Activity Number:
|
577
- Statistical Models in Ecology
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistics and the Environment
|
Abstract #307320
|
|
Title:
|
A Time Series Clustering Approach for Classification of Intermittent Streams
|
Author(s):
|
Claudio Fuentes* and Jeffrey Mintz and Xiaohui Chang and James Molyneux and Ivan Arismendi
|
Companies:
|
Oregon State University and Oregon State University and Oregon State University and Oregon State University and Oregon State University
|
Keywords:
|
Clustering;
Classification;
Time Series;
Shape Similarity;
Error Rate
|
Abstract:
|
Intermittent/ephemeral streams characterize more than half of the length of the Earth's river networks. Freshwater ecosystems from arid or semi-arid landscapes are especially vulnerable to changes in human-related water use as well as shifts in terrestrial climates. Yet, the description and quantification of patterns of flow permanence in these systems are limited due to difficulties in instrumentation and detection. In this paper, we use hydroclimatic time series from multiple locations at two watersheds in the Northwest U.S. to develop a straightforward approach to classify flow permanence in stream reaches based on time series of stream temperature. By exploiting the qualitative differences in daily temperature patterns, we propose a method in which we weigh the similarity of time series to wet and dry cluster centers. While the cluster centers are determined in a supervised fashion, the results presented here can be easily extended to situations where the ground truth is unknown. We provide examples of how to combine different assessments of shape similarity in order to improve classification error rates.
|
Authors who are presenting talks have a * after their name.