A Study of Daily Rainfall Pattern and Its Changes in Kathmandu Valley Nepal during Summer Monsoon Season
Department of Hydrology and Meteorology, Government of Nepal
Email: email@example.com, firstname.lastname@example.org
The daily rainfall pattern over Kathmandu valley was investigated on the basis of observed rainfall data from 6 stations for the period of 1971-2005. The rainfall data of the south west monsoon season, which accounts to about 80% of the annual rainfall, was used in this study.
Heavy rain events with rain rate > 30 mm/day were found to be more frequent on mountain slopes compared to the valley floor while light rain events with rain rate <10 mm/day were found more frequent in valley floor relative to that of mountain slopes.
Analysis of the change in daily rainfall pattern over the study period revealed a significant rising trend in days with rain rate 0.1- 10 mm/day; however, the total monsoon rainfall remained unchanged. The significance of this trend was tested by using Mann Kendall Method, and is postulated to result from an increasing number of condensation nuclei due to urban air pollution in Kathmandu valley.
Key words: daily rainfall pattern, condensation nuclei, Mann Kendall method
Rainfall is the dominant meteo-climatic feature of the Kathmandu valley. Most studies in Nepal are based on monthly, seasonal and annual rainfall data derived from daily rainfall data focused outside Kathmandu valley. However, information on daily rainfall distribution is of relevance for the efficient management of water resources, as well as for the better understanding of the processes producing rainfall. In addition, studies on changes in daily rainfall over the urban valley are important to differentiate the global climate change impact on rainfall from local effect due to urbanization.
A study conducted by Shrestha et al. (2000) did not find a distinct long term trend in total annual precipitation in Nepal, whereas trend analysis in daily climatic extremes of precipitation in Nepal carried out by Baidya et al. (2008) found an increasing trend in heavy precipitation events in most of the stations but no significant pattern in trend was observed in extremely wet days (annual total precipitation greater than 99th percentile). These studies, and their findings, stressed the need to study the change in daily rainfall pattern, and the findings are significant because a change in daily rainfall pattern would have a subsequent effect on urban flooding.
This study investigates rainfall patterns and trends in the Kathmandu valley on a daily timescale over a three decade period. The focus is on the south west monsoon season (June – September) because it accounts to about 80% of annual rainfall over most part of the country, yet little is known regarding the daily rainfall pattern during this season or how it may have changed through time in Kathmandu valley. The goals are to assess 1) the frequency of rainy events and rainfall accumulated by events of various intensity during monsoon season and 2) change in daily rainfall patterns during the monsoon season. The significance of temporal trends of rainfall characteristics over Kathmandu valley was tested using the Mann Kendall method, a widely used method to test the significance of monotonic trend in time series.
2. Study Area and Precipitation Climatology during Monsoon Season:
Kathmandu, the capital city of Nepal, fills the flat Kathmandu valley which lies between1300 to 1400 m.a.s.l., and is surrounded by hills ranging from 2000 to 2750 m.a.s.l. in elevation: the Himalaya to the north and the Mahabharata mountains to the south.
In this section, a brief overview of what is known about the monsoon precipitation patterns across the watershed is presented, prior to introducing the current study. In Kathmandu valley 80 % of total rainfall occurs during the monsoon season (JJAS). Over the study period, the highest rainfall amounts are observed in the Northwestern part of valley around the periphery of Kakani with more than 2300mm of rain falling in the four monsoon months. The second highest rainfall was observed in the Northern part of the valley around the periphery of Sundarijal and Sankhu with more than 1700 mm of monsoon rainfall.
Pockets of notably high monsoonal rainfall were also noted in the mountainous eastern, western and southeastern portions of the catchment, with monsoon rainfall of more than 1500 mm at the Nagarkot, Thankot and Godavari stations respectively. The lowest rainfall was observed on the southern side of valley floor at the periphery of Khumaltar with monsoon rainfall averaging < 1000 mm. Fig-1 clearly shows the spatial variation of precipitation in Kathmandu valley during monsoon season.
Rainfall in the whole Kathmandu valley showed a marked increase with elevation. This relationship is more pronounced in the northern mountains than the other surrounding mountains. The mean value of rainfall during this season was 1462 mm with the value ranging from 926 mm at Khumaltar to 2366 mm at Kakani. The main cause for the highest monsoonal rainfall occurring in the northwestern mountains is the moisture laden air entering from the southwestern side of valley and traveling towards the northern mountain and the lowest rainfall in the southern part of the valley floor is due to subsidence of air. (Karki, 2008)
3. Materials and Methods:
Mann-Kendall tests are non-parametric tests for the detection of monotonic trend in a time series. These tests are widely used in environmental science, because they are simple, robust and can cope with missing values and values below a detection limit.
