Climatic trend in different bioclimatic zones in Chitwan Annapurna Landscape, Nepal

Depending upon altitudinal gradient in the Himalayas, the rate of climate change varies from lowland to upland. The Chitwan Annapurna Landscape (CHAL) is the central part of the Himalayas and covers all bioclimatic zones. Analysis of time series data (1970-2019) of temperature and precipitation was carried out in seven bioclimatic zones extending from lowland Terai to higher Himalayas. The non-parametric Mann-Kendall test was applied to determine the trend, which was quantified by Sen’s slope. Annual and decade interval average temperature, precipitation trends, and lapse rate were analyzed in each bioclimatic zone. Out of seven bioclimatic zones, four zones showed a decreasing precipitation trend (lower tropical, upper tropical, upper subtropical, and alpine bioclimatic zones)at the rate of 1.8, 1.98, 2.06, and 1.80 mm/year, and in lower sub-tropical, temperate, and lower subalpine bioclimatic zones, increasing at the rate of 0.45, 1.81 and 1.28mm/year, respectively. Precipitation did not show any particular trend at decade intervals. The average annual temperature at different bioclimatic zones clearly indicates that temperature at higher elevations is significantly increasing more than at lower elevations. In lower tropical bioclimatic zone (LTBZ), upper tropical bioclimatic zone(UTBZ), lower subtropical bioclimatic zone (LSBZ), upper subtropical bioclimatic zone(USBZ), and temperate bioclimatic zone(TBZ), change site and region-specific, this information highlights the need to mitigate climate change in different bioclimatic zones.

ii) examining decade interval climate change patterns in temperature and precipitation at different bioclimatic zones, and iii) determining the average temperature, precipitation, and lapse rate of temperature and precipitation in different bioclimatic zones of the area.

Study area
The study area is Chitwan Annapurna Landscape (CHAL), located in central Nepal between mountains (2200-4000m), and high Himalayas (>4000 ml). In this study, CHAL is categorized into seven bioclimatic zones [22]and [23]with slight modification to evaluate the climatic trends as lower tropical bioclimatic zone (<500m) (LTBZ), upper tropical bioclimatic zone (500- in the high mountains [24]. The average annual rainfall ranges from 165mm at Lomanthang (Mustang) in the northern part to 5,244mm at Lumle (highest rainfall in the country) in mid-hills [2]. Orographic effects cause high spatial variation in precipitation in different zones of the landscape. Nearly 78%of the total annual precipitation occurs during the monsoon season between June and September [16]. Occasional winter rainfall is common including short rainfall in the Siwalik and mid-hills and snowfall is common in high altitude regions.

Climate analysis
Both precipitation and temperature data for all stations in CHAL were procured from the Department of Hydrology and Meteorology (DHM), Government of Nepal. There are 81 and 32 weather stations to measure rainfall and temperature within CHAL, respectively (Table-1).
However, time series data (1970-2019) of daily temperature and rainfall of 26 and 52 stations, respectively, were only used in this analysis due to presence of complete data set of different bioclimatic zones in CHAL (Table-1). The distribution scenarios of functional stations used to depict inferences along different bioclimatic zones are presented in Figs2 and 3. The station details, including the data available date, are presented in Annex 1 and 2.
The rainfall and temperature trends were calculated in spatial (seven bioclimatic zones based upon altitude, climate, and vegetation) and temporal (annual and decade) scales.

Data management
The stations level data were manually checked to determine the odd and unusual patterns. The station data that showed outliers and unusual values were corrected for individual stations by replacing them with the average of the value from the previous day and the next day [2]. If there is an outlier or missing data for about a month or more, that particular year is omitted from the analysis.

Trend Analysis
Mann-Kendall test and Sen's slope methods were used to analyze the climatic trend, magnitude, significance of temperature, and precipitation data in bioclimatic zones. The Mann-Kendall test is a non-parametric tool and one of the best methods to analyze the presence and significance of monotonic positive or negative trend in time series climatic data [25,26,2] .Existence of positive or negative trend among the considered climatic variables was determined by using Mann-Kendall trend test, its quantification was done by using Sen's slope method, and significance by Mann-Kendall Methods in R package version 3.4.4 [27]. The Mann-Kendall method of significance test uses the hypothesis testing approach. In the testing mechanism null hypothesis (Ho) there is no monotonic trend in climatic data, and with alternative hypothesis (H1), there is a monotonic trend in climatic data at significant level. Significant tests at 0.05 confidence levels were used.

