OSSE Assessment of Ocean Observing System Enhancements to Improve Coupled Tropical Cyclone Intensity Prediction

George R Halliwell Jr1, Michael F Mehari2, Ricardo M. Domingues3, Vassiliki Kourafalou4, Robert M Atlas1, HeeSook Kang4 and Matthieu Le Henaff5, (1)NOAA Miami, Miami, FL, United States, (2)Cooperative Institute for Marine and Atmospheric Studies Miami, NOAA/AOML/PhOD, Miami, FL, United States, (3)National Centers For Environmental Prediction-Environmental Modeling Center, College Park, MD, United States, (4)University of Miami, Miami, FL, United States, (5)NOAA Atlantic Oceanographic and Meteorological Laboratory, PhOD, Miami, FL, United States
Abstract:
A new ocean OSSE system validated in the tropical/subtropical Atlantic Ocean is used to evaluate ocean observing strategies during the 2014 hurricane season with the goal of improving coupled tropical cyclone forecasts. Enhancements to the existing operational ocean observing system are evaluated prior to two storms, Edouard and Gonzalo, where ocean measurements were obtained during field experiments supported by the 2013 Disaster Relief Appropriation Act. For Gonzalo, a reference OSSE is performed to evaluate the impact of two ocean gliders deployed north and south of Puerto Rico and two Alamo profiling floats deployed in the same general region during most of the hurricane season. For Edouard, a reference OSSE is performed to evaluate impacts of the pre-storm ocean profile survey conducted by NOAA WP-3D aircraft. For both storms, additional OSSEs are then conducted to evaluate more extensive seasonal and pre-storm ocean observing strategies. These include (1) deploying a larger number of synthetic ocean gliders during the hurricane season, (2) deploying pre-storm synthetic thermistor chains or synthetic profiling floats along one or more “picket fence” lines that cross projected storm tracks, and (3) designing pre-storm airborne profiling surveys to have larger impacts than the actual pre-storm survey conducted for Edouard. Impacts are evaluated based on error reduction in ocean parameters important to SST cooling and hurricane intensity such as ocean heat content and the structure of the ocean eddy field. In all cases, ocean profiles that sample both temperature and salinity down to 1000m provide greater overall error reduction than shallower temperature profiles obtained from AXBTs and thermistor chains. Large spatial coverage with multiple instruments spanning a few degrees of longitude and latitude is necessary to sufficiently reduce ocean initialization errors over a region broad enough to significantly impact predicted surface enthalpy flux into the storm. Error reduction in hurricane intensity forecasts resulting from the additional ocean observations is then assessed by initializing the ocean component of the HYCOM-HWRF coupled prediction system with analyses produced by the OSSE system.