Application de la télédétection à l'analyse de la variabilité climatique des régions boréales et subarctiques du Canada et à la validation du modèle régional canadien du climat

The subject of the present thesis fits within the scope of global climate change analysis. It focuses on two main areas of interest, namely the study of recent past climate over the last few decades and the development of tools for modeling future climate. Study of past climate.The first part of this thesis deals with the interpretation of NOAA-AVHRR (National Oceanic and Atmospheric Administration - Advanced Very High Resolution Radiometer) time series at the continental spatial scale.The data constitute an archive of decadal or weekly composite spanning over two decades which was developed for studying and monitoring boreal ecosystem activity in Canada. This study uses classical parameters derived from remote sensing in the visible and thermal infrared spectra (NDVI vegetation activity index, land-surface temperature Ts), as well as indicators chosen for their robustness.The latter parameters (length of the growing season, annual growing degree-days, and ecotone displacement) were selected so as to minimize problems related to instrumental drift and inter-satellite adjustment.The study of the twenty year NOAA-AVHRR satellite archives, permitted the observation of recent climate variations. There were compared with daily meteorological surveys of temperature and precipitation as well as with the length of the snow cover period. It was possible to observe rapid cycle climatic phenomena such as the El Niäno/La Niäna couple whose effect on surface temperature and the length of the growing season was found to be more significant over the central region of Canada.The NAO (North Atlantic Oscillation) and AO (Arctic Oscillation) oscillations were also found to have an influence on the climatic regime of Canada.The effects of these climatic cycles on the annual growing degree-days are more accentuated over the Northeastern region of Canada in spring. Significant climatic trends over the last twenty years were also observed. We demonstrated, based on the use of the indicators developed in this thesis, a North/South disparity over Canada.The region north of the 55th parallel indicated a warming trend (increase in the annual degree-days, reduction in the length of the snow cover period, increase in the length of the growing season, increase in the air temperature), southern regions of Canada appeared to be cooling (based on the same indicators). Validation of the Canadian regional climate model (towards the development of future climate prediction tools).The second part of this thesis focuses on the validation, using remotely sensed measurements, of a specific surface field of the CRCM model (Canadian Regional Climate Model), namely the land-surface temperature.The comparison over a short time scale between the ground temperature values modeled by the CRCM with composited satellite temperatures shows the feasibility of validating climatic models using remote sensing.The results show a slight under-estimation of the CRCM ground temperature during the summer. This is possibly due to an overestimation of the precipitation rate which in turn generates excessive surface wetness and an excessive evaporation rate (thus cooling the surface).The agreement which was observed between model and measurements strongly suggest that climate models of the type used in this work should facilitate reliable predictions of future climate trends and help in orienting the decision making process for the world community as we collectively fact the prospect of climate imbalance