Describing and detecting changes at various scales in geomorphological features: the example of the Rhone river bathymetry
Abstract
Wavelet-based decomposition is increasingly used in geosciences to analyze time series. Its use for spatial series is sparse, although it can be particularly useful to describe longitudinal structures such as river profiles. We use bathymetric data of the Rhone River (Southern France) to illustrate 1) the intricate nature of this kind of geomorphological signal in terms of scales interlocking 2) the challenges associated with the interpretation of its scale-specific subsignals 3) the insight provided by some associated statistical methods to interpret these subsignals. We design a statistical method to analyze the scale-specific subsignals. This method consists in an unsupervised segmentation of the signal into portions with homogeneous variance. The significance of the changes between adjacent portions is then tested. We carry out a simulation study to quantify the power of this segmentation algorithm and associated test to detect scale-specific changes, according to, e.g., rate of change, series length, changes at close scales, etc. The segmentation and testing method is applied to the bathymetric data of the Rhone River, and its results are discussed both from a geomorphological perspective and based on the results of the simulation study.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFMEP43A0817V
- Keywords:
-
- 1805 HYDROLOGY Computational hydrology;
- 1808 HYDROLOGY Dams;
- 1984 INFORMATICS Statistical methods: Descriptive;
- 1986 INFORMATICS Statistical methods: Inferential