Weather Models part 1

Hi It's Rebecca again. Weather Models part 1 will cover the basics of weather models. I will explain what a weather model is and cover some of the models out there. As well as some of their weak and strong points. I've decided to divide the subject of weather models into two sections.

In the previous post;  I went into too much detail on Skew-T's; as a result I fear I may have lost some of you in the details. Therefore, this part will cover only the basics. So that when you hear a certain model name you will have an idea of what it is and how it works.  I've been fielding a few questions from some who want to know a little more about how this stuff works.  So for those adventurous souls, in the second part of this installment, I will go into a little more detail. The section will cover things like 500mb and 700mb Heights, Vorticity, 850mb Temperatures, and 1000mb to 500mb thickness, and maybe a few other things. I will give you the fundamentals on how to read a model to make a forecast. because winter is around the corner (yes it's still Fall in spite of the past nor'easter). I will focus on winter forecasting.  I hope this is acceptable for everyone. It's my hope that if I tackle it this way,  you can read the section you want.

What Are weather Models?

Weather models are computer generated forecasts of how the current or initial state of the atmosphere will change over a period of time. There are many different models, and they all have their uses. Some computer models, such as the Global Forecast System (GFS) run forecasts out to 16 days while other models, such as the Rapid Update Cycle (RUC) will only run for 12 hours.

 How does a weather model work?  

Each computer model will forecast weather for a different amount of time but all models are based on initial weather observations and weather data. This data is then ingested into a powerful supercomputer. The mathematical equation used by the computer involves uncertainty analyses (which is hard to explain in this format) Anyway, computers running these forecasts take in such things as temperature, pressure, humidity, wind speed and direction, etc in an attempt to represent the current weather conditions, then process the data and output it in a certain amount of resolution (detail) attempting to make a prediction of the future state of the atmosphere. Resolution is based on a grid system that  divides up the atmosphere at the mandatory  levels in the vertical, from the surface  to the stratosphere. At each level the horizontal spacing between the points is very regular. The closer the points the higher the resolution ( Supposedly, the higher the resolution the more accurate the model)  Some of you might be asking. Then why not make all the models high resolution? The answer, the computer processing power needed is the fourth power of the increase, Therefore if you want to make the grid half as small you would have to increase processor power by 16 times. So you can see based on current computer technology it becomes impossible at a certain point.

The Models.

Mid and Long Range Models:

NAM (North American Mesoscale) This is one of 4 American models. The NAM runs 84 hours into the future. Updates are produced every six hours at 0z, 6z, 12z and 18z,

Strong points:

It normally has a good handle on temperatures. This model is very useful in forecasting winter storms and severe weather that are expected occur within 48 hours or less.

Weak points:

Normally it shows systems moving slower than they actually are. It also likes to show more precipitation than actually accrues. The model has a hard time handling shortwaves.

GFS (Global Forecast System) Model Forecasts are also produced every 6 hours. (00, 06, 12, & 18). Like the NAM they are available in 6 hour increments. The GFS is a longer range model and goes out to 384 hours. However, the current GFS output from NCEP is 3-hourly up through hour 192, after that it then shifts to 12-hourly. There is also a resolution change at the 192 hour time.

Strong points

It's great for looking at overall patterns and temperatures.
Great for finding the conditions for major winter storms.

There are associated ensemble forecasts that run off the GFS operational run that are used for Quantitative Precipitation amounts (QPF)

Weak points:

It's not very accurate out past 5-6 days. It can have a storm in the 11-15 day range, only to "lose" it for a few days and then "find" it again when the storm is 3 or 4 days out.

GFS can miss the short term forecast in bizarre ways, such as the location a front will be six hours from the initialization time.

It has a northern dry bias

Like the NAM it has a tendency to overdo the precipitation a bit especially when the precipitation is light and it tends to over forecast the amount and extent of precipitation in cold air masses.

It has a very large grid spacing

ECMWF (European/Euro) Some of the data from the Euro is available by subscription only. Many meteorologist call the Euro the King, because it's known for  its accuracy rate in the mid-range. The model runs 240 hours into the future. However, it's between the 72hr and 120hr that it is the most reliable. The Euro It runs every 12 hours, at 0z and 12z, and there are no "off-hour" runs.

Strong points

Handles winter weather events very well.
Like the GFS it has its own ensembles.

Weak points 
Doesn't handle shortwave interaction with storms during winter as well as some of the other models.
Forecasts using this model are a bit more complicated

There are two other models I will briefly mention. The first is the UKMET. It is run 4 times a day; the 00z and 12z runs go out 120 hours; the 06z and 18z runs go out through 72. Most of the time it's used to collaborate with other models during the winter. The second is the GGEM. This is a Canadian model; it runs twice a day out to 144 hours. A major problem with this model is it has a warm and wet bias

Short Range Models:

RUC (Rapid Update Cycle) This model updates every three hours beginning with 0z, 3z, 6z, etc. It takes current surface readings add any changes into the model.  It is a high resolution model that's  intended to forecast looking out for 12 hours; because of this it's very valuable for now casting.  Because of its rapid updates and higher resolution it's great for severe and winter weather forecasting.  

HRRR (High-Resolution Rapid Refresh) The HRRR is a 3-km resolution, hourly updated, cloud-resolving atmospheric model. Its enhanced Rapid Refreshing capability replaced the Rapid Update Cycle (RUC) earlier this year. This model has a much higher resolution than the RUC. The HRRR "looks" farther out into the future than the RUC.

What the heck is Z time

The forecast hours on weather models is  UTC, which is Coordinated Universal Time; this is also called or Zulu time (That's where Z comes from) which is military time, or GMT, Greenwich Mean Time.

0z is midnight at the Greenwich Meridian and 12z is noon.


Here is something to help you convert Eastern time to Z time.

 

Eastern Standard Time (winter) is 5 hours behind UTC
Ex 12z = 7am
00z= 7pm (Remembering that 00z is actually midnight)

During Eastern Daylight Savings (summer) is 4 hours behind the UTC
Ex 12z = 8am
00z = 8pm

How to read a weather model

Here are the basics. I will  focus on the GFS and NAM.

As I said above the GFS is a long range model, going out to for 16 days. whereas, the NAM is used in the shorter term going out 84 hours. On each one you will see OZ, 6Z, 12Z and 18Z. you will have to check to see what your local time is and then convert it to Z time.... Z time  is the time they run the model, each runs four times a day.

 

Now you will scroll down the page and click on an hour. My  advice is to click on fine and then  select the 850 MB and 6 hr Precipitation.  A map should then appear.

Now you will see a lot of  graphics on the map. Don't let the graphics scare you.  The graphics will give you all kinds of information. However,  you are most likely looking  for a storm. So look for green. The amount of precipitation is broken down by color.  The colors move from light green to blue to purple. Light green represents light precipitation and Purple represents very heavy precipitation.

If you are looking for a snowstorm, you must look for the O line. It will be a blue line labeled 0. The color blue represents cold air. If any blue line is to your south and you see precipitation it will likely be something in the frozen state. The orange line represents warmer air and means rain.

To give you an idea of what a model looks like here is the 1000-500 mb thickness from todays GFS. I added an arrow that points to the 540 0c line.

Here is a link to the e-wall Penn State site.

http://www.meteo.psu.edu/~gadomski/ewall.html

Well that's about it.....This should give you an understanding of what a weather model is and what their used for.........I will be back with part 2 Monday or Tuesday.


Rebecca.