Evaluation of thermal-water stress of forest in southern Québec from satellite images
|Advisor:||Viau, Alain A.; Anctil, François|
|Abstract:||The thermal-water stress of the vegetation canopy was evaluated for southern region Québec from SPOT VEGETATION (VGT) and NOAA AVHRR images, and for the 1999 and 2000 reproduction seasons. To retrieve surface temperatures from AVHRR images, the algorithm of Coll et al. (1994b) was found to be the optimal method for our study area by comparing six algorithms. Artificial Neural Networks (ANNs) were trained for cloud detection on daily synthesis (S1) and P data of the SPOT VGT system. The analysis demonstrated the superior classification of the network over the standard cloud masks provided with the data. The network detected not only bright thick clouds but also thin or darker clouds. As another application study, ANNs were employed for estimating air temperatures. The input variables for the networks were the five bands of the AVHRR image, surface altitude, the solar zenith angle, and Julian day. The network using all the input data provided the best results, with 22 nodes in the hidden layer. The Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI) were derived from SPOT VGT for evaluating vegetation water status and surface temperature was retrieved from AVHRR for the thermal status. The two vegetation indices were integrated for evaluating the vegetation condition and water status as a new index, namely the Normalized Moisture Index (NMI). A trapezoid was defined by the NMI and by surface temperature and the thermal-water status of the vegetation canopy was determined according to the four sectors of the trapezoid. The thermal-water status was validated by comparing it with the indices of the Canadian Forest Fire Weather Index System.|
|Document Type:||Thèse de doctorat|
|Open Access Date:||11 April 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.