Proceedings of the 12 th International Coral Reef Symposium, Cairns, Australia, 9-13 July 2012 10A Modelling Reef Futures Ghost of bleaching future: Seasonal Outlooks from NOAA's Operational Climate Forecast System C. Mark Eakin 1 , G. Liu 1 , M. Chen 2 , A. Kumar 2 1 NOAA Center for Satellite Applications and Research, Coral Reef Watch, College Park, MD, 20740 USA 2 NOAA National Centers for Environmental Prediction, Climate Prediction Ctr, College Park, MD, 20740 USA Corresponding author: mark.eakin@noaa.gov Abstract. Using models to inform the possible future of coral reefs can happen on many scales. At the seasonal timescale, the National Oceanic and Atmospheric Administration’s (NOAA) Coral Reef Watch (CRW) and National Centers for Environment Prediction (NCEP) recently made a major advance in NOAA’s ability to predict thermal stress capable of causing mass coral bleaching: a newly-developed global seasonal outlook system based on NOAA’s operational Climate Forecast System (CFS). These outlooks predict the probability of thermal stress events capable of causing large-scale, mass coral bleaching, using a weekly, 28-member ensemble of sea surface temperature forecasts from the CFS. The new system builds upon the first global seasonal bleaching outlook system; collaboration between CRW and NOAA’s Earth System Research Laboratory used a statistical climate model to produce the first seasonal bleaching outlook system released in 2008 at the 11 th International Coral Reef Symposium. This paper describes the new CFS-based outlook, initial testing using a series of hindcast and forecast simulations, and the performance of the system during recent bleaching seasons. Key words: Coral bleaching, modeling, prediction, SST Introduction Mass bleaching of coral reefs has occurred with increasing frequency in recent decades. The US National Oceanic and Atmospheric Administration’s (NOAA) Coral Reef Watch (CRW) provides critical information to reef managers and scientists based on near-real-time satellite monitoring of thermal stress conducive to coral bleaching (Liu et al. 2006). However, many users have requested information on the likelihood of coral bleaching months in advance. Longer lead-time information helps managers prepare for the bleaching, as many actions laid out in bleaching response plans are expensive and/or controversial (Maynard et al. 2009). In 2008, CRW released the world’s first prediction tool for forecasting coral bleaching weeks to months in advance (Liu et al. 2009). However, that system was based on a statistical global sea surface temperature (SST) forecast model, the Linear Inverse Model (LIM) (Penland and Matrosova 1998), limiting the system to a single, deterministic seasonal bleaching outlook. CRW has now partnered with the NOAA National Centers for Environmental Prediction (NCEP) to develop a next-generation global seasonal bleaching outlook based on an operational ensemble prediction system for SST, providing a dynamical, probabilistic seasonal coral bleaching thermal stress outlook delivering advance warning to coral reef managers, scientists, stakeholders, and the public. Here we introduce this bleaching outlook system and provide an initial evaluation of its performance. Material and Methods NCEP Climate Forecast System The new seasonal bleaching outlook system is based on NOAA NCEP’s Climate Forecast System Version 1 (CFSv1) (Saha et al. 2006), a fully coupled ocean– land–atmosphere dynamical seasonal prediction system. For its atmosphere, CFSv1 uses the 2003 version of the NCEP Global Forecast System model at T62 horizontal resolution with 64 vertical layers, and for its ocean it uses the Geophysical Fluid Dynamics Laboratory Modular Ocean Model version 3 (MOM3) (Pacanowski and Griffies 1998) with zonal resolution of 1° and meridional resolution of 1⁄3° between 10°S and 10°N, gradually increasing to 1° poleward of 30°S and 30°N. The ocean has 40 vertical layers with 10 m vertical resolution from the surface to 240-m depth. The atmosphere and ocean components are coupled daily without any flux adjustment. The CFS is initialized using oceanic conditions from the NCEP Global Ocean Data Assimilation System (GODAS) (Saha et al. 2006), with SSTs relaxed to Reynolds OISST (Reynolds et al. 2002) with a time scale of 5 days. Atmospheric and terrestrial initial conditions are obtained from near- real-time observations assimilated using the