The weather glossary is a list of every meteorological term that I use often organized alphabetically. If a term you do not understand or know is not on this glossary, please comment below and I will be sure to add it.
AAO- Antarctic Oscillation. This is the least important of the four teleconnections, with little research being done on how it affects the Northern Hemisphere. I only use it to see when there are big swings, which can signal global shifts in the pattern. It can be found here.
Advisory- This is a product issued by the National Weather Service. It is a signal that conditions will be dangerous either due to flooding, fog, wind, snow, or wintry weather. However, an Advisory does not mean conditions will be life threatening, just that dangerous conditions are imminent.
Analog- In weather, when making a longer term forecast or looking at storms, it is often good to compare to previous patterns or storms, called analogs.
AO- Arctic Oscillation. This is one of the three most important teleconnections. When the AO is negative, there is stronger blocking in the atmosphere that sends sets colder air up better to eject into the eastern United States. When the AO is positive, it is harder to get colder air into the eastern United States. A negative AO is more favorable for snow, and the AO forecasts can be found here.
ARW- Advanced Research WRF model. This is one of two 48 hour short range, high resolution models. It is used to help forecast precipitation totals and fine tune forecasts right before storms hit. Similar to the NMM and 4km NAM. It can be found here: http://www.meteo.psu.edu/~fxg1/ewallhires.html
Blocking- In simple terms, blocking in the atmosphere is formed by ridges or areas of stagnant high pressure that dictate storm movement and can slow storms down. In a progressive pattern, there is less blocking so storms generally move from west to east, though with blocking the pattern is more amplified so storms can move in many different directions, making them harder to forecast but also increasing the chance that a storm moves up the east coast and becomes a Nor’ Easter, hitting Southwestern Connecticut with snow. Details from the National Weather Service can be found here.
BUFKIT- This is a program that takes a vertical cross-section of the atmosphere and shows temperatures and other data throughout all of it. At any certain time, it enables the viewer to see the temperature at any point in the troposphere, helping forecast precipitation types and snowfall ratios. This also takes model output for a specific location and shows what predicted accumulations from rain/snow are. More details on how to get BUFKIT for free can be found here.
CMC- Canadian Meteorological Centre, which hosts the CMC weather model. This weather model is on average slightly less accurate than the GFS and ECMWF weather models, and does well in patterns with a lot of blocking or amplification, and tends to do much worse with amplified patterns. It is run twice a day, and comes out around midnight and noon. 0z run can be viewed here and the 12z run can be viewed here.
CPC- Climate Prediction Center, which forecasts droughts and longer term weather, such as 3-month forecasts, 2-3 week forecasts, etc. Their products I find are rarely very informative across microclimates like Southwestern Connecticut, which is why I rarely reference them, unless they highlight an area as having potential for heavy rain/snow. The CPC’s homepage is here.
DGEX- This is one of the least accurate weather models, which occasionally can show a decent solution but is rarely used seriously. It is run twice a day at 6z and 18z, and essentially comes out at 5 AM and 5 PM. 6z run can be found here and 18z run can be found here.
ECMWF- European Center for Medium-Range Weather Forecasting. They produce one of the most accurate weather models, called either the ECMWF or the Euro, along with one of the most accurate model suites. The model comes out twice a day with a 0z run and a 12z run, but the preliminary model suite is typically available around 2 AM and 2 PM, with full model information available around 3:30 AM and 3:30 PM. It is operated privately and only crude images of forecasts every 24 hours are available to the public, which can be found here for 0z and here for 12z. I will reference it a lot but rarely, if ever, link it, as I have to pay for access. The model typically does better in periods with more blocking, and was the one model to nail Hurricane Sandy, giving it more credence in the meteorological community.
Ensembles- These are runs of a model run with slightly different different conditions multiple times. Most models have around 12 ensemble members, which are then averaged together to get an “ensemble mean”. Because of the number of the options in these, these means are generally more accurate. The GFS, ECMWF, NOGAPS, and CMC models are the ones that have ensembles that are used the most often, with the GFS and ECMWF means usually being the most accurate. I will always post on the blog when I am referencing a specific ensemble so you can view it, unless they are the ECMWF ensembles which I cannot post.
