The front page of the model system has a brief explanation of update times: http://www.hurricanecity.com/models/ In the disclaimer at the bottom. I'll include the whole thing because most of it is relevant. "The current position and intensity of any storm on this page does not come from the latest advisory from the National Hurricane Center (NHC) for the Atlantic and East Pacific basins or the Central Pacific Hurricane Center (CPHC) for the Central Pacific basin. For this important data you must refer to the agency responsible for issuing forecasts in the particular basin you are looking at. The current position and intensity on this page is provided to show what data the early cycle models initialized with. An invest area, noted with a number from 90 to 99, represents an area that is being watched for possible development. Refer to the tropical outlooks at the NHC and the CPHC for guidance on the possibility of development from these areas and look for the latest official advisory data if the storm is a tropical depression or higher. The best track data, model data and center fix data come from the Automated Tropical Cyclone Forecasting System (ATCF). Our site downloads the latest files from here and processes the data to be visually displayed. That system and the processing system our site uses may contain errors at times. Please consult the agency responsible for issuing forecasts in your basin, as this information, in addition to possibly containing errors from time to time, is only updated every 6 hours by the NHC. Updates made through regular and special forecast advisories may not be reflected here for three hours or more since ATCF data is usually offset from NHC advisory data by 3 hours. Normal ATCF update times are 0Z, 6Z, 12Z, and 18Z. Since the positions are valid at the time noted, it may take an hour for them to be posted to the ATCF system and then be downloaded by our site. Under most normal circumstances, best track data and NHC model data come in starting around 0:30, 6:30, 12:30, and 18:30Z on our site. (And it may be a half hour after that for late posted data.) Sometimes 6Z best track data may not be available for weaker storms. Normal NHC advisory update times are 3Z, 9Z, 15Z, and 21Z, with special advisories possible at any time, and these do not appear here. The models available on this site are for educational purposes only. Meteorologists use these models along with many other tools in order to produce their forecasts. You can't simply look at these models alone and determine where a storm will go. Certain models are more applicable than others, but which ones are more applicable can only be determined by those who know how to use this data. One model is never always right. It is up to weather professionals to look at these models and see which ones are more applicable at the moment and use that knowledge along with the many other resources they have to come up with the best possible forecast. These models are provided for those who are interested in learning more about the tools that weather professionals use. They are not provided for any other reason. For all official hurricane information, refer to the agency responsible for issuing forecasts in your basin. DO NOT RELY ON ANY UNOFFICIAL SOURCES IN LIFE OR DEATH DECISIONS. By using these models, you agree that this site is in no way responsible for the decisions you make based on these models. If you do not agree to this, you cannot use the models on our site." As for Zulu. That is needed. I provide the current time in Zulu on my site. It is visited in many different time zones and it needs to be so that someone can compare the current Zulu time to the model, best track, or recon times. It would be confusing to have local time. I figure people have local time on their computer screen already so they can then see the current Zulu time to see the offset between the two. (Personally I have two clocks on my computer, one in my time and one in Zulu, but I'm nerdy.) CaneTalk displays both the Eastern time zone (where most people on CaneTalk are) and Zulu time on the front page. All time is standardized because there are different time zones. If I say the models come out at 0Z, you know you can convert that to your time zone easily. That is why things are released on Zulu time. People using the data can easily convert to their time. If I had local time where they are, then the update times would be confusing. 0Z, 6Z, 12Z, and 18Z is easy. If I had different numbers depending on the local time zone you are in and someone then forgot to add the proper time zone when they posted about it somewhere, then people in other time zones would have bad info. For me in the Eastern time zone on daylight saving time, I know that this time of year that this is when things update: 2AM, 8AM, 2PM, 8PM. Post that in the central time zone and it would be the wrong times. Someone can correct me if I am wrong on all the following. Global models give you an idea of the surrounding environment but have less resolution on the smaller scale stuff. Models designed for tropical systems focus on the storm environment more finely and use global model data as well. You don't use the GFDL to see where the highs and lows are. You don't use the GFS to see the maximum intensity of the storm. The highs and lows from a global model can allow the GFDL to forecast a track but to get a good view of the highs and lows you want to look at a global model. Some people will look at the heights of certain isobars in a model and compare that to actual data to see if the model got the height right. If it is off and a ridge is stronger than the model expects then a storm will have less ability to move that way. If the ridge is weaker than the model then the model might be showing that the