Let Y be rainfall characteristics and X year then
H0: T = 0, No correlation exist between X and Y
And Alternative Hypothesis,
H1 : T ≠ 0, X and Y are correlated.
Then S = A-B
Where A= Number of cases when Yi<Yj for i<j B= Number of cases when Yi>Yj for i<j
T =, where n is number of year
The variable T is normally distributed with mean zero and has a variance
sz2 = ,
ZT = T * ( )1/2
Where ZT is the standard Normal variate.
The null hypothesis is rejected at significance level a if /ZT/ > Za/2, and the trend is significant. In the case of significant positive trend ZT> Za/2, and in case of significant negative trend ZT< Za/2 .
4. Result and Discussions
4.1 Distribution Pattern of Daily Rainfall during Monsoon Season:
Based on the data, two distinct types of daily rainfall distributions (Mountain slopes and Valley floor) were found in Kathmandu valley during monsoon season.
During the southwest monsoon season, the rainfall amount was larger in the mountain stations than in the central part of valley floor with average monsoon value of about 1978 mm and 1151 mm in the mountain side and valley floor (KathmanduAirport) respectively. The average number of rainy days during the monsoon season in all the stations lasted about 90 days, except at Kakani, which had 100 rainy days.
The amount of rainfall contributed in percentage (% of total monsoon rainfall) by various precipitation rate classes (i.e., 0.1-10, 10-20, 20-30, 30-40,40-60, 60-90, and greater than 90 mm/day) for various stations of Kathmandu valley was calculated and categorized into mountain and valley floor. Similarly, frequency of rainy days in percentage (% of total number of rainy days in monsoon season) by various classes of precipitation rate was also calculated and categorized (Fig 3 and Fig 4).
While about 51% [mountain (41%) and valley floor (61%)] of total rainy events had precipitation rate less than 10 mm/day, these events contributed only 13% [mountain (9%) and valley floor (17%)] of the total rainfall.
On the other hand, the frequency of heavy rainfall events (30-90 mm/day) dropped sharply because of low chances of the event. However, the rainfall contributed from these rainfall events was about 46% [mountain (53%) and valley floor (39%)] with corresponding frequency of only 17% [mountain (23%) and valley floor (10%)] of the total number of rainy days. Extremely heavy rainfall events > 90 mm/day were very infrequent with less than 2% of the total occurrence. These events happen about once in one summer and the contribution was averaged 6% of the total rainfall. The percentage of total rainfall contributed and the frequency of rainy days on the category (10-30 mm/day) was almost equal (30-35%).
These data clearly show that about half of all rainy days belonged to precipitation rate less than 10 mm/day but the contribution of these events to the total monsoon rainfall was only about 15%. Generally, half of the total monsoon rainfall was contributed by rain rate 30-90 mm/day whereas frequency of these rainfall events was only about 15% of the total number of rainy days.
Higher frequency of heavy rainfall events (30-90 mm/day) and fewer light rain (<10 mm/day) events at the mountain stations relative to that of valley floor led to the high amount of rainfall on the mountain side compared to valley floor and this is due to the orographic effect on rainfall.
4.2 Study of Change in Total Monsoon Rainfall through time:
The time series was separated into 1971-1989 and 1990-2005 to assess the change in total monsoon rainfall. The difference of monsoonal rainfall between the two periods (1971- 989 and 1990-2005), for the stations of Kathmandu valley indicated rainfall enhancement of 106 mm, 109 mm, 160 mm, and 165 mm in the Sankhu, Kathmandu Airport, Nagarkot and Kakani stations respectively but there was reduction in rainfall by 207 mm and 108 mm at Thankot and Godavari respectively in the later period (1990-2005) compared to the earlier period (1971-1989).
The Mann Kendall method was performed to test the significance of monotonic trends in total monsoon rainfall in all 6 stations (1971-2005). The test performed at 5 % significance level showed no trend in 4 stations Kakani, Sankhu, KathmanduAirport and Godavari, but a falling trend in Thankot and a rising trend in Nagarkot were significant. This study did not find evidence of change in total monsoon rainfall in Kathmandu valley.
