An operational oceanographic forecasting and observing system for the Eastern Mediterranean Levantine basin: The Cyprus coastal ocean forecasting and observing system (original) (raw)

Operational coastal ocean forecasting in the Eastern Mediterranean: implementation and evaluation

Ocean Science Discussions, 2006

The Cyprus Coastal Ocean Forecasting and Observing System (CYCOFOS) has been producing operational flow forecasts of the northeastern Levantine Basin since 2002 and has been substantially improved in 2005. It is the first system in the Mediterranean to produce a forecast every day at the coastal scale. CYCOFOS uses a the POM (Princeton Ocean Model) flow model, and recently, within the frame of the MF-STEP project (Mediterranean Forecasting System, Toward Environmental Prediction), the flow model was upgraded to use the hourly SKIRON atmospheric forcing, and its resolution was increased from 2.5 km to 1.8 km. The CYCOFOS model is now nested in the ALERMO (Aegean Levantine Eddy Resolving Model) regional model from the University of Athens, which is nested within the MFS (Mediterranean Forecasting System) basin model. The Variational Initialization and FOrcing Platform (VIFOP) has been implemented to reduce the numerical transient processes following initialization. Moreover, a five-day forecast is repeated every day, providing more detailed and more accurate information, of particular value to coastal end users. Forecast results are posted on the web page http://www.ucy.ac.cy/cyocean. The new, daily, high-resolution forecasts agree exceptionally well with the ALERMO regional model. The agreement is better and results more reasonable when VIFOP is used. Active and slave experiments suggest that a four-week active period produces realistic results with more small-scale features. Bias with respect to the slave mode is negligible and there is no detectable bias with remote sensing images (for September, 2004). In situ observed hydrographic data from south of Cyprus are similar in many ways to the corresponding forecast fields. Plans for further model improvement include assimilation of observed temperature profiles (XBT), conductivity-temperature-depth (CTD) profiles from drifters or research cruises, and CT data from the CYCOFOS ocean observatory.