Extreme Precipitation in Nepal. Trends and Key Processes

Abstract

Nepal is located at the rim of the Himalayas and houses the highest mountains of the world. Owing to the complex topography and a seasonal monsoon climate, Nepal experiences precipitation events of considerable intensity. Large amounts of rain lead to natural hazards like landslides, floods, infrastructure damage, agricultural losses, and human casualties. It is therefore important to understand whether there are changes in extreme precipitation in Nepal, and which physical processes lead to an extreme event while taking into account the spatial variability of rainfall. To approach these questions the setup of this thesis is threefold. First, a measurement based climatology of precipitation was established and trends in extreme precipitation were detected. Second, synoptic scale conditions associated with extreme precipitation in Nepal were revealed, and third, a case study was used to proceed to the process level and obtain a better understanding of how involved processes interact and finally end in an extreme event. The first manuscript (Paper I) aims to assess the rainfall climatology and trends in extreme precipitation based on rain gauge data in Nepal from 1971-2010. Rain gauge data show that most of the annual precipitation is recorded during the Indian summer monsoon with considerable variability in time and space. Upper quantiles and annual maxima occur mainly during the Indian summer monsoon. The seasonal precipitation varies with the El Niño-Southern Oscillation (ENSO). This correlation vanishes with increasing quantiles. Trends in precipitation extremes were assessed using linear regression, quantile regression, and non-stationary extreme value theory. Moreover, parameter estimation for the non-stationary extreme value distribution was performed applying a maximum likelihood and a Bayesian approach. Multiple approaches add information regarding the method sensitivity of the trends. The study concludes that despite high spatial variability in the trends of extreme precipitation, Far-West Nepal shows a robust positive trend in extreme precipitation across the different methods. The significant changes in extreme precipitation found in Paper I urge a better understanding of the involved physical processes, which motivates the second and third manuscript (Paper II and Paper III). Paper II investigates atmospheric synoptic scale conditions and moisture sources related to extreme precipitation events in Nepal. The high spatial variability in daily rainfall was taken into account by clustering daily precipitation from rain gauges using K-means clustering. As a result, spatial patterns of daily rainfall were established dividing Nepal into West, Central, and East Nepal. The study focuses on extreme precipitation events during which the 99.5 percentile was exceeded at least at five stations at the same time in one cluster. Based on the resulting set of extreme precipitation events, a composite study was conducted for each cluster using meteorological fields from Era-Interim reanalysis. The study shows that large scale atmospheric flow was angled toward the Himalayas at the cluster location during an extreme event following mid- and upper-tropospheric trough structures in geopotential height. Tracking of low pressure systems indicates that the large scale flow conditions guided the low pressure systems toward the Himalayas where they rain out. These results show that the large scale flow conditions mainly determined the location of the extreme event. A Lagrangian moisture source diagnostic reveals anomalously abundant moisture sources over land, particularly over the Indo-Gangetic plain, along the path of the low level flow. The moisture was likely provided by foregone precipitation events over this region preconditioning the soil moisture for additional uptake. It was further found that monsoon break conditions were prevailing during 25 %-43% of all extreme events during July and August. To go deeper into the responsible physical processes and their interplay, Paper III focuses on one case, the extreme precipitation event on 19 July 2007 in Nepal. This extreme event was part of a sequence of precipitation events contributing to the South Asia flood 2007 affecting 20 million people. The study is based on rain gauge data, TRMM 3B42, Era-Interim reanalysis, Lagrangian trajectories, and a high resolution numerical simulation. The combination of these different datasets allows a multiscale analysis of the considered extreme precipitation event. The evolution of the extreme event started with individual convective cells forming over Nepal that were invigorated by moist low-level inflow with high convective available energy. The individual cells organized upscale into an intense wide convective system and resulted in torrential rain with over 250mm within 24 hours. The synoptic scale conditions were similar to Paper II, permitting and orchestrating the development of this extreme event. The following conditions were identified: prior to the extreme event precipitation events preconditioned the soil moisture along the Indo-Gangetic plain, anomalously high moisture sources were available along the path of the low level flow which was characteristic for monsoon break periods, abundant moisture sources enabled the formation of moist airmasses fueling the convective system, and the airmasses were destabilized by topographic and quasi-geostrophic forcing where the final trigger mechanism was probably the upslope flow. Besides investigating an interesting extreme precipitation event, this study shows how synoptic conditions can co-exist and interact to form a system of unusual intensity. Together, the three studies provide the basis for a comprehensive understanding of extreme precipitation events in Nepal. The interplay between atmospheric circulation and moisture sources are of particular importance. The conditions, as described in Paper II and III, have to be just right to provoke an extreme event and should therefore be usefull to increase the ability of forecasting an extreme event. Challenges resulting from the pronounced changes in extreme precipitation (Paper I) can be approached supported by the conditions found in Paper II and III. The involved processes can be persued in future studies to gain further insights which will hopefully foster new research and useful findings for Nepal.