Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods

Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods

Probabilistic flood forecasting, which provides uncertain information in the forecasting of floods, is practical and informative for implementing flood-mitigation countermeasures. This study adopted various machine learning methods, including support vector regression (SVR), a fuzzy inference model...

Full description

Bibliographic Details
Main Authors: Dinh Ty Nguyen, Shien-Tsung Chen
Format: Article
Language:English
Published: MDPI AG 2020-03-01
Series:Water
Subjects:
flood
probabilistic forecasting
support vector regression
fuzzy inference
<i>k</i>-nearest neighbors
Online Access:https://www.mdpi.com/2073-4441/12/3/787
id doaj-4368fccc6d024f5997fc1f22445bf988
record_format Article
spelling doaj-4368fccc6d024f5997fc1f22445bf9882020-11-25T00:44:43ZengMDPI AGWater2073-44412020-03-0112378710.3390/w12030787w12030787Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning MethodsDinh Ty Nguyen0Shien-Tsung Chen1Ph.D. Program for Civil Engineering, Water Resources Engineering, and Infrastructure Planning, Feng Chia University, Taichung City 407, TaiwanDepartment of Hydraulic and Ocean Engineering, National Cheng Kung University, Tainan City 701, TaiwanProbabilistic flood forecasting, which provides uncertain information in the forecasting of floods, is practical and informative for implementing flood-mitigation countermeasures. This study adopted various machine learning methods, including support vector regression (SVR), a fuzzy inference model (FIM), and the <i>k</i>-nearest neighbors (<i>k</i>-NN) method, to establish a probabilistic forecasting model. The probabilistic forecasting method is a combination of a deterministic forecast produced using SVR and a probability distribution of forecast errors determined by the FIM and <i>k</i>-NN method. This study proposed an FIM with a modified defuzzification scheme to transform the FIM&#8217;s output into a probability distribution, and <i>k</i>-NN was employed to refine the probability distribution. The probabilistic forecasting model was applied to forecast flash floods with lead times of 1&#8722;3 hours in Yilan River, Taiwan. Validation results revealed the deterministic forecasting to be accurate, and the probabilistic forecasting was promising in view of a forecasted hydrograph and quantitative assessment concerning the confidence level.https://www.mdpi.com/2073-4441/12/3/787floodprobabilistic forecastingsupport vector regressionfuzzy inference<i>k</i>-nearest neighbors
collection DOAJ
language English
format Article
sources DOAJ
author Dinh Ty Nguyen
Shien-Tsung Chen
spellingShingle Dinh Ty Nguyen
Shien-Tsung Chen
Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
Water
flood
probabilistic forecasting
support vector regression
fuzzy inference
<i>k</i>-nearest neighbors
author_facet Dinh Ty Nguyen
Shien-Tsung Chen
author_sort Dinh Ty Nguyen
title Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
title_short Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
title_full Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
title_fullStr Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
title_full_unstemmed Real-Time Probabilistic Flood Forecasting Using Multiple Machine Learning Methods
title_sort real-time probabilistic flood forecasting using multiple machine learning methods
publisher MDPI AG
series Water
issn 2073-4441
publishDate 2020-03-01
description Probabilistic flood forecasting, which provides uncertain information in the forecasting of floods, is practical and informative for implementing flood-mitigation countermeasures. This study adopted various machine learning methods, including support vector regression (SVR), a fuzzy inference model (FIM), and the <i>k</i>-nearest neighbors (<i>k</i>-NN) method, to establish a probabilistic forecasting model. The probabilistic forecasting method is a combination of a deterministic forecast produced using SVR and a probability distribution of forecast errors determined by the FIM and <i>k</i>-NN method. This study proposed an FIM with a modified defuzzification scheme to transform the FIM&#8217;s output into a probability distribution, and <i>k</i>-NN was employed to refine the probability distribution. The probabilistic forecasting model was applied to forecast flash floods with lead times of 1&#8722;3 hours in Yilan River, Taiwan. Validation results revealed the deterministic forecasting to be accurate, and the probabilistic forecasting was promising in view of a forecasted hydrograph and quantitative assessment concerning the confidence level.
topic flood
probabilistic forecasting
support vector regression
fuzzy inference
<i>k</i>-nearest neighbors
url https://www.mdpi.com/2073-4441/12/3/787
work_keys_str_mv AT dinhtynguyen realtimeprobabilisticfloodforecastingusingmultiplemachinelearningmethods
AT shientsungchen realtimeprobabilisticfloodforecastingusingmultiplemachinelearningmethods
_version_ 1725273869920501760

Similar Items