No abstract

No abstract

The evaluation of hurricane forecast skill requires ensembles of historical forecasts. The purpose of this article is not to undertake such an evaluation, but rather to demon strate the current status of satellite physical retrievals and their potential to provide valu able information for such evaluations and contribute to model improvements. Figure 3 shows a pictorial example of the 120-hour accumulated surface rainfall from satellite retrievals, and from single high-resolution forecasts from ECMWF and NASA models.Predictions The hurricane in the ECMWF forecast, though, deviates by two to three degrees east of the best track, and makes landfall between Ala bama and Florida about 12 hours late. These differences in the hurricane track and accu mulated precipitation may reflect inadequa cies in the large-scale circulation provided in the initial conditions, or imperfect model physical parameterizations, but also may be due to the system's lack of predictability. Developments in Hurricane ForecastsAdvances in spaceborne observations and numerical weather prediction (NWP) models provide new opportunities for improving hurricane forecasts. Apart from their impor tance for NWR global atmospheric models of hurricanes and their forecasts represent an important and unique test bed of model formulations.Recent developments that include moving from synoptic-scale-resolving to mesoscaleresolving global models show some very encouraging results. In addition to increasing resolution and including more physically based parameterizations on mesoscale effects in conventional general circulation models, cloud-scale-resolving global models-in which the cloud dynamics and mesoscale processes are explicitly resolved-also are being devel oped and could be used as a parallel approach to more realistically simulate hurricanes in global models in the future.Better resolution of the hurricane struc ture and larger-scale steering circulation, along with improved initial conditions pro vided by high-resolution satellite data and sophisticated data assimilation systems, could lead to better detection, monitoring, under standing, and prediction of the genesis and development of hurricanes that have such a devastating impact on society.

The REMUS-100 Hull and Harbor vehicle is being developed by the Ocean Systems Laboratory (OSL) at the Woods Hole Oceanographic Institution, under Office of Naval Research (ONR) funding. The purpose of the vehicle is autonomous search and survey in close confines, such as harbors, finger piers and ship or submarine hulls. Initial tests of the system confirmed the unique suitability of this vehicle to this mission.Second generation developments leverage the strengths of the first generation engineering and continue to make the vehicle more autonomous, more productive and easier to apply to the specific restrictions of rapid, confined area search and survey.

No abstract