Japan/East Sea Model Predictability
Abstract
Three major input uncertainties (initial velocity field, open boundary conditions, and atmospheric forcing) limit the ocean modeling capability. The Princeton Ocean Model (POM) implemented to the Japan/East Sea (JES) is used to investigate the ocean predictability due to the input uncertainties. Two-step (pre-simulation and simulation) initialization is used to obtain "standard initial velocity". Twelve experiments are conducted with one control run and 11 sensitivity runs. The control run is to integrate POM-JES from the "standard initial velocity" with the lateral transport (unperturbed) and the daily surface wind stress, net heat flux, and fresh-water flux interpolated from the COADS monthly mean data (unperturbed). The sensitivity runs are to integrate POM?JES with replaced initial velocity fields (with or without diagnostic initialization), and noisy winds and lateral boundary transports. Model uncertainty due to uncertain input data (initial velocity, winds, and lateral boundary transport) is significant. Level independent relative root mean square error for the whole JES is 0.2-0.5 for uncertain initial velocity field, 0.19 for uncertain surface winds with 0.5 m/s noise, and 0.20 for uncertain lateral transport with 5% noise. The maximum level dependent relative root mean square error reaches 0.6 at the surface for uncertain winds (0.5 m/s noise), and 0.18 at the bottom for uncertain lateral transport (5% noise). Model uncertainty reduces with time for uncertain initial velocity field oscillates with an evident error growing trend for uncertain winds, and oscillates with no evident error growing trend for uncertain lateral transport. Furthermore, there is no difference using and not using the diagnostic (velocity) initialization and no difference in choosing periods (30-90 days) for diagnostic initialization.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 29, 2005
- Accession Number
- ADA593769
Entities
People
- Chang S. Kim
- Chin-lung Fang
- Peter Cheng Chu
Organizations
- Naval Postgraduate School