Derivation of River Channel Morphology Information from Time Series Satellite Images
Topics:
Keywords: Floods, Satellite Image, Phase-code Algorithm, Machine Learning
Abstract Type: Paper Abstract
Authors:
Pawan Thapa, University of Alabama
Hongxing Liu, University of Alabama
Lei Wang, Louisiana State University
Rupesh Bhandari, University of Alabama
,
,
,
,
,
,
Abstract
River channel morphological attributes are essential for estimating stream discharge, hydraulic modeling, and predicting floods. This research presents and evaluates an efficient method for extracting and quantifying river channels from time series image data from multiple satellite systems. A machine-learning-based region-growing algorithm is implemented to extract river channels from multispectral satellite images, a phase-coded disk algorithm is implemented to derive river width, central river line, and other attributes, and a dynamic segmentation data model is utilized to represent the topological and morphological information of river channel network. Our method will be tested and evaluated for rivers in the Mobile River Basin using time series Landsat-8/9 and Sentinal-2A/B image data.
Derivation of River Channel Morphology Information from Time Series Satellite Images
Category
Paper Abstract
