Author(s): Shivaanand V; Priyanga S A; Manjula R

Linked Author(s):

Keywords: ANN; Flow data prediction; Water level prediction

Abstract: ANN is a versatile tool which can be used in many applications. One such application is in predicting ungauged data using data from other stations. For a tool which only uses single type of data for predicting the ungauged data, ANN has relatively good accuracy. For prediction of flow data for any station requires daily data from upstream stations. This study uses ANN for prediction of water level and flow data for Musiri by taking daily data from 1 Jan 2006 till 31 Dec 2011. Comparative study is performed by taking daily data from 4 other stations and predicted the flow data and water level at Musiri. Prediction of Musiri data with single separate stations to do comparative study of how many stations and the distance between them is to needed good accuracy. Flow data prediction for Musiri using 4 other stations gives = 0.917, while prediction using the farthest station (Thevur) gives = 0.8414. Water level prediction for Musiri using 3 other stations gives = 0.9562, while prediction using farthest station (Urachikotti) gives = 0.9377. This paper provides a study for using distant stations for prediction during the unavailability of data for the required station.

DOI: https://doi.org/10.64697/HIC2024_P379

Year: 2024