The GNOME model is used in all NOAA Emergency Response Division (ERD) spill responses that require modeling. The model's algorithms function perfectly, but the user is the key to SETTING up the model in Diagnostic Mode for spill response. Because the primary forces that move the oil (wind, currents, etc.) are not generated by GNOME, but input by the user, the user is ultimately responsible for the results.

During spill response at NOAA ERD, the TRAJECTORY ANALYST selects the physics she believes are most appropriate for the spill and sets up the model accordingly. She selects the appropriate hydrodynamic model for the currents, simulating tides and hydrology as necessary, obtains a wind forecast from the NOAA National Weather Service, and estimates other parameters from reviewed literature and experience. Once the model is set up, a prediction is made and later compared with overflight information from on scene.

If the current and wind predictions are accurate, and if the windage and diffusion parameters are set accurately in GNOME, then GNOME will generate very good trajectories. A 48-hour prediction of where the oil will go can be expected to be within 1-2 miles. If the model and observations differ, hindcast model runs are made to tune the model to the local conditions before the next forecast is made. The forecast winds and currents are usually not accurate enough to generate trajectories within 1 mile of accuracy after 48 hours. This is why GNOME supports user-specified uncertainty bounds, which are set according to the uncertainty in the input data. This is also why the GNOME input data is constantly updated during an event, and the model is rerun at least once a day during a spill.

Making predictions in tidally dominated areas is much easier than in eddy-rich offshore environments. The availability of local real-time data and good wind forecasts (from the NOAA National Weather Service) are key to good spill trajectory estimates.

GNOME's Location FILES are tuned to simulate local climatological conditions to the best of our knowledge. Though tides are predictable, other environmental conditions are not so simple. Location Files are not appropriate for spill response, just as an almanac is not appropriate to predict the weather for a particular day.

The GNOME model is used in all NOAA Emergency Response Division (ERD) spill responses that require modeling. The model's algorithms function perfectly, but the user is the key to setting up the model in Diagnostic Mode for spill response. Because the primary forces that move the oil (wind, currents, etc.) are not generated by GNOME, but input by the user, the user is ultimately responsible for the results.

During spill response at NOAA ERD, the trajectory analyst selects the physics she believes are most appropriate for the spill and sets up the model accordingly. She selects the appropriate hydrodynamic model for the currents, simulating tides and hydrology as necessary, obtains a wind forecast from the NOAA National Weather Service, and estimates other parameters from reviewed literature and experience. Once the model is set up, a prediction is made and later compared with overflight information from on scene.

If the current and wind predictions are accurate, and if the windage and diffusion parameters are set accurately in GNOME, then GNOME will generate very good trajectories. A 48-hour prediction of where the oil will go can be expected to be within 1-2 miles. If the model and observations differ, hindcast model runs are made to tune the model to the local conditions before the next forecast is made. The forecast winds and currents are usually not accurate enough to generate trajectories within 1 mile of accuracy after 48 hours. This is why GNOME supports user-specified uncertainty bounds, which are set according to the uncertainty in the input data. This is also why the GNOME input data is constantly updated during an event, and the model is rerun at least once a day during a spill.

Making predictions in tidally dominated areas is much easier than in eddy-rich offshore environments. The availability of local real-time data and good wind forecasts (from the NOAA National Weather Service) are key to good spill trajectory estimates.

GNOME's Location Files are tuned to simulate local climatological conditions to the best of our knowledge. Though tides are predictable, other environmental conditions are not so simple. Location Files are not appropriate for spill response, just as an almanac is not appropriate to predict the weather for a particular day.