Context

Climate change in Tibet may have a potentially significant impact due to the role played by Tibetan glaciers in terms of large river hydrology, irrigation water availability and water use by local populations. Anthropic greenhouse gas emissions could cause a rapid retreat of a significant portion of these glaciers, with negative consequences for water resources. However, the glacier water balance is influenced not only by local temperature variations, but also by precipitation variations. In southern Tibet, precipitation events exhibit a strong interannual-to-decadal variability directly related to the variability of Indian monsoon rainfall. Within the framework of a thesis project co-sponsored by LSCE and the Botany Institute of Beijing, Chunming Shi has studied the potential of an isotopic analysis of high-altitude forest tree rings in south-east Tibet. The first results obtained are very promising, with good correlation between summer precipitation and ¹⁸O concentrations in cellulose over the past 50 years. 

Objectives

These encouraging results led to the creation of the PLUIES-TIBET project. This project aims to:
• analyse interannual, decadal and centennial precipitation variability in south-east Tibet based on tree data, precipitation isotope composition records and/or climate variables;
• assess the accuracy of simulations of precipitation and forest response (in terms of productivity) to climate variability in the region.

Methodology

Geographical framework

Given the quantity of material necessary to conduct isotopic analyses, the natural reserve of Gangcun (Chayu valley, south-east Tibet) has been chosen as the pilot site. This site allows us to study a relatively homogeneous set of trees (pine) exhibiting rapid growth, with annual tree ring widths sufficiently large to conduct annual analyses, sometimes on a tree-by-tree basis. In addition, this site is located near three meteorological stations (Bomi, Linzhi et Chayu) providing meteorological records since the 1960's, therefore allowing us to compare dendro-isotopic measurements with local and regional climate variability data. The first results obtained within the scope of Chunming Shi's thesis work concerned the isotopic analysis of annual tree ring samples (from 1960 to today) taken from five homogeneous-growth trees within the site.

Location of sampling sites and three meteorological stations (Bomi, Linzhi and Chayu).

Research phases

Phase 1: Reconstruction of precipitation history in Chayu valley (south-east Tibet)

1) Pursuit of paleoclimatic reconstruction efforts

Initially, we will pursue paleoclimatic reconstruction efforts using the oldest trees (probably 250 to 300 years old) already sampled in the natural reserve of Gangcun. Chunming Shi's thesis work will first entail dating these trees and determining a regional growth index (work conducted in Beijing) and, subsequently, tree ring cutting and sampling, and cellulose extraction and isotopic analysis (performed at LSCE). The pilot study has shown that five trees constitute a sufficient population to avoid inter-tree isotopic variability. The ¹⁸O isotope analysis of these samples will be supplemented with a carbon isotope composition analysis so as to obtain a second marker of the isotopic functioning of trees, more directly influenced by photosynthesis processes and stomata openings. This second parameter is particularly important, since it gives access to carbon isotope discrimination and tree water use efficiency variables that can be compared with the results of vegetation functioning models (ORCHIDEE model developed at IPSL).

2) Comparison of empirical data with the results of vegetation models

Shilong Piao (Ecology Department, Beijing University) is currently developing modifications of the ORCHIDEE model specifically adapted for the high-altitude plain vegetation of the Tibetan plateau. A joint work meeting will be scheduled to compare vegetation productivity and isotope discrimination simulations with wood growth and discrimination measurements taken in the Chanyu site. Local meteorological data will be provided to force the ORCHIDEE model. The comparison of simulation results with dendro-isotopic measurements will allow us to characterise the ecophysiological functioning of trees and the key processes involved in discrimination variability. These comparisons should lead to better models of atmospheric carbon dioxide content and local forest reaction to climate variability.

3) Comparison of the ¹⁸O isotope composition of tree ring samples and ice cores from Himalayan glaciers

Cellulose oxygen isotope composition data covering several centuries will be compared with isotope composition data from Himalayan glaciers and with estimates of past variations in accumulation rate based on ice core analyses (Yao Tandong database from several coring campaigns in the Himalayas). This will allow two very different records (tree ring and ice core data) from the same region, but at different altitudes, to be compared for the first time.

Ice core analyses used to determine past climate conditions.
Source: E. Lefebvre, Glaciology and Environmental Geophysics Laboratory (LGGE). 

4)  Assessment of precipitation variability based on historic data

French and European Tibetologists (CNRS Himalayas Research Group in particular) will collect historic climate data. Biographies, which are sometimes highly detailed in this region, should provide the most useful source of information. Precipitation variability could be described in terms of floods, droughts or famines. Climate reconstructions will be compared with this historic data.

Phase 2: Analysis of variations in local precipitation isotope composition and their relation with climate variability

This second phase takes advantage of the opportunity to compare tree ring data with precipitation sample measurements. For several years now, the Tibetan Plateau Research Institute has collected precipitation, river and lake samples, as well as glacier core samples from south-east Tibet. At present, this work is conducted within the scope of the doctoral thesis project of Gao Jin, a Chinese student whose thesis is supervised by Yao Tandong, director of said institute.

1) Isotopic analysis of precipitation samples

Within the framework of the PLUIES-TIBET project, there are plans for Gao Jin to visit the LSCE laboratory for 3 months so as to take accurate measurements of the isotope composition of precipitation samples. These measurements will be extremely useful for Jin's thesis and, for the purposes of the present project, they will improve our understanding of the acquisition of precipitation and tree-ring isotope data.
The spatial and temporal variability of precipitation isotope composition (intra and interannual) will be interpreted in terms of climate variability and precipitation origin (using retro-trajectory calculations based on atmospheric analyses).

2) Modelling of temperature and precipitation in the Tibet region

The goal here will be to supplement the observations collected during phase 1 with a model-based approach. Simulation results obtained using the LMDZ model (general atmospheric circulation model developed by IPSL) and the associated precipitation isotope module (LMDZ-isotopes) will be compared to the meteorological and isotopic data available for the Tibet region.
After this evaluation phase, the LMDZ-isotopes module will be used to study relationships between precipitation origin, temperature, rainfall, relative humidity, and isotope composition of water vapour, precipitation and soil water. These different variables will be combined to represent the tree-ring cellulose isotope composition expected as per available knowledge of isotopic fractionation in trees. This approach should allow us to understand the close relationship already observed between cellulose isotope composition variability and summer precipitation.
The local precipitation variability will also be compared with the larger-scale climate variability, particularly to test the local signature of Indian summer monsoon precipitation. This analysis will be conducted using the LMDZ model and by means of statistical studies based on local meteorological data and monsoon indexes.
To conclude, the current variability of local precipitation will be interpreted in terms of past variability, via reconstructions (conducted during phase 1) and long-term simulations (IPCC simulations starting in 1850), and in terms of future variability (IPCC simulations). This will allow us to compare the precipitation variability spectrum derived from instrumental data with that derived from climate models, to test the accuracy of interannual-to-centennial variability simulations, and to interpret potential climate change in terms of past data.