How does the NHC come up with their forecasts/advisories?

[Ladylight]

Obviously they look at all the models since they mention them and point out when some are in disagreement with the track they're going with. They also mention when they're in agreement with the model consensus. But at other times, they mention being left or right of the consensus. So I was wondering, do they have their own models they use to come up with their official track and form their advisories from that? Or do they simply base it on their professional experience as pro mets? Or is there something else they use? Does anyone know? I've always wondered about this so would appreciate any insight. Thanks.

[Cyclenall]

I'm certain it would be based off of their professional experience as Hurricane Specialists combined with what you mentioned about the models and data.

[supercane]

Making a forecast includes more than just looking at the models. A simplified timetable is presented below:

6 Hour Forecast Cycle
When a storm threatens the following occurs
0:00 A new hurricane forecast cycle begins.
0:45 Receive the location of the center of the hurricane.
1:00 Initialize or start the hurricane models with the storm's location and intensity
1:20 Receive model guidance and prepare a new hurricane forecast.
2:00 Coordinate with National Weather Service and Dept. of Defense.
3:00 Issue the full hurricane advisory package.
3:15 Participate in the Federal Emergency Management Agency (FEMA) conference call with the affected states.
6:00 A new hurricane forecast cycle begins.

Source: Hurricane preparedness: Forecast process

[Ladylight]

Thank you both so much. I understand better now. There's so much that goes into it!

[theavocado]

I'll try to shed a little more insight to the forecasting process, if I may.

One of the key first steps, and this is usually done before the locations from TAFB and NESDIS, is to conduct a thorough analysis of the synoptic pattern. An identification of the low, mid and upper level major features and flow patterns using oberservations from ground stations, aircraft and satellite data is conducted.

The analysis is then compared to the initialization of the models. Sometimes models initialize poorly: sometimes the features are the wrong intensity, in the wrong location, or are moving the wrong speed. These abborations are noted and often times the best models are identified. Also, statistics are often run with the models regarding how they have been handling the storm up to the current point.

After the locations come in from TAFB and NESDIS, the current best track location is determined and then the models are interpolated to the best track location (the models are 6 hours old, and the location and time need to be adjucted to the current best track).

Once the model tracks, along with the main and secondary consensus tracks, are laid out, then consideration is taken if the models aren't in perfect agreement (they are never in perfect agreement). The forecaster will adjust the track from the consensus based on how the models are behaving based on the earlier evaluation, and consideration is taken for how the current analysis looks. For example, if a system underwent rapid intensification, and all the models missed it, the system is likely steering at a mean flow than the models computed. So, in this case, the forecaster might adjust the track closer to the 850-500mb mean flow if it looks like the models are still tracking it close to the 850-700mb mean flow.

During and after the track is laid, the forecaster has to consider the intensity of the system. Global models are notoriously poor at forecasting intensity, and mesoscale models are notorious regarding track. The forecaster has to look at the global tracks and mesoscale intensities, along with statistical intensities, climatological intensities, oceanography, and upper level features to forecast the official intensity. There is really no good guidance outside of STIPS (a model and statistical blend, if you'll excuse my over simplification) to accomplish this task.

Then, they try and express all of this work from the last 6 hours into a short message and a line on a map...

[JtSmarts]

Excellent post theavocado, this is something I've always wondered about but never asked. Definitely makes me appreciate what they do over there at the NHC much more.

[Ladylight]

Wow! Thank you so much for such a great and detailed explanation. I had no idea just how much adjusting and assessing went into it. We really should appreciate those folks more!

I assume there is a team of mets who do this dueling every shift. Do you know, by any chance, how much influence the lead met (the one who puts his/her name to the final report) has over it's contents? Obviously, some of this is very much subject to interpretation. So I'm wondering if the met reporting gets to make the final calls or it's all by consensus or general guidelines

Thanks again. You're great!

[theavocado]

To be honest, I don't really know how the interaction between the Specialists at NHC plays out. At my organization, there is only one met working on each system. Generally they are autonomous, but there are a few triggers that would cause the branch chief to get involved (new forecast track deviates from old track by XX miles, forecast track deviates from consensus by XX miles, etc.), and then several people discuss it and the Branch Chief has final say.