Monitoring subarctic environments using x- and c-band radar imagery.

Duguay, Yannick; Bernier, Monique; Dominé, Florent et Lévesque, Esther (2013). Monitoring subarctic environments using x- and c-band radar imagery. In: ArcticNet Annual Scientific meeting - ASM 2013, 9-13 décembre 2013, Halifax, Canada.

Prévisualisation

PDF Télécharger (6MB) | Prévisualisation

Résumé

Subarctic environments are signifi cantly aff ected by the climate warming observed in recent decades. Th ese perturbations lead to many environmental changes such as permafrost thawing and expansion of shrub vegetation cover. Snow, with its insulating properties, also plays an important role in these processes maintaining relatively warm ground temperatures during the winter while protecting the vegetation from the cold and wind. Th e presence of shrubs may, in turn, trap wind blown snow and creates a positive feedback favoring the expansion of shrub vegetation at the expense of the tundra. Th e objective of this study is to develop methods for monitoring simultaneous subarctic environments of these two elements to ensure eff ective monitoring of subarctic environments. Satellite Synthetic Aperture Radar (SAR) allow, among other things, provide information on the vertical structure of the observed objects and are considered here to estimate the height of the vegetation and snow cover thickness. Th e study area is a 60 km2 region situated around the Umiujaq community (56.55° N, 76.55° W) in northern Quebec, Canada. Th e area can be divided into two distinct environments: the coastal region to the east and the Lac Guillaume-Delisle graben to the west. Th e vegetation in the coastal region is very sporadic and dominated by tundra vegetation, while the graben vegetation is mainly shrublands with patches of conifers. A series of polarimetric RADARSAT-2 C-band images (HH, VV, HV, VH polarizations) and dualpolarized TerraSAR-X X-band images (HH, HV polarizations) have been acquired over the area between October 2011 and April 2012 during the fall and winter seasons. Field measurement campaigns where performed during the summer of 2009 and the winter of 2012 to collect data on the vegetation and snow characteristics respectively. Temperature and soil moisture sensors were also installed at 6 selected sampling sites. In-situ observations have shown that the height of the shrub vegetation infl uences the depth of the snow cover. Preliminary results show an increase in the RADARSAT-2 backscattering with vegetation height while the TerraSAR-X backscattering seems to saturate with higher vegetation. Th e fall images, representing the baseline without snow cover, are then compared with the winter images to evaluate the eff ect of snow cover on the SAR signal. In the presence of snow, RADARSAT-2 signal is attenuated while the TerraSAR-X signal increases slightly. Th e relationship between snow depth and radar parameters is relatively weak. A fi rst classifi cation with and unsupervised Wishart method was performed using the polarimetric RADARSAT-2 data to delineate vegetated areas and allowed to classify three types of vegetation. Given the correlation between the heights of vegetation and snow cover, improved segmentation based on the height of the shrub cover should allow a better estimation of the snow cover parameters.

Type de document:	Document issu d'une conférence ou d'un atelier
Informations complémentaires:	Affiche
Mots-clés libres:	environnement subarctique; végétation; neige; imagerie radar; surveillance
Centre:	Centre Eau Terre Environnement
Date de dépôt:	12 nov. 2020 14:59
Dernière modification:	12 nov. 2020 14:59
URI:	https://espace.inrs.ca/id/eprint/4051

Gestion Actions (Identification requise)

Modifier la notice