GFS- This is the Global Forecasting System weather model, which is the main American weather model used in the medium to long range. It typically has a bias to be slightly too progressive, which is why it often places storms too far to the east, and often suffers from feedback at times as well. I will outline whenever I think there are outlines with the model, but I still do use it often in forecasts. It is for times a day with data from 6z, 12z, 18z, and 0z, and comes out around 5 AM, 11 AM, 5 PM, and 11 PM respectively. It can be found on the Penn State Ewall, which is here.
GEFS- Global Ensemble Forecast System. These are the ensemble members of the GFS, and sometimes instead of saying GFS ensembles it is easier to abbreviate as just GEFS. These come out about an hour later than the GFS, and I will post either the individual members or the GEFS mean. They can also be found here.
GOA- Gulf of Alaska. Low/high pressure systems in this region can have significant impacts on the overall pattern, which can sometimes be discussed in blog posts where I am analyzing the overall pattern.
GOM- Gulf of Mexico. This typically provides moisture for storms coming up into the Northeast.
HPC- Hydrometeorological Prediction Center. This issues updates inside 3 days on winter weather, and their products are sometimes posted/tweeted as they focus specifically on precipitation, flooding potential, heavy snow/ice/sleet potential, etc. I find they are typically fairly accurate, and since their forecasts are short term they are in depth and their products useful.
HRRR- High Resolution Rapid Refresh Model. This is a short range model that only goes out 15 hours from the point. It measured many different levels of the atmosphere for temperature and other indices, and can be used for development of snowfall banding and can forecast short term snowfall accumulations. While sometimes hard to find, it can be found here. I reference this model a lot as storms are occurring, but it is so short term it isn’t useful except for when storms are hitting.
HWO- Hazardous Weather Outlook. This is issued by the National Weather Service typically days in advance of storms to highlight potentially dangerous weather in the future. These typically do not make headlines, and just highlight potentially dangerous weather sometime in the future. I will reference them on Twitter when they issue them, as they affirm that I am seeing similar things to the National Weather Service.
mb- While mb (millibars) is a measure for pressure, this is also used to refer to different levels in the atmosphere. The most common level I refer to is 850mb, which is at 5,000 feet, 500mb (close to 20,000 feet) and 925 mb (around 2,500 feet). Refer to the bottom for details on each.
MJO- Madden/Julian Oscillation. It is divided up into eight different octets, and each octet results in different conditions around the world. It is mainly based off of where convection is most prevalent, and depending on where it is it can often signal stormier patterns, colder patterns, or dry/warmer patterns. Little other research has been done except to see the impacts that the different octets can bring. Octets 1 and 8 in winter months are most favorable for snow and cold, while octets 4 and 5 are warmest and usually lack snow. More details can be found here.
MOS- Model Output Statistics. MOS readings from weather models are helpful in determining temperatures at the surface given on models, and are useful to help predict high and low temperatures. I do not use MOS often in forecasts, but I do reference them occasionally. Details can be found here.
MREF- Medium Range Ensemble Forecasts. This is another way to say GEFS, essentially, as it is the only ensembles readily available with individual members in the medium term. Technically, the CMC and NOGAPs ensembles apply too, but generally this is used in place of GEFS. They can be seen here.
NAM 4km- The North American Model 4km resolution is a model that goes out 60 hours and is a high definition model. It comes out in both a 4km resolution and a 12km resolution. A 4km resolution means every 4 kilometers there is a data point that the model projects a forecast at, which means it is extremely high resolution. It forecasts specific snowfall amounts, winds, temperatures, etc. and can be important in forecasts, though often overdoes rainfall and snowfall. It can be found here.
NAM 12km- The North American Model 12km resolution is a model that goes out 84 hours and is also high resolution, but not to the level of the shorter range higher resolution NAM 4km. This model is only really accurate inside of 48-60 hours, which is when it is used most, and also can overdo snowfall or rainfall with storm systems. However, its higher resolution makes it quite accurate when inside 48 hours, when it gets weighted heavily in forecasts. It is run four times a day, 0z, 6z, 12z, 18z, though 0z and 12z are the two times it has fresh data. It can be found on the Penn State Ewall here.
