test is usually favoured (Hirsch et al. 1991).

The MK test is a distribution-free test for estimating a time series trend (Mann 1945; Kendall 1975), i.e., even if the time series do not adhere to a certain distribution, it may be employed. The test assumes that the data are not serially correlated and largely depends on the quantity of data points. The first point is true for all statistical tests; however, the second cannot be done if comparing data sets from several time periods is the goal. We deal with a lot of data like this in trend analysis. A non-parametric regression approach called Sen’s slope (SS) test is almost always used in conjunction with the MK test to evaluate the trend magnitude and to know whether the \(y\) values are going up or down over time.

The MK test statistics (S) is expressed as (Salas 1993) follows:

$$S= \sum_{i=1}^{n-1}\sum_{j=i+1}^{n}\text{sgn}({x}_{j}-{x}_{i}).$$

Here, \(n\) denotes the number of records, \({x}_{i}\) and \({x}_{j}\) are the data values for the time series i and j (where j > i), respectively and \(\text{sgn}({x}_{j}-{x}_{i})\) is the sign function defined as:

$$\text{sgn}({x}_{j}-{x}_{i})=\left\{\begin{array}{c}1, \quad \;\;\text{if}\;\left({x}_{j}-{x}_{i}\right)>0\\ 0, \quad \;\;\text{if}\;\left({x}_{j}-{x}_{i}\right)=0 \\ -1, \quad \text{if}\;\left({x}_{j}-{x}_{i}\right)<0\end{array}.\right.$$

represents the number of negative differences. The test is carried out using a normal distribution (Helsel and Hirsch 1992) with the mean and variance as follows, for large samples (\(n\) > 10).

$$E\left[S\right]=0$$
$$\text{Var}\left(S\right)= \frac{n\left(n-1\right)\left(2n+5\right)-{\sum }_{i=1}^{m}{t}_{i}({t}_{i}-1)(2{t}_{i}+5)}{18}$$

where \(n\) indicates the overall number of data points; \({t}_{i}\), the number of ties of size i; and m, the number of coupled groups. Here, coupled groups refer to a collection of data points of the same kind. The standard normal test statistic Zs (Z-statistics) is calculated when there are more than 10 data points (\(n\) > 10) using (Hirsch et al. 1993):

$${Z}_{s}=\left\{\begin{array}{c}\frac{S-1}{\sqrt{\text{Var}(S)}}, \quad \text{when} \ S>0\\ \\0, \qquad \quad \;\;\;\text{when} \ S=0\\ \\ \frac{S+1}{\sqrt{\text{Var}(S)}}, \quad \text{when} \ S<0\end{array}\right. .$$

An upward trend is shown when \({Z}_{S}\) is more than zero, while a downward trend is indicated when \({Z}_{S}\) is less than zero. The alternative hypothesis \(({H}_{1})\) is chosen if there is a positive or negative trend in the variable, whereas the null hypothesis \(({H}_{0})\) is chosen if there is no trend. If \(\left|{Z}_{S}\right|>{Z}_{1-\alpha /2}\), a strong trend is detected in the data series, and the null hypothesis is discarded. The normal distribution table in its conventional form may be used to determine the value of \({Z}_{1-\alpha /2}\). If \(|{Z}_{S}|\) > 1.96 at the 5% significance level (α = 0.05) as determined by the standard normal cumulative distribution chart, the null hypothesis of no trend is rejected (Lund et al. 2007; Costa and Soares 2009). The significance levels, \(\alpha = 0.01\;(99\%), 0.05\;(95\%) \;\text{and}\; 0.1\;(90\%),\) were employed in this investigation.

Sen's slope estimator (Sen 1968), a nonparametric technique, is applied in the study of climatic time series and is employed by many researchers (Lettenmaier et al. 1994; Yunling and Yiping 2005; Partal and Kahya 2006; El-Nesr et al. 2010; Tabari et al. 2011, 2012). Sen's method, which is frequently employed to assess the strength of trends in hydro-meteorological time series, presupposes a linear trend in the time series (Lettenmaier et al. 1994; Yue and Hashino 2003; Partal and Kahya 2006). It roughly represents the trend slope, Q, in the sample of N pairs of data points and is calculated as the median of:

$${Q}_{i}=\frac{{x}_{j}-{x}_{k}}{j-k},\quad i=1, 2, \dots , N.$$

where \({x}_{j}\) and \({x}_{k}\) are the data values at times \(j\) and \(k\) \((k > j)\), respectively, i.e., Sen’s slope, \(Q\) = median (\({Q}_{i}\)).

