Thèse de doctorat en Géophysique et géochimie
Sous la direction de Claude Boutron.
Soutenue en 1995
à Grenoble 1 , en partenariat avec Laboratoire de glaciologie et géophysique de l'environnement (Grenoble) (laboratoire) .
Le président du jury était Laurent Charlet.
Les rapporteurs étaient Freddy Adams, Kevin Rosman.
It is now weIl recognized that the huge amounts of metal discharge from industrial emissions into the air, water and soils have no doubt resulted in the great impact on the regional and global cycling of trace metaIs since the Industrial Revolution [Nriagu, 1990]. To evaluate the impact of anthropogenic emissions on the natural cycles oftoxic heavy metals, it is needed to understand a general historical trend in occurrence of these elements in the atmospheric system which is a main route of the transport of pollutant aerosols emitted from various anthropogenic sources. This thesis consists of eight components in the form of articles (1-VIll) which present all the analytical procedures and resnlts obtained by analysing fourty seven ice core sections from GRIP deep ice core. In 1990-92, the multinational European Greenland Ice-core Project (GRIP) drilled an ice core at Summit (72° 58'N, 37° 64W, elevation of 3238m above sea level, annual mean air temperature -32 °C) in central Greenland. The 3028. 8-m-Iong core was recovered by an electromechanical drill, which covered the past 250 kyr according to the time scale given from Dansgaard et al. . The author analysed for Pb, Cu, Zn, and Cd fourty seven ice core sections which cover the Holocene period " 10,000 years B. P. ) and the last full glacial cycle (10,000 - 150,000 years B. P. ). As many other investigations were conducted on the same core sections, only a limited part of the core cross section was available for this study. Each core section is 55cm long and thus the time span of the accumulation of ice ranges from three years for the shallow depth to about 200 years for the deep one. This allows us to obtain an average concentration value of more than one year of accumulation of ice in each core section for eliminating a possible influence of seasonal variation in concentration. The time scale used here was establisbed by counting annuallayers back to 14. 5 kyr B. P. And beyond that by ice flow modelling [Johnsen et al. , 1992; Dansgaard et al. , 1993].