Lapse rate of temperature and precipitation
The lapse rate of temperature and precipitation in different bioclimatic zones along altitudinal gradient in CHAL were calculated following [28] and [29] with slight modification. The lapse rate of precipitation was derived by analyzing annual average precipitation sums of the stations lying in each bioclimatic zone by using the following equation: Where PPt LR is precipitation lapse rate, P1 and P2 are the precipitation sums of the highest and lowest bioclimatic zones (in mm) and Z1 and Z2 are their respective elevations of higher and lower bioclimatic zones. The value of Z1-Z2 considered 1 because our calculation is converted into 1 kilometer along elevation change.
All temperature data were aggregated to yearly values for the period of 50years. Temperature lapse rates were calculated through temperature difference between two successive bioclimatic zones i.e. temperature-elevation space [28]. Average temperature is normally assumed to decrease linearly with increasing elevation in CHAL. So, Temp LR was calculated as: Where Temp LR is Temperature lapse rate, T1 and T2 are the yearly average temperatures of the highest and lowest temperature in each bioclimatic zone ( o C), and Z1 and Z2 are their average elevations (m). We calculated the lapse rates based on the strong relationship between ambient air temperature and elevation (30,28,29].

Temperature in different decades
Average temperature in every bioclimatic zone of CHAL has been increasing over the past five decades (

Average annual temperature and precipitation
The average temperature and precipitation values of all stations in each bioclimatic zone have been analyzed for 1970-2019 (Table-5 This result indicates that with increasing altitude from LTBZ to UTBZ, annual average precipitation increases, but in other BZ the precipitation successively decreases (Table-5).

Lapse rate of temperature and precipitation lapse rate
The temperature lapse rate showed2.1 to 2.3 o C decreases with every increase of 500m altitude along different bioclimatic zones in CHAL (Table-5). It implies that every one-kilometer that altitude rises, the temperature decreases by 4.2 to 4.6 o C/km in CHAL.