2.3.2 Linear regression

A statistical technique called simple linear regression analysis is used to establish or investigate the relationship between two variables of interest. In this regression model, one variable is called an explanatory (independent) variable, and the other one is a response (dependent) variable. The regression coefficients \({\beta }_{0}\) and \({\beta }_{1}\) in the following equation are estimated through linear regression:

$$y={\beta }_{0}+{\beta }_{1}x+\varepsilon$$

where \(y\) is the study variable (or dependent variable) and \(x\) is the explanatory variable (or independent variable). \({\beta }_{0}\) and \({\beta }_{1}\) are termed as an intercept term (a constant) and the slope parameter/coefficient for the variable \(x\), respectively. The component, \(\varepsilon\), also known as the residuals, is the unobservable error component or the model's error term. It is responsible for the data's inability to fall in a straight line and reflects the mismatch between the true and observed realisations of \(y\). It is important to note that the aforementioned equation yields \(y\) from \(x\). The residual includes the effects of all the various causes on the value of \(y\).

3 Results and discussion

Distinct variability in temperature trends both at spatial and temporal scales during the period 1901–2020 over NEI and surrounding areas was observed (figure 2). The spatial distribution of near-surface temperature during four seasons (MAM, JJA, SON, DJF) retrieved from CRU for 1901 and 2020, as well as associated biases are shown in figure 2. Since the available data series from CRU was relatively lengthy (about 120 years), it was divided into three sub-periods: (a) pre-satellite era (1901–1950), (b) transition era (1951–1978), and (c) satellite era (1979–2020), to perform the trend analysis. By examining the era of pre-satellite, which represents the early 20th century, a time before the availability of satellite-based observations, we aimed to capture the trends and patterns in climate that predate modern observational tools, allowing us to establish a baseline for natural climate variability. The mid-20th century marks the advent of satellite technology, providing a more comprehensive and global perspective on climate phenomena, and the inclusion of this period allows us to assess the impact of improved observational capabilities on our understanding of climate dynamics and to identify any noticeable transitions or changes in trends. The last period encompassing the era of advanced satellite observations marks the deployment of satellites that have significantly enhanced our ability to monitor various climate variables with higher spatial and temporal resolution. This phase is crucial for examining recent trends, identifying anomalies, and assessing the influence of human activities on observed climate changes. By partitioning the analysis into these three blocks, we aimed to delineate the coaction of natural and anthropogenic factors over distinct periods in the history of climate research, which in turn allows us to discern better long-term trends, potential abrupt shifts, and the impact of technological advancements on our understanding of climate variability. Mann–Kendall and Sen’s slope test statistics results for monthly mean near-surface temperature during 1901–1950, 1951–1978 and 1979–2020 based on CRU, and 2m-temperature during 1979–2020 based on ERA5, for the study region, are presented in table 1.

Figure 2
figure 2

Spatial distribution of near-surface temperature in 1901, 2020 and bias for the seasons: (a) pre-monsoon (MAM), (b) summer/monsoon (JJA), (c) post-monsoon (SON), and (d) winter (DJF), based on CRU.

Full size image
Table 1 Results of Mann–Kendall and Sen’s slope test for near-surface temperature from CRU and 2-m temperature from ERA5 averaged over the entire NEI and adjoining regions at different time periods.
Full size table

3.1 Spatiotemporal distribution of near-surface temperature

For each year in consideration, near-surface mean temperatures to the north of ~28ºN were relatively on the lower side, whereas to the south of ~28ºN, the temperatures were generally higher (figure 2). Areas around Bangladesh, which are notably polluted, as well as parts of Myanmar and the Indian states of West Bengal, Bihar, Jharkhand and Chhattisgarh exhibited greater mean temperatures than the rest of the region, in general. The post-monsoon season (SON) displayed a temperature rise (~0.6°–2°C) over the entire region with gradually increasing magnitudes from the southwestern region to the northeastern stretches of the study domain between 1901 and 2020. However, the winter season (DJF) showed a rise in temperatures for most parts of the area except a tiny portion near southern Bangladesh, West Bengal and eastern Myanmar, where an insignificant fall in temperatures (~0.4°–0.1°C) was observed. During the monsoon season (JJA), although the areas experiencing a fall in temperatures were relatively greater, there was a rise in temperatures observed for most parts of the areas lying east of the study domain (~0.2°–0.9°C). The pre-monsoon season (MAM) showed a similar tendency, with the southeastern regions seeing the largest rise in mean temperatures (~0.4°–0.8°C) during the same time period.

