Preview

image preview

Faculty Mentor

Christopher Badurek

Creation Date

4-2022

Document Type

Open Access Document

Abstract

This poster reports on supervised image classification performed on archived UAS data of the Plymouth, NH ice jam of February 2017. Ice jams are a costly occurrence as noted by the USACE Ice Jam Database (2020) and referred to as ‘nuisance flooding’ during spring thaws. Data collected with a DJI Inspire 1 with FC550 sensor over four miles of the Pemigewasset River resulted in a final 3 cm resolution mosaic image. We accessed the data set from USGS Earth Explorer and processed it with the free and open source GIS application Quantum GIS (QGIS). First, we classify the image using a training data set and the maximum likelihood algorithm. We next determine the extent of land cover classes to differentiate ice, snow, mixed snow-water, and water classes. Finally, we perform an accuracy assessment using a random sample to determine accuracy of the land cover map of the ice jam area.

Keywords

Ice Jams, supervised image classification, Pemigewasset River, QGIS, Transformations

Share

 
COinS