Publication: A study on the development of time series prediction model for phosphorus in reservoirs
Loading...
Date
2020
Authors
Dahan Mohammad Hatem
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This project is to study the development of time series for prediction phosphorus in Fei Tsui reservoir. project is drawn to produce a prediction model, that can be used for water quality monitoring. The test was conducted mainly in Fei Tsui reservoir, Taipei city, Taiwan. The purpose for this project is monitor the phosphorus concentration in the reservoir and build a model based on the previous event records and without the needs to sit visit to test the phosphorus concentration. Phosphorus was a critical issue for past decades in Fei Tsui reservoir. This matter has affected water drinking supplies in Taipei city due to etherification issue in reservoir, as Fei Tsui reservoir considered as main water supply for Taipei city. 27 elements of 10 years average monthly records and 13 years of average annual records have picked and correlated to see which one has strong relationship with phosphorus. NH3, Mg, NO3 and TDS are the parameters that have active relationship with phosphorus. A model was constructed to predict phosphorus concertation in water body based on linear regression. MATLAB software is used to build the prediction model as it considers one of the machine learning software. The result showed that the model of 13 years average annual data records has accuracy more than the 10 years average monthly records due to messing data records in some months. The missing data is as a result of no time series data availability recorded in the events at Fei Tsui reservoir.
Description
TD395.D34 2020
Keywords
Reservoirs