3.2 Trend analysis of surface temperature

During the last 120-year period, there has been an increasing trend of magnitude ~0.11°C decade−1 in February, followed by ~0.10°C decade−1 in December and up to ~0.02°C decade−1 in July (figure 3). The trend analyses for other sub-periods were performed using M–K and Sen’s slope test (table 1).

Figure 3
figure 3

Temporal variation of near-surface temperature across various months, retrieved from CRU, over the NEI and adjoining land areas for the period of 1901–2020. The dashed line through the data points represents the linear least square fit indicating the long-term trend in 2m-temperature, its slope yielding the trend (°C/yr).

Full size image

The time series of surface temperature across various months revealed significant warming trends for the mean annual temperatures during the most recent period, 1979–2020, as compared to the variability in the past century (1901–2020) and periods like 1901–1950, 1951–1978 over the entire NEI and adjoining areas. However, the duration 1951–1978 exhibited declining trends for most of the months.

During the first half of the previous century (1901–1950), all the months exhibited increasing trends, with a  maximum increase in December (~0.3°C decade−1) followed by March (~0.2°C decade−1) at 99% and 95% confidence level, respectively (table 1). The temperature increases at a rate of ~0.1°C decade−1 in the rest of the months, at a 99% significance level in August and September; at 95% level in January, February and November; at 90% level in April and May and a non-significant increase in June, July and October. Between 1951 and 1978, the majority of the months displayed declining temperature trends, although non-significant, with the exception of March and November. However, in April and May, the decrease is significant with a rate of ~0.3°C decade−1 at 95% level of significance and ~0.2°C decade−1 at 90% level, respectively. Contrarily, a relatively higher warming rate was observed in the most recent period (1979–2020), especially during the months of summer monsoon (JJAS), post-monsoon (ON) and winter (DJF) seasons for both CRU and ERA5. Warming at a rate of ~0.1°C decade−1/~0.2°C decade−1 to ~0.2°C decade−1/~0.3°C decade−1 at 99% confidence interval was observed in July, August and September for both the data sources. Similarly, warming in the winter months of December and February at a rate of ~0.2°C decade−1 and ~0.2°C decade−1/~0.3°C decade−1 at a 95% confidence interval  was observed. For CRU, May, June and October exhibited a warming of ~0.1°C decade−1 at 90% significance level, while for ERA5, it was observed only during April. However, in November, the former shows a warming of ~0.2°C decade−1 at the 95% level and the latter exhibited the same at ~0.3°C decade−1 at the 99% level. The warming was observed to be non-significant during January and March for both type of data sets. It is worth mentioning here that the spatial resolution for ERA5 (0.25º × 0.25º) is finer than that of CRU (0.5º × 0.5º).

Mann–Kendall and Sen’s slope test revealed an overall significant warming trend for each of the 11 locations over the decades with large variability in magnitude ranging between 0.1° and 0.5°C decade−1 mostly at 99% significance level during June–November (table 2). However, the warming in November at SHN, DHB, and DAC and that in October at THM and TWN were at 95% level, while at AGA, the same was at 90%. Moreover, a non-significant warming trend was observed over DBR. In December, however, high latitude locations such as TWN and THM exhibited the highest warming rate of 0.5°C decade−1 and 0.7°C decade−1 , respectively, at 99% confidence level, followed by BNK and IPH, where warming was at a rate of 0.3°C decade−1 and 0.2°C decade−1, respectively at 95% confidence level. The rate for the month of December in AZL was 0.2°C decade−1 at a 90% level, while the same in the low- altitude locations were non-significant. However, in DAC, temperature trends decreased non-significantly during December and at a confidence interval of 95% during January. The locations close to DAC, viz. DHB and AGA also exhibited non-significant decreasing temperature trends in January. Strong warming trends prevailed in February over most of the locations, with steeper increasing trends with increasing altitudes, the maximum rate being over THM, TWN and IPH (0.5°C decade−1) at a 99% confidence level. March, April and May also showed significant trends across stations with greater elevations ranging between 0.2 and 0.4°C decade−1  and relatively higher trend magnitudes at AZL and IPH. The months of February and between June and November were found to be the time with the greatest number of stations exhibiting significant increasing trends (figure 4).

