North Atlantic OSSEs in support of improved hurricane forecasting: Nature Run evaluation

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 5 January 2015: 4:30 PM
131AB (Phoenix Convention Center - West and North Buildings)
Villy H. Kourafalou, Univ. of Miami/RSMAS, Miami, FL; and G. R. Halliwell Jr., R. Atlas, H. S. Kang, M. F. Mehari, M. Le Henaff, L. K. Shay, R. Lumpkin, and G. Goni

A prototype ocean OSSE system, first developed and data validated in the Gulf of Mexico, has been applied on a North Atlantic Ocean domain (5S to 45N and 98W to 20W), chosen to contain the extended North Atlantic hurricane region. The main objectives of this study are: a) to contribute toward a fully relocatable ocean OSSE system by expanding the Gulf of Mexico OSSE to the North Atlantic Ocean; b) demonstrate and quantify improvements in hurricane forecasting when the ocean component of coupled hurricane models is advanced through targeted observations and assimilation. The study is part of a Sandy Supplemental Project and substantially addresses two of the three expected outcomes with respect to hurricanes and other storms: (1) improve the Nation's ability to observe coastal and oceanic conditions associated with these storms; and (2) improve NOAA's capability to accurately predict future storms at longer lead times.

The system is based on the Hybrid Coordinate Ocean Model (HYCOM) and has been applied on a 1/25 degree Mercator mesh for the free-running Nature Run (NR) and on a 1/12 degree Mercator mesh for the data assimilative forecast model (FM); both are nested in the 1/12 degree global HYCOM. A “fraternal twin” system is employed, using two different realizations for NR and FM, each configured to produce substantially different physics and truncation errors. Building two different HYCOM configurations is a unique capability of the HYCOM code, as it contains multiple choices of numerical algorithms and subgrid-scale parameterizations that can be varied to produce the different physics and truncation errors between the two models, as required for a credible OSSE system. The chosen difference in horizontal resolution introduces additional truncation errors. The NR and FM mean and RMS amplitude maps are in good agreement with ocean climatology, such as the AVISO-derived Sea Surface Height Anomaly. Detailed NR evaluation is based on data products, specifically developed from RSMAS and AOML archives. These include upper ocean profiles from air-borne instruments, surface velocity maps derived from the historical drifter data set and tropical cyclone heat potential maps derived from altimetry observations.