Document Details

Document Type : Thesis 
Document Title :
Using Artificial Neural Networks for Estimating the Roles of Atmospheric Parameters On Sea Level Variations in Eastern Central Red Sea
استخدام الشبكات العصبية الاصطناعية لتقدير دور العوامل الجوية في تغيرات منسوب سطح البحر بالوسط الشرقي من البحر الأحمر
 
Subject : Faculty of Marine Sciences 
Document Language : Arabic 
Abstract : Sea level variations, in general, are known to be induced by the variations of the forcing Astronomical, Meteorological, and Oceanographic functions. Sea level near Jeddah, which is located on the eastern central part of the Red Sea in the proximity of the amphidromic point of the dominating M2 tidal component, experiences only small variations due tidal (Astronomical) forcing. Sea level variations induced by Meteorological and Oceanographic functions, however, are more pronounced near Jeddah due to the variations of the atmospheric parameters that are associated with the monsoonal effects. This study attempted to investigate the role of atmospheric parameters on sea level variations near Jeddah using an Artificial Neural Network (ANN) approach. Non-linear Auto-Regressive Network with eXogenous inputs (NARX) was initially used to simulate the de-tided (residual) sea level record for the year 2011 with tapped-delayed inputs that consist of atmospheric parameters (Atmospheric Pressure, Air Temperature, Cross- and Along- Shore Wind Stresses and Evaporation Rate). The high performance shown by the NARX-based model in accurately simulating the sea level residual indicated that the model was able to capture the non-linear relationships between the residual sea level and the forcing atmospheric parameters. The developed model, therefore, was used to investigate the role of each atmospheric parameter through separate simulations with one of the highly varying atmospheric parameters replaced by its own polynomial fit one at a time. The investigations clearly showed that the model is properly responding to the short-term variations of the atmospheric parameters that are inducing the residual sea level variations. The comparisons made with the simulated sea level obtained using all varying parameters indicated that the roles played by atmospheric parameters do vary seasonally and in magnitude. Atmospheric Pressure role was more pronounced during winter and summer with an effect on sea level in the range of ±4 cm. The effect of Air Temperature was also in the range of ±4 cm; however, the role was more pronounced during spring and fall. Both Cross- and Along- Shore Wind Stresses have shown small contribution in the order of ±2 cm in spring and fall periods during which the roles were more pronounced. Evaporation role was more pronounced during winter and summer with effect range of -7 cm to +4 cm in the winter and -3 to +4 in the summer. Apparently, the results of the carried-on investigations in this study have clearly indicated the ability of an ANN technique (NARX) as a powerful tool that can be used to accurately simulate the residual sea level variations that are induced by the variations of atmospheric parameters. It also can precisely quantify the role played by each atmospheric parameter. 
Supervisor : Dr. Khalid Mahmoud Zubier 
Thesis Type : Master Thesis 
Publishing Year : 1438 AH
2017 AD
 
Added Date : Thursday, June 8, 2017 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
لينه سمير عيونيEyouni, Lina SameerResearcherMaster 

Files

File NameTypeDescription
 40868.pdf pdf 

Back To Researches Page