methods of weather forecasting

Forecasting methods we use at The Weather Center… Before I get to that…. Long before someone can really forecast the weather you need to have an understanding of the atmosphere and its physics, along with SFC terrain as terrain can and does effect the forecast on a local scale.. Ok lets talk about some of the methods we use to come up with our forecasts.


1. Persistence method….. This method assumes that the current weather pattern won’t change through the forecast cycle ..This method will bust if the weather patterns are changing rather fast… So this method is not the best in the short term forecast, however does work out rather well in the long term forecasts, out past 10 days and even month to month forecast….

2. Trends method….. This method involves mathematics in determining the speed/direction weather systems are moving… This method works rather great for nowcasting (now to 2hrs out in time) for an example lets say there is a severe thunderstorm west of the area by 60 miles… We have figured out the storm is moving East at 30 MPH, this would put the storm in the area in about 2 hrs…. Caution has to be used with this method in nowcasting as storms can pick up speed and changed direction of travel.. This method is rather well for short term forecasting a day or two if the weather pattern is not being blocked.

3. Climatology method…… This is another rather easy method to use, however do require a little more time… As one will be looking at years of weather data for a day one is forecasting for…. You will look at the SFC OBS/reports from the past then you will need to avg them all out then you have your climatology forecast for that day… Now this method only works if the weather pattern in its self is similar to the data…. This method will bust if the weather pattern is not typical for that time of year….

4. Analog method…. This method is a little more complicated however it is used… What you want to do is look at today’s weather scenario then remembering a day back in time that the same or close to the same setup as today’s…. You will never find the exact match however you will get a close match which will allow you to come up with a forecast… This method is good for detailing larger events…

5. NWP for known as the Numerical Weather Prediction, or as the Computer models…. This method is very difficult to master, there is a lot of information one has to understand which is another whole write up, so for now will touch on some of it…. First off you need to understand each forecasting model and its limitations, along with its bias… You need to understand the gaps in the system as this will play havoc on the forecast...If not all the info is in the computers it will through off the analysis which in turn can/will effect the forecast.. This has been better over the years….One needs to look at all the forecasting models to see which models agree and ones that don’t… One has to also look at past runs to see how much things have changed in between the model its self… One wants to match the SFC OBS/Satellite imagery to the forecasting model analysis to see how close the model is matching up in real time.. This just a very limited amount of things one must know…. Another key factor is one has to look at all the computer models data, not just the temps and precip data…. One needs to looks at all levels to see what is taking place higher up than the SFC…. Knowledge is the key factor when interpreting forecast models… One can tell a beginner forecaster from a long time forecaster just by reading their forecast discussions … … Example…. A beginner or one with less knowledge forecast discussion may read like this… Models have a low over CO.. Models don’t agree on the track of the low… Well really that isn’t saying much at all… What its telling me is the forecaster is unsure of them self’s and I wonder how they came up with their forecasts… Now a long time or more knowledgeable forecaster’s discussion would read more like this…. GFS has a 997 MB low pressure over CO..This model is forecasting the storm to move East and so one… Then you will see another model brought up like so…. Nam as the low pressure system a tad North of GFS and a little stronger down to around 994 MBS Nam takes the low more North than GFS… ECMWF also agrees with NAM however ECWMF is a little weaker with the low, as it at 999 MBS…. So will use a blend on the NAM and ECWMF to come up with this forecast…. That is just an example…. If one wants to keep the forecast easy that is great but there should be no reason to talk about the forecasting models then…That will confuse people if the forecaster can’t or won’t explain them..

We do use all the methods talked about above some more than others…. Plus we use other tools like skew, bufkit data just to name a few…