Discussion
A mixed pattern of precipitation in different bioclimatic zones is in congruence with the precipitation pattern for the whole of Nepal [31]. Previous work in different parts of CHAL were localized to districts or whole regions at small scale analysis based on some meteorological stations of Lamjung [32], Syangja [33], and other districts [2]. According to Poudel and Shaw [32], annual precipitation was increasing in Khudi (855m) and Kunchha (823m) [2]in Terai, annual average rainfall was found increasing at the rate of 0.49mm/year, but Siwalik, Mid-mountain, High mountains, and High Himalayas showed decreasing at the rate of 1.48mm, 1.58mm, 3.17mm, and 1.46mm/year, respectively. For the whole of Nepal, total annual precipitation since the 1960s has been decreasing at the rate of 3.7 mm per year [34]. District wise, annual precipitation trend in CHAL showed a different pattern. Among 19 districts, annual average rainfall significantly decreased in 10 districts and decreased in nine districts [2], which is consistent with the present study. There is no previous comparison at the bioclimatic zone level in CHAL.
In Nepal, average annual rainfall is 1858.6 mm [35]; however, the spatial pattern of annual rainfall in the country depends on topography. Altitude further affects rainfall pattern; total annual rainfall increases with altitude up to approximately 3,000m above sea level and then diminishes at higher elevations [36]. However, in CHAL, average annual rainfall was found inconsistently increasing and decreasing from lower to higher altitudes because of mountains blocking the monsoon wind. Seasonal analysis of station-based rainfall pattern in Gandaki river basin (major part of CHAL) revealed significantly decreasing winter rainfall but increasing monsoon rainfall [16]. The precipitation was found significantly increasing in the high mountains of CHAL.
Precipitation in each bioclimatic zone in the mountains depends upon many factors such as topography, strength of moisture-bearing wind, and the orientation of the mountain range with respect to the prevailing wind direction. Thus, precipitation processes and distribution in a region is influenced by aforementioned factors, including steep altitudinal contrast [37]. Variation in elevation within a very short distance of about 185km creates dissimilarity in precipitation within particular regions in Nepal [31,38]. In this study, we found that the rainfall in LTBZ and UTBZ is decreasing annually but in LSBZ it is increasing, and again significantly decreasing in USBZ.
This unusual pattern of precipitation along different bioclimatic zones of CHAL is therefore due to orography and the spatial arrangement of topographic gradients, which control the precipitation patterns [37]. These spatial arrangements of topographic gradients, wind direction, aspects, and slopes of mountain may alter consistency in precipitation along different ecological zones within a short distance in CHAL.
In the higher elevation above 2000m, annual precipitation was found to be much lower than precipitation in lower elevations. Increases in annual precipitation from LTBZ to higher elevation up to 2000m at USBZ, and decreases in precipitation above TBZ, was well noted. The highest annual rainfall was found in USBZ in CHAL. The Lumle station in CHAL received the highest amount of annual precipitation in Nepal of about 5500mm [39,35]  According to [15], there are two high monsoonal rainfall zones in Nepal, one around 600m and another around 2100m, with different rain patterns. This record of gradual increasing precipitation from LTBZ to USBZ, with highest at around 2000m altitude, may be due to presence of high monsoonal rainfall zones. From the comprehensive precipitation observations in the Annapurna range, [40] showed that the annual precipitation gradually increases from lowland and had a strong peak at about 3000m altitude, then decreases as elevation increases. However, present analysis showed that strong peak of precipitation at 2000m then decreases as elevation increases in CHAL, which is inconsistent with the result of [40].
According to the precipitation trend analyzed from data of 80 stations in Nepal, most of the Terai area and Western Nepal observe a negative trend [41]. The hills and mountains of Western Nepal and the northern part of Eastern Nepal have a positive trend. Based on data from 1947 to 1993, [42] found that the precipitation trend in the Koshi Basin (Eastern Nepal) showed an increasing trend. The overall average trend for Nepal indicates that the annual average precipitation is decreasing at the rate of 9.8mm/decade [5]. Our findings on decade level precipitation trend in different bioclimatic zones of CHAL showed consistency with previous findings from Nepal.
According to [43], global mean surface temperature has increased on average by 0.  [2]. In the CHAL, rise in average temperature was found considerably higher than the global average.
At the district level, the highest significant positive trend was observed in Manang district (0.12 o C/yr) [2]. Average annual temperature increased in Terai (<200m), Siwalik (200-1500m), Middle mountain (1500-2500m), High mountain (2200-4000m), and High Himalayas at the rate of 0.020, 0.023, 0.031,0.032, and 0.035 o C/year, respectively, which is a similar pattern with the present study. The temperature increase was noted higher in our study than DHM (2017). This speedy warming trend in high elevation zones compared to lower altitudes is due to melting of snow and ice [44] and cold air pooling along local heating by combination of topography and synoptic conditions [45].
In Nepal, average temperature and precipitation analysis for different bioclimatic zones in particular areas has not been assessed before. However, average temperature and precipitation analysis along different physiographic regions had been done. The range of temperature in Terai and Siwalik, mid-hills, and mountain physiographic regions is 20-25, 15-20, 10-20, <3-10 o C, respectively [24,36]. [20] analyzed the variation in temperature by increasing every two hundred meters altitudes and found that temperature at <200m was 24.8 o C. The temperature at <3800m was 7.8 o C without any consistency pattern of increasing temperature in between lowest to higher elevation.
Thus, these reasonable elevation-dependent temperature and precipitation trends vary due to elevation-based differential changes in climate drivers, such as snow/ice cover, clouds, water vapor, aerosols, and soil moisture, or differential sensitivities of surface warming due to changes in these drivers at different elevations. However, mountain systems are inherently difficult to understand as a result of their complex topography, which leads to a high level of spatial and temporal variability in their climatic responses [4].
According to [29], the annual average TLR on the southern slope of the central Himalayas is −5.2 °C/km, which is −0.4 to 0.8°C/km higher than lapse rate at different bioclimatic zones of CHAL. The lapse rate of temperature depends on surface air temperature on elevation and varies due to different seasonality, interplay of radiation, slope and aspect of the mountains, and several local factors [46]. However, the precipitation lapse rate shows irregular patterns in different bioclimatic zones in CHAL. When altitudes rise from LTBZ to UTBZ, PLR shows positive trends of 611mm/500m elevation i.e. 1222mm/km. However, moving upwards from UTBZ to LSBZ, the precipitation lapse rate (PLR) shows negative rate of 778.4mm/km, and then again shows the rate of 1845mm/km up to 3000m.Then, PLR continuously decreases as altitude increases. The decreasing rate is higher at uppermost elevation (Table-5). This unusual pattern of increasing and decreasing PLR is consistent with [47] in Langtang valley within CHAL. The difference in PLR between different bioclimatic zones may be due to the local effects of topography, slope, and aspects of mountain as well as the direction of wind and many other local factors.

Conclusion
The result of the present study contributes to understanding temperature and precipitation trends in different bioclimatic zones in Nepal. There was considerable variation in precipitation and a mixed pattern from lower tropical to alpine bioclimatic zones. The average precipitation is significantly decreasing at the rate of -2.06 mm/year in USBZ, but increasing at the rate of