NAO- North Atlantic Oscillation. This is arguably the most important teleconnection for forecasting winter snow storms along the east coast. When in a negative phase, there is a ridge near Greenland creating blocking and a storm path typically favors storms moving up along the east coast, throwing back snow into the area. When positive, more options are on the table, with more storms likely resulting in rain. While one of many different factors in a storm, a pattern with a negative NAO typically tries to surprise the area with snow. NAO forecasts and additional discussions can be found here.
NHC- National Hurricane Center. I don’t really ever mention them in winter, but in summer when tracking tropical systems they produce numerous useful products and also produce data showing the current strength of tropical systems. To view the NHC homepage, just click here.
NMM- Nonhydrostatic Mesoscale Model. This is similar to the ARW in that it is a very high definition 4km model that comes out twice a day, at 0z and 12z, and runs out only 48 hours. I have questioned its accuracy in the past as, similar to the NAM, it tends to slightly overdo precipitation for the area, but it is good for temperatures and has a useful simulated radar as well. I only reference it when storms are imminent since it only goes out 48 hours maximum. All runs can be found here.
NOGAPS/NAVY- This is the Navy’s operational weather model, which is in a similar camp as the DGEX in that it preforms very poorly in winter months. It was created to predict tropical systems, and thus handles storm systems in the winter awfully, typically being far too progressive and showing nothing on the east coast when large storms are likely. It very rarely gets a storm correct, but sometimes I reference it to show which ways models are trending, and if it is less progressive than another model that can also be a red flag.
NWS- National Weather Service. This is the government agency that issues Watches, Warnings, Advisories, HWOs, SWSs, etc. I reference them a lot and use their forecasts to compare mine to when making a forecast for SW CT to see who is accurate more often, and where the other goes wrong.
PNA- Pacific/North American Pattern. This is the final teleconnection that I use in forecasts, and it also plays an important role in temperatures, storm tracks, and the overall pattern. A positive PNA means there is ridging near the Pacific Northwest, which sets up a pattern with troughs in the east helping storms move up the east coast and throw back more snow. A positive PNA, especially when paired with a negative AO, is what can set up some of the coldest patterns for SW CT. A negative PNA results in a more progressive pattern typically with less cold air, resulting in storms along the east coast that typically are not as strong and/or do not have as much cold air, so SW CT sees more rain/sleet in them, especially near the coast. Of course, these are generalities, but that is what the PNA is generally used for. PNA forecasts can be found here.
PV- Polar Vortex. This is a large vortex of the colder air, and the placement of this typically dictates where the pattern is going to bringing colder air as well. A Polar Vortex up in the GOA (Gulf of Alaska) or anywhere around that area typically keeps the cold air bottled up there, while a PV further southeast in central Canada or even further south results in more sustained cold weather for SW CT.
Radar- This is the main tool to see current precipitation following across any region. When storms are occurring, I will reference the radar often to see what precipitation is moving this way and what is already falling. The radar I use the most often and find the easiest to use with the most options is here.
RAP- RAPid Refresh Model. This is similar to the HRRR in that it is very high resolution but only forecasts out for a short amount of time. The RAP only forecasts out 18 hours from when it begins, so it is only used right before storms hit or as storms hit. It is weighted similarly to the HRRR in forecasts, and as it is so short range it is typically fairly accurate. I continue to use it and the HRRR even when nowcasting. It can be found here.
Ridge- A ridge is essentially the opposite of a trough. Think of them similar to this: ^ where they make an upside-down V or an upside-down U in the atmosphere. The opening to the south allows warmer air from the south to flow up, while areas east of the ridge typically get cold air as it slides down the east side. A ridge over the Pacific Northwest with a positive PNA is thus beneficial as cold air flows down the east side of that ridge into the United States, but a ridge over the eastern United States allows warmer air to flow up from the south and pushes storm systems up over it and to our west, meaning we are in their warm sector and get rain and fog.
RUC- Rapid Update Cycle Model. Just this past year this model was upgraded to the RAP, so it is no longer seriously used, but I sometimes consult other versions of it and will reference them if I can find them.
