Data-driven approaches for Chao Phraya River management

Name: Phakawat Lamchuan

LCSH: Kasetsart University -- Dissertations. Ph.D. (Water Resources Engineering) 2021

Classification :.LCCS: TD313.T48

LCSH: Kasetsart University. -- Department of Water Resources Engineering -- Dissertations

LCSH: Water quality -- Thailand -- Chao Phraya River

LCSH: Suspended sediments -- Thailand -- Chao Phraya River

LCSH: Neural networks (Computer science)

Abstract: This study utilized a data-driven approach for estimating and forecasting water quality parameters in the Great Chao Phraya River Basin. The study focused on the physical properties of three parameters, namely suspended sediment load, salinity, and turbidity. This study has two main scopes: 1. To estimate suspended sediment load by data-driven approach and compared with the traditional method (Sediment rating curve, SRC) determination by the Royal Irrigation Department. The study area was the Upper Chao Phraya River Basin, consisting of the Ping River Basin, Wang River Basin, and Yom River Basin. 2. To forecast salinity and turbidity for the early-warning system for producing water supply in the study area of the Samlae raw water pumping station of the Metropolitan Waterworks Authority, which is in the lower Chao Phraya River Basin. The study results were categorized according to two main scopes: 1. The results showed that our new approach for all three study areas (PLR and ANNs) gave better results with the observed data than the traditional SRC method, except for MLR, SLR, and QLR. For P1, ANNs provided NSE of 0.96 and RMSE of 769 tons/day, while PLR showed NSE of 0.82 and RMSE of 1,553 tons/day. For W4A, ANNs provided NSE of 0.78 and RMSE of 2,382 tons/day. PLR presented an NSE of 0.82 and RMSE of 2,382 tons/day. For Y14, the result of PLR (NSE of 0.94 and RMSE of 2,411 tons/day) is better than ANNs (NSE of 0.81 and RMSE of 4,451 tons/day). 2. The accuracy of the confusion matrices (True positive rates) for salinity forecasted using the selected ANNs combined model with forecast periods from 24 to 48 hours is 0.840 and 0.780, respectively. 2. The salinity can be forecasted at the Samlae raw water pumping station with an RMSE of 0.054 g/l at 24 hours and up to 0.107 g/l at 120 hours. The accuracy of the confusion matrices (True positive rate) for salinity forecasted using the selected ANNs combined model with forecast periods from 24 to 48 hours is 0.840 and 0.780, respectively. The turbidity illustrated forecast results with an RMSE of 5.372 NTU at 24 hours and up to 15.440 NTU at 120 hours. The accuracy of the confusion matrices (True positive rates) with 0.982 at 24 hours and up to 0.557 at 120 hours. This study demonstrated that data-driven approaches can effectively estimate and forecast river water quality.

Kasetsart University. Office of the University Library

Address: Bangkok

Email: tdckulib@ku.ac.th

Name: Adichai Pornprommin

Role: Thesis Advisor

Name: Jiramate Changklom

Role: Thesis CO-Advisor

Created: 2021

Modified: 2025-07-25

Issued: 2025-07-25

วิทยานิพนธ์/Thesis

application/pdf

URL: https://www.lib.ku.ac.th/KUthesis/2564/phakawat-lam-all.pdf

CallNumber: TD313.T48 .P53

eng

Spatial: Thailand Chao Phraya River

Spatial: Thailand Chao Phraya River

DegreeName: Doctor of Philosophy

Level: Doctoral Degraee

Descipline: Water Resources Engineering

Grantor: Kasetsart University

©copyrights Kasetsart University

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	phakawat-lam-all.pdf	8.99 MB

ใช้เวลา

0.03506 วินาที