The Geostationary Ocean Colour Imager (GOCI) is a visible band ocean colour instrument onboard the Communication, Ocean, and Meteorological Satellite (COMS) scheduled to be in operation from early 2010. The instrument is designed to monitor ocean water environments around the Korean peninsula in high spatial and temporal resolutions. We report a new imaging and radiometric performance prediction model specifically designed for GOCI. The model incorporates the Sun as light source, about 4000km x 4000km section of the Earth surrounding the Korean peninsula and the GOCI optical system into a single ray tracing environment in real scale. Specially, the target Earth section is constructed using high resolution coastal line data, and consists of land and ocean surfaces with reflectivity data representing their constituents including vegetation and chlorophyll concentration. The GOCI instrument in the IRT model is constructed as an optical system with realistic surface characteristics including wave front error, reflectivity, absorption, transmission and scattering properties. We then used Monte Carlo based ray tracing computation along the whole optical path starting from the Sun to the final detector plane, for simultaneous imaging and radiometric performance verification for a fixed solar zenith angle. This was then followed by simulation of red-tide evolution detection and their radiance estimation, in accordance with the in-orbit operation sequence. The simulation results prove that the GOCI flight model is capable of detecting both image and radiance originated from the key ocean phenomena including red tide. The model details and computational process are discussed with implications to other earth observation instruments.