A Cyberinfrastructure-based Approach to Real Time Water Temperature Prediction

TL;DR

This paper introduces WT-Agabus, a cyberinfrastructure system supporting real-time water temperature prediction using neural networks and online data, enhancing aquatic ecosystem management capabilities.

Contribution

It presents a novel cyberinfrastructure platform that supports multiple prediction models and introduces a neural network-based water temperature prediction model using publicly available data.

Findings

Successful implementation of WT-Agabus system

Neural network model achieves accurate predictions

Supports real-time water temperature forecasting

Abstract

The prediction of water temperature is crucial for aquatic ecosystem studies and management. In this paper, we raise challenging issues in supporting real time water temperature prediction and present a system called WT-Agabus to address those issues. The WT-Agabus system is designed to be a cyberinfrastructure and to support various prediction models in a uniform way. In addition, we present a neural network-based water temperature prediction model to use only data available online from Korea Meteorological Administration (KMA). In this paper, we also show the current prototype implementation of the WT-Agabus system to support the prediction model