This study aims to enhance Fiji's tsunami
early warning system by employing a tsunami database, which includes 64 (four
magnitudes and depths at four source points) precomputed simulation results
stored in a MySQL database. The current system predominantly relies on manual
input of forecasting messages, making it time-consuming and error-prone. Five
methods for retrieving tsunami height from the database were assessed,
revealing significant variations in the estimated tsunami heights. The most
simple approach was defined as Method 2, which selects the maximum tsunami
height among the four source points based on the searched magnitude and depth.
This method was used for its simplicity and ease of implementation, primarily
due to its avoidance of complex interpolation techniques. Compared to the
existing system, which takes 10 to 15 minutes to generate and dispatch tsunami
warnings, the enhanced system can produce forecasting messages in seconds using
Method 2 after earthquake parameters are obtained.
Keywords: Tsunami Early
Warning, Tsunami Database, Tsunami Height Estimation, Information
Dissemination, Tide Gauges.