SREFs- Short Range Ensemble Forecasts. These are ensemble members of the NMM, ARW, and other short range models, and they are then averaged together to find a mean. They go out the same amount of time as the NAM does, so they are often used together in forecasts. They are not very accurate until they get inside 60 hours, which is when they begin to get more so. They forecast percentage chances of different snowfall accumulation amounts, precip types, winds, temperatures, etc. so they provide me with a lot of different data. They can be found here.
SWS- Special Weather Statement. This is typically issued by the National Weather Service when they expect dangerous conditions in the short term, but conditions likely are not bad enough to warrant an Advisory. For example, if only an inch or two of snow is falling a Special Weather Statement would be issued instead of a Winter Weather Advisory.
Trough- This is a feature that I mention most at the 500mb level in the atmosphere. When a trough is over the east coast, it creates almost a U with the east coast in the middle, resulting in cold air being able to flow down from Canada into the east and storms typically ride on the east side, which is why troughs most often result in east coast snow storms. This is a very simplified explanation of a very complex meteorological process at 500mb, but troughs over the east coast typically are beneficial to snow and snow storms. If I am talking about troughs in other areas, just remember they are U shaped and I should explain the rest in the blog post.
UKMET- This is the United Kingdom’s Meteorological Model, which is in many ways similar to the ECMWF, and the UKMET is not very high resolution nor does it present many details like other models, so it is never used very often. I with reference it typically just to add in another possible solution, and if it stands out from the other models it can be noticeable, but it does not preform exceptionally well. It is run twice a day and only goes out 144 hours, so it is short to medium range with no long range component. It is run twice a day, and the 0z run can be found here with the 12z run found here.
Vorticity- Energy at the 500mb level. See 500mb for more details.
Warning- This is issued by the National Weather Service when dangerous to life-threatening conditions are expected within the next 24 hours. A Warning is issued for a snow storm when 6 inches of snow are expected within 12 hours, or 8 inches within 24 hours. There are warnings whenever life-threatening weather of any sort threatens, and these should always be taken with the utmost seriousness. I will post the details of any warning issued.
Watch- This is issued by the National Weather Service to indicate that a Warning conditions are possible within 48 hours, but we are not within the time range yet to issue or a Warning. Warnings are whenever life-threatening conditions are “imminent” or, for winter storms, within 24 hours, while Watches are when they are “possible”, or, for winter storms, within 48 hours. While not as serious as Warnings, these typically can be followed by Warnings.
12z, 0z, etc.- These are timing signals for models, essentially done in military time with 24 hours. The main model times are 0z, 6z, 12z, and 18z, though most models come out at 0z and 12z. New data is taken officially every 12 hours, as 0z and 12z, which is why those are the two model runs with fresh data. The timing is actually that in London, where the first weather balloons were launched and timing needed to be standard across the globe for weather. Thus, 0z is midnight in England and 7 PM during winter in SW CT. 12z is noon in England, 7 AM here in SW CT. Models come out anywhere between 2 and 6 hours later than the timing in z shows they will, but the time stamp just shows what data was input.
500mb- This level in the atmosphere is just below 20,000 feet, and this is where models measure energy, sometimes referred to as vorticity. This energy is what drives overall patterns, and this is also where ridges and troughs in the upper atmosphere are found. When delving deeper into meteorology, one finds that it is the 500mb setup that drives the overall pattern and dictates storm direction, which is why I refer to this level so much. It does take years to get accustomed to the different complexities at the 500mb level, and this is where blocks can be observed as well. I tend to try and explain exactly what is going on at this level in the atmosphere whenever I use 500mb charts, but troughs in the east allow colder air down and support snow storms off the east coast, while ridges on the east coast allow warm air to flow up along the east and push storms to our west, resulting in rain.
850mb- This level is one where all models output temperatures, and it is around 5,000 feet up in the atmosphere. As a VERY general rule, whenever temperatures at 850mb are below freezing, snow or sleet can be expected at the surface. If surface temperatures are significantly warmer than 850mb/5,000 feet, then precipitation can fall as rain, but the 850mb freezing line typically is very important when predicting precipitation types, which is why I reference it so much.