Abstract
In topographically complex mountain environments, such as the High Himalayas, knowledge of spatial and altitudinal variations of orographic rainfall and surface water distribution is essential to understanding the hydroclimatic processes that translate into natural hazards and impact the local ecology and land use. These facets of the Earth’s surface system are all exposed to the effects of rapid climate change.
Since instrumental records are rare in the High Himalayas, I integrate precipitation data derived from a dense rain-gauge network with hydroclimate-sensitive proxy: Oxygen stable isotope ratios in wood cellulose of tree-rings provide valuable constraints on the local climatic and hydrologic conditions over decades to centuries.
After a general introduction, I investigate the spatiotemporal pattern of precipitation in 24 trans-Himalayan catchments in Chapter II of this thesis. For this purpose, I apply an objective statistical method to determine onset and withdrawal of the monsoon and to produce a 65 year long seasonal precipitation time series from the gridded APHRODITE precipitation product. Defining the orographic barrier as the 500 mm/yr isohyet, the Himalayan domain is divided into catchments draining the northern and southern flanks of the mountain belt. I show that these catchments can be grouped into three geographical clusters, located in the western, central and eastern part of the Himalayan arc, and each characterized by consistent and distinct spatial and temporal precipitation dynamics. Although the monsoon season dominates the annual precipitation budget throughout the region, clusters are exposed to distinct long-term influence dominated by different seasons, with different epoch of occurrence. In the East, I observe a strong monsoon decline in the south of the range, continuous for the full record, and a moderate pre-monsoon increase north of the range, starting in the late 1960s. In the Center, a substantial pre-monsoon increase and moderate winter increase occurred mainly from 1950s to 1980s, on both side of the range. In the West, the southern flanks are stable while the northern have experienced a substantial decline of mainly winter and monsoon that have occurred near the start of the record. These behaviours are discussed in terms of the various sources of moisture and the synoptic patterns that may drive the different clusters. A focus on the processes driving these clusters as well as on the moisture source of winter and pre-monsoon precipitation, which have been largely disregarded in the past, is a key to better understanding of the climate change and its effects in the Himalayas.
In Chapter III, I use the ratios of water stable isotopes recorded in tree ring cellulose to obtain information about moisture sources and precipitation intensity in the Kali Gandaki catchment within the central cluster defined in Chapter II. Tree ring δ18O records were established for two sites along the upper Kali Gandaki valley, one in each of two distinct physiographic Himalayan regions: the wet High-Himalayas and the Trans-Himalayan dryland to the north. I used precipitation data from the APHRODITE network as well as highly time-resolved monitoring data (relative humidity, temperature, rainfall δ18O, and radial tree growth) from sensors I deployed along the valley from 2015 to 2017. Empirical correlation and regression analyses were combined with mechanistic model simulations to explore the sensitivity of tree ring δ18O to the sub- and inter-annual specifics of the hydroclimatic forcing. I find that the onset of monsoon conditions represents an abrupt change in the local precipitation regime and its isotopic signature, and that the tree vegetative period straddles this transition. As a result trees recruit water from both pre-monsoonal and monsoonal inputs, which are integrated into a compound δ18O in the cellulose of High-Himalayan trees. In contrast, the δ18O records of trees at the Trans-Himalayan site appear to mainly register monsoon inputs. The difference in sensitivity of trees at these two sites offers an opportunity to reconstruct the hydroclimatic seasonality over the Himalayan orographic region, with an emphasis not only on monsoon strength, but also on the influence of the pre-monsoon season.
In Chapter IV, I explore the spatio-temporal variability of the tree ring δ18O records in the same region by studying the properties of annual anomalies as well as decadal trends at the two sites presented in Chapter III. For this purpose, I consider the full record length (80 to 115 years) and add data from a site located between the two presented in the previous chapter. Detailed comparison of these three closely spaced (~40 km) sites gives insight into the connectivity between the monsoon-dominated High Himalayas and the high-elevated cold and arid Tibetan plateau, across the orographic barrier. I show that the three sites share δ18O annual anomalies, correlated with large-scale climatic indices such as the Monsoon Index and the El Nino Southern Oscillation Index. However, strong discrepancies exist between the decadal fluctuations of the three records, especially in the period 1940-1980. This interval coincides with global temperature dimming. Moreover, trends at the southern and northern sites, though synchronous, are reverse. This suggests that, although monsoon moisture is channeled through the Kali Gandaki valley, by translating the year-to-year moisture availability into a precipitation signal in Southern Tibet, opposite climatic regimes dominate over decadal scales in front and behind the Himalayan orographic barrier. This may have important implications for the way in which we combine and interpret long-term isotopic records from neighboring sites that are influenced by different climate regimes, for example across an orographic barrier.
In Chapter V, I summarize the major observations and findings that I collected during this work and discuss two major conclusions, which have consistently emerged from the three reported studies. One is the unexpected finding of an influential role of the pre-monsoon season in setting the hydroclimatic boundary condition in the Himalayas. The other is the spatial compartmentalization of hydroclimatic regimes across and along the Himalayas.
In conclusion, this study provides original insights on the temporal and spatial evolution of seasonal precipitation over the Himalayas under changing climate, with strong implications for water resource management and paleo-proxy interpretation.