4.3 Study of Change in Daily Rainfall rate in the Kathmandu Valley during Monsoon Season:
The difference (average of 1990-2005 minus average of 1971-1989) of accumulative rainfall from each class of precipitation rate was calculated to investigate the change in rainfall amount (mm) as a function of precipitation rate in Kathmandu valley and is shown in Fig-5.
The accumulated rainfall in the categories 0.1-10, 40-60 and >90 mm/day has increased in the decades since 1990 in all the stations of valley except Thankot. Similarly, there was decrease in accumulated rainfall in the categories 60-90 mm/day in all the stations of valley except Sankhu in the recent decades. The change in rainfall amount in other categories varies from station to station. The Mann-Kendall method was again employed to test the significance of monotonic trends in whole time series (1971-2005) to the above change in days and accumulated rainfall for the various threshold of precipitation during the monsoon season. The test was performed at 5% significance level and the result showed rising trends in number of rainy days in the category 0.1-10 mm/day in all the stations but the significance is lower on valley floor compared to mountain side stations. Similarly, the rising trend in accumulated precipitation in the categories 40-60 mm/day and >90 mm/day and falling trend in the categories 60-90 mm/day were significant in some stations of valley but trends in the valley-wide average time series of the above categories were not significant. Only the overall rising trend in days with rainfall rate less than 10 mm/day was statistically significant (Fig-6). This rising trend in turn has led to an increase in total number of rainy days. However totals monsoon rainfall remained unchanged as the amount of rainfall contributed by these classes of precipitation to total monsoon rainfall was small. The test statistics for Mann Kendall method is presented on Table 2. The rising trend in days with precipitation rate less than 10 mm/day might be due to increasing number of condensation nuclei in air as a result of urban air pollution but further similar type of study on non-urban environment is necessary to verify this hypothesis.
Daily rainfall analysis over Kathmandu valley during monsoon season, clearly showed that light rains (<10 mm/day) were more frequent in the valley floor relative to that of mountain side whereas, heavy rain rate (>30 mm/day) were more frequent in mountain side relative to the valley floor. Change in daily rainfall pattern has been also studied and the significance of temporal trends was tested using the Mann Kendall method. The result indicated significant rising trends in days with precipitation rate 0.1-10 mm/day however total monsoon rainfall remained unchanged as the amount of rainfall contributed by these threshold to total monsoon rainfall was marginal. The significant rising trend in the number of rainy days with precipitation rate 0.1-10 mm/day might be due to the increasing number of condensation nuclei in the air as a result of urban air pollution.
However, the causes for the change in the threshold of precipitation on decadal scale in the region are very complicated and they may involve the natural variability, anthropogenic change and localized effects as well as interaction between them. So, detailed analysis of the change with more data and similar type of study on non-urban environment is necessary.
The author wishes to thank the Department of Hydrology and Meteorology, Nepal for providing the rainfall data of Kathmandu valley. The author would also like to express his sincere thanks to Prof. Lochan Prasad Devkota, Central Department of Hydrology and Meteorology, Tribhuvan University, Nepal, Dr. Lindsey Nicholson, Institute of Geography, University of Innsbruck, Austria, Saraju Kumar Baidya, Deputy Director General of Department of Hydrology and Meteorology, Nepal for their valuable suggestions and comments during the preparation of this paper.The author thanks anonymous reviewers for their valuable comments that have helped to revise the paper.
1. S.K. Baidya, M.L. Shrestha, M.M. Sheikh, Trends in Daily Climatic Extremes of Temperature and Precipitation in Nepal, Journal of Hydrology and Meteorology, 5, 2008
2. Intergovernmental panel on climate change (IPCC) Fourth assessment Report, Climate change 2007, The physical Science Basic, 2007
3. R.C. Karki, Study on rainfall pattern over Kathmandu valley during summer monsoon season and its long term change, a dissertation submitted to Central department of Hydrology and Meteorology, Tribhuvan university unpublished Master’s thesis 2008
4. M. G. Kendall, Rank Correlation Methods, Griffin, London, 1955
5. A.B. Shrestha, C.P. Wake, J.E. Dibb, A.P. Mayewski, Precipitation fluctuations in the Nepal Himalaya and its vicinity and relationship with some large scale climatological parameters, International Journal of Climatolology, 20, 2000