storm will have less ability to move that way when in reality it may be able to move that way a little more so. Ensemble members from global models are not for intensity forecasts. I believe global models don't have the resolution to do that. So if the ensemble member says 50 knots and intensity models like SHIPS says 100 knots, go with the intensity models. The GFS for this storm had ensemble members going all over the place at times, mostly either left or right. This means uncertainty. I don't know how it works exactly, but the way I understand it ensemble members tweak certain conditions. I don't know how or what they do, but they tweak things about different things. If all the ensemble members are fairly close together, that means that confidence is higher because even if conditions change somewhat the storm will probably still move along the same general path. If the ensemble members are all over the place, that means confidence is lower because depending on the conditions, there could be more significant changes in the track depending on what conditions actually transpire. This storm is an epic example of that when you have a trough, upper level low and high. Tweak some things about the various players, movements and strengths, and perhaps initial conditions and some ensemble members go one way and some go the opposite way, meaning the placement and movement matter greatly to the path of the storm, so until you have a good handle on what will likely happen, the track is very uncertain. I don't know all the details, but the GFS at 0Z and 12Z are usually better than 6Z and 18Z. I can't find the reasons at more of an official site at the moment. On my site, I purposefully don't try to explain too much about things. I provide a page like this: http://www.tropicalatlantic.com/models/more/ And try to leave out what might be considered helpful. I do it for a lot of reasons, such as one model is never always right which is why we have more than one model, models can't always be compared equally (early/late), not every run is created equal (noted in sentence above this paragraph), some models are better at certain aspects rather than other models (ex: BAMS, shallow storms, BAMD, deeper storms), I'm not a weather professional, and for legal reasons. Eventually, I plan to offer NHC advisory data on my site. I currently do not. Until I do, I don't want too many people relying on my site's data too much as what you are seeing is not advisory data but best track data. So most of my site is not for most people. It tries to provide just the facts and allows someone who knows how to use the information to use it properly. I don't want to give someone the wrong impression amongst a lot of facts. On a message board like this, people know what to expect. You will have a lot of opinions from all sorts of people and you know that some will be better than others. I can say something here that I would not want to say on my site amongst all the other data because some important sites link to my site and I want to provide just the facts since I am not a weather professional. I provide links to other information, but try to avoid implying one thing is better than another thing. As for the model listing: http://www.hurricanecity.com/models/data.cgi?page=models The reason that page is the way it is relates to how it is created. It is automatically created from data within the model system that has the model identifier, model name, and a tiny bit of other official information so that the system can run fast. Every time the system is run that model information is loaded, so it needs to be very concise and the format of that page is created automatically, not manually, so I can't adjust anything because I can make a simple change to the main place where that information is stored in the model system, when I need to reflect the NHC changing or adding something, and can have it deployed throughout the site, including that page, instantly. I know you want an explanation of the models, rather than just a name, but there are links to the NHC's model page on that listing page. An understanding of the models is complex. Unfortunately, that cannot be helped. I don't want to simplify it to much to make someone think they then know everything there is to know. The NHC page has about the best explanation that you can get and that page I often link to: http://www.nhc.noaa.gov/modelsummary.shtml I would rather everyone go there for the early/late, statistical/dynamical, interpolated, and global. Who better to explain it than the NHC? It's better to drive someone to the NHC because there explanation on these important things are far better than one I could add and copying it from their site is not as good as someone actually seeing it on the NHC's site. I know it is complex, and can be daunting at times, but it is very important people learn about it properly if they want an understanding of it and it is going to contain some lengthy reading. I can provide little bits of information amongst the facts on my site but then you have a problem. If you present one little bit of information people think of it as either most important or of the only importance. That may not be true. If I provide a little bit of explanation, then you have to present a whole lot of explanation, some of which I don't know, so that they don't get the wrong impression. So rather than leave out some vital information, I try to have it just contain the facts. I simply process the data and leave out the commentary. I know it doesn't seemingly come out as helpful that way, but I am helping people by not having incomplete, misleading, or sometimes invalid commentary in certain situations. With that said, if you do have any comments on specific areas I can try to improve it here and there. |