Table 2 Results of Mann–Kendall and Sen’s slope test for 2m-temperature trend at different locations for the period 1979–2020 using ERA5.
Full size table
Figure 4
figure 4

Number of stations showing positive (increasing) significant and non-significant; negative (decreasing) significant and non-significant at 90% level of significance; and no-trend across different months during 1979–2020, based on ERA5.

Full size image

4 Summary and conclusions

The broad spatiotemporal variability in temperature over NEI and the neighbouring areas is caused by the non-uniform distribution of GHGs, aerosols, and particulate matter in the area (Gogoi et al. 2009; Pathak et al. 2016; Biswas et al. 2017; Pathak and Bhuyan 2022). Pollution attributable to human activity due to a relatively larger population is the primary cause of areas around Bangladesh, Myanmar and the Indian states of West Bengal, Jharkhand, Chhattisgarh and Bihar displaying greater mean temperatures than the rest of the region (Sharma 2003). While both the pre-satellite and satellite eras displayed only a rise in mean annual temperature, the transition era showed both increases and decreases in temperature across different months. Nevertheless, the last four decades showed a tendency towards a warmer climate over the region mainly due to increased anthropogenic activities. Both near-surface and 2m-temperature displayed positive trends for the period 1901–1950, 1979–2020 and during the whole duration 1901–2020.

With changes in height and orographic winds further altering the temperature gradient along various elevation belts, the temperature across the study region is very unpredictable. Additionally, seasonal variations in solar insolation and the amount of available atmospheric moisture influence it. Temperature fields with sharp variations show the region's vulnerability to climate change. It has been observed that regions with higher elevation have experienced a higher warming rate than regions with low elevation. The warming rate is found to be maximum during JJAS at ~1000–3000 m elevation belt, which includes locations such as THM, TWN, SHN and AZL. The NEI's TWN area has the highest warming rate of 0.5°C decade−1. Except for January, all of the summer (JJAS), post-monsoon (ON), and winter (DJF) months showed a larger warming rate for most of the sites, indicating a longer summer and warmer winter and, thus a changing general climate prevailing over the region.

Over the area, the surface temperature is generally on the rise. Trends are driven by trace gases, especially GHGs and aerosols, including PM, emitted primarily by anthropogenic activities due to the growing population accompanied by increasing urbanisation, changes in land use, major roadway expansions, increased deforestation, biomass burning, fossil fuel consumption, leading to an increasing concentration of gaseous constituents including greenhouse gases (GHGs) and aerosols (Mitchell et al. 1995). Despite being less inhabited and having less industrial and human activities than the rest of India, the region serves as a sink to the emissions from the industrialised west and highly populated Indo-Gangetic plains (IGP), where the concentration of both gaseous pollutants and particulate matter is relatively high. The climate prevailing over the NEI and the surrounding regions is conducive for not only formation but also redistribution and thus accumulation of both local and remote atmospheric composition particularly in dry months (Pathak and Bhuyan 2022).

Due to climate change, crop production in the region can be affected. Extreme weather due to a change in the climate might occasionally cause crop damage, whereas long-term changes in means or climate variability can have an impact on agricultural production. Modelling of the impact of increasing temperature means and variability on agricultural output has shown that crop productivity decreases when temperature variability rises in various situations (Semenov and Porter 1995; Mearns et al. 1996; Riha et al. 1996).

Seasonal temperature variations in the region, in particular, surface temperature trends, can significantly influence the monsoon's activity and the effects it has on agricultural production. With summers extending beyond September and winters getting shorter and warmer, this indicates a significant acceleration of the warming over the past four decades, the effect being majorly felt in relatively higher altitude areas and is consistent with current global trends. According to the results of the current study, the future climate of the NEI will experience warming/wetter scenarios as a result of a rise in temperature. The biodiversity, hydrology, and ecosystems in the area may be dramatically affected further by this forecasted scenario.