Variations of pre-monsoon precipitation derived from tree rings in the central Himalayas

Abstract

The Himalayas are the longest and highest mountain system of the world with a variety of climates and abundant forest resources from tropical to alpine growth conditions. This region, being highly sensitive to the global climate change, however, is poor in instrumental climatic records in terms of quality and longevity. Tree rings have been playing an important role in reconstructing past climatic change. However, dendroclimatic studies in the Himalayas, in particular in Nepal, were limited. To date, a few tree-ring based climate reconstructions were confined to winter or spring temperature. Little was known about variations in precipitation in a long-term context in the central Himalayas. In addition, most dendrochronologial studies were based on tree-ring samples from coniferous tree species. The dendrochronological potential of broadleaf tree species was less known. In fact, Himalayan birch (Betula utilis D. Don) has a wide distribution in the Himalayas. It is also a widespread timberline species. However, we did not know whether this species has dendroclimatic potential. This study focused on dendrochronology of timberline Himalayan birch in the central Himalayas and tried to answer several questions: Whether is Himalayan birch useful to develop the long tree-ring chronology? Whether are tree rings of Himalayan birch useful for climatic reconstruction, in particular precipitation reconstruction? Whether is there a teleconnection between climate changes in the central Himalayas and other large-scale environmental events or indices? In addition, most meteorological stations are located at low elevations in the central Himalayas. It is not clear whether variations in monthly climate records at low elevations can be used to quantitatively represent those at higher elevations in the central Himalayas, and hence for dendroclimatic calibration. Focusing on the above questions, this study can be concluded as: In order to investigate climatic linkages between different elevations, we compared the climatic data representing a wide range of climatic conditions from sub-tropical (130 m a.s.l) to sub-alpine (5050 m a.s.l.) from January 2005 to December 2008. In terms of magnitude of their means and distributions, temperature and precipitation across different altitudes at varying time scales are significantly different to each other. In spite of these differences, the variations of temperature and precipitation are consistent in different altitudes and their agreement increases with lengthening time windows. Strong and significant correlation of temperature was observed between elevations [except between low–elevation (plan area) with the mid hills as well as the high Himalayas]. The slopes of the regression model (R2>0.5) indicated similar changes in temperature between different elevations. As commonly used variable for dendroclimatic calibration, the variations of monthly mean temperature records at lower elevations are better representatives to those at higher elevations. Precipitation data also showed a similar pattern, although the associations between the stations at different elevations were not as stronger as the temperature, however, significant in most of the cases. In summary, we found that it is possible to use lower-elevation monthly climate records to quantitatively assess those at higher elevations in the central Himalayas. However, variations of climatic records in the plane area cannot represent well those in mid hills and the high Himalayas. Based on Himalayan birch tree-ring samples from the Langtang national park of central Nepal, we developed a 458-year ring-width chronology. The chronology statistics showed a high dendroclimatic potential. The tree-ring growth of Himalayan birch demonstrated positive and significant (p<0.001) response to pre-monsoon precipitation. Beyond our expectation, tree rings of timberline Himalayan birch provide a rare opportunity to show variations of past pre-monsoon precipitation in the central Himalayas. However, such a climatic response needed to be tested through a large-scale tree-ring network. A large-scale of tree-ring network of timberline Himalayan birch (7 forest sites) in the central Himalayas further supported that its growth was primarily controlled by moisture stress rather than by low temperature. In particular, its growth at the timberlines is dominated by moisture availability during the pre-monsoon season, being different with tree growth at other alpine and arctic timberlines. Such a climatic response of timberline tree species is closely related to the world‟s largest elevation gradient in association with decreasing precipitation with increasing elevation (above 2000-3000 m a.s.l). On the other hand, it is related to the ecophysiological trait of Himalayan birch preferring to grow in the rain shadow. Given a wide distribution of Himalayan birch forest in the Himalayas, its timberline represents an exceptional case to investigate mechanism driving timberline formation. Based on the above tree-ring network in the central Himalayas, we developed a 460-year regional mean chronology (RC), which is the longest chronology of this species from the high Asia. Taking into account strong relationship between the RC and pre-monsoon (March-May) precipitation, we reconstructed spring precipitation back to AD 1552. This is, to date, the first climate reconstruction based on this species and the first precipitation reconstruction from the central Himalayas. The reconstructed precipitation series showed annual, multiannual to decadal variations. The years 1999, 1813 and 1954 experienced the driest springs, whereas 1775, 1557 and 1988 the wettest years. It showed the driest decades in 1811-1821 and 1995-2005. The decrease in precipitation was in phase with the unknown volcanic eruption around 1809/10 and the Tambora eruption (1815), suggesting that the subtropical and tropical volcanic eruptions may cause dry conditions in the central Himalayas. The reconstructed precipitation also showed significant correlation with March-May precipitation in Kathmandu and other large-scale regional indices that were used to represent the South Asian monsoon rainfall. This study represented a new contribution to the dendroclimatology and timberline ecology in the central Himalayas and showed insight into the variability of past precipitation and its driving forcing.