Grouping river flow patterns based on nonlinear features of short time series data

River flows provide information on the availability of water and its variability in space and time. Identifying similar grouping of river flow patterns is useful so that the knowledge of homogeneous sites can be used to manage water resources more efficiently. However, the grouping of river flow is...

Full description

Bibliographic Details
Main Authors: Mansor, Nur Syazwin, Ahmad, Norhaiza
Format: Conference or Workshop Item
Published: 2023
Subjects:
_version_ 1817854218804920320
author Mansor, Nur Syazwin
Ahmad, Norhaiza
author_facet Mansor, Nur Syazwin
Ahmad, Norhaiza
author_sort Mansor, Nur Syazwin
collection ePrints
description River flows provide information on the availability of water and its variability in space and time. Identifying similar grouping of river flow patterns is useful so that the knowledge of homogeneous sites can be used to manage water resources more efficiently. However, the grouping of river flow is not easy because river flow can contain several regimes to account for different behaviours. In this study we use a two-stage approach to identify groupings of river flow patterns on a short time series record for rivers located in the state of Johor, Malaysia. Specifically, we use significance tests to detect any nonlinearity features and existence of structural change in the river flow time series. Then, we fit a Self-Exciting Threshold Autoregressive (SETAR) model to identify the regime switching model for rivers that exhibit nonlinearity features. The results show that there are two main groups amongst the rivers in Johor. All but one river in Johor for the period considered in the study can be explained by significant nonlinear features in their river flow process. Specifically, Johor river, Sayong river and Segamat river that exhibit nonlinear feature patterns, can be further categorized into two different regimes of time series based on the fitted SETAR models.
first_indexed 2024-12-08T06:54:39Z
format Conference or Workshop Item
id utm.eprints-107981
institution Universiti Teknologi Malaysia - ePrints
last_indexed 2024-12-08T06:54:39Z
publishDate 2023
record_format dspace
spelling utm.eprints-1079812024-10-16T07:04:46Z http://eprints.utm.my/107981/ Grouping river flow patterns based on nonlinear features of short time series data Mansor, Nur Syazwin Ahmad, Norhaiza QA Mathematics River flows provide information on the availability of water and its variability in space and time. Identifying similar grouping of river flow patterns is useful so that the knowledge of homogeneous sites can be used to manage water resources more efficiently. However, the grouping of river flow is not easy because river flow can contain several regimes to account for different behaviours. In this study we use a two-stage approach to identify groupings of river flow patterns on a short time series record for rivers located in the state of Johor, Malaysia. Specifically, we use significance tests to detect any nonlinearity features and existence of structural change in the river flow time series. Then, we fit a Self-Exciting Threshold Autoregressive (SETAR) model to identify the regime switching model for rivers that exhibit nonlinearity features. The results show that there are two main groups amongst the rivers in Johor. All but one river in Johor for the period considered in the study can be explained by significant nonlinear features in their river flow process. Specifically, Johor river, Sayong river and Segamat river that exhibit nonlinear feature patterns, can be further categorized into two different regimes of time series based on the fitted SETAR models. 2023 Conference or Workshop Item PeerReviewed Mansor, Nur Syazwin and Ahmad, Norhaiza (2023) Grouping river flow patterns based on nonlinear features of short time series data. In: 5th ISM International Statistical Conference 2021: Statistics in the Spotlight: Navigating the New Norm, ISM 2021, 17 August 2021-19 August 2021, Virtual, Online, Johor Bahru, Johor, Malaysia. http://dx.doi.org/10.1063/5.0111090
spellingShingle QA Mathematics
Mansor, Nur Syazwin
Ahmad, Norhaiza
Grouping river flow patterns based on nonlinear features of short time series data
title Grouping river flow patterns based on nonlinear features of short time series data
title_full Grouping river flow patterns based on nonlinear features of short time series data
title_fullStr Grouping river flow patterns based on nonlinear features of short time series data
title_full_unstemmed Grouping river flow patterns based on nonlinear features of short time series data
title_short Grouping river flow patterns based on nonlinear features of short time series data
title_sort grouping river flow patterns based on nonlinear features of short time series data
topic QA Mathematics
work_keys_str_mv AT mansornursyazwin groupingriverflowpatternsbasedonnonlinearfeaturesofshorttimeseriesdata
AT ahmadnorhaiza groupingriverflowpatternsbasedonnonlinearfeaturesofshorttimeseriesdata