Author(s): Stephen D. Newman; Steven F. Daly; Lin Neifert
Linked Author(s): Steven F. Daly
Keywords: No keywords
Abstract: A variety of remote sensing techniques have been used to analyze river ice conditions and river ice jams including satellite imagery, aerial photography, and web cameras. Satellite imagery provides large scale images displaying the entire length of ice covered river reaches but often at poor spatial resolution. Recently, multiple image dates were acquired of an in-place ice cover on the Winooski River at Montpelier, VT by the QuickBird satellite. QuickBird is a high resolution commercial satellite which collects panchromatic (black & white) imagery at 60 cm resolution and multispectral imagery at 2.4 meter resolution. The images were collected on 09 March 2007 and 22 March 2007 during a period when the City of Montpelier was threatened by flooding resulting from the breakup of the ice cover and the formation of ice jams. This research seeks to examine the capability of using high resolution QuickBird imagery to classify river ice types and to determine ice cover characteristics. Multiple spatial and spectral analyses were performed to quantify the variation in ice conditions. Processing techniques included pan-sharpening each image (fusing the panchromatic band with the multispectral bands to create a high resolution multispectral image) and subsetting to the extent of the river channel banks and bridges. Texture and edge detection algorithms were used to characterize the spectral variation in rough ice conditions and classify the structural characteristics of river ice cover. Results show these techniques can be successfully used to identify variations in river ice cover characteristics and could provide an additional tool for monitoring river ice conditions in high hazard areas.
Year: 2008