Advice on Forecasting

from Dennis Eckert


Fri, 14 Jan 2000

I do soaring forecasting for Region 8 comps (Pacific NW), and perhaps the following points might be useful. Looking at various web sites to get data appears to be only half the challenge. The other half is having a method by which to analyze all the data to make actual forecasts. There are some web sites at some of the universities in the USA that have posted course work for their various undergraduate meteo courses (Texas A&M among others), and these include various exercises. So far, the best and latest non-web source of how to interpret weather for soaring conditions (IMHO) is Tom Bradbury's Meteorology and Flight (2nd edition). One excellent part of Tom's book is near the end where he summarizes a list of weather features most favorable to thermal flying (IMHO, he also has the most up-to-date results of mountain wave research presented in a very readable format). To round his approach out, Dennis Pagen's Understand The Sky really adds a lot as it focuses on how to read terrain and conditions, and is thus more aimed at certain strategic/tactical actual flying.

I'm self taught, and did have a great opportunity to audit the senior-level last undergrad course taught at the Univ. of Washington (Seattle). This forecasting class tied all the previous under-grad courses together and had some of the key following elements:

-- a one-hour lab each day where we were trained how to draw on a map of the USA isobars, cold, warm, occluded fronts. Doing this every day taught us how systems move, etc.

-- a 2nd one-hour lab each day where we had to forecast for four cities.

The key here was not only forecasting for each city a lot of data but also being scored on the data, and thus it served as a contest of sorts.

Many of the univerities in the US who have good meteo programs have these forecasting contests and there is even a national one where each university is ranked. Another key to this lab was how to interpret upper-air charts. For example, the upper air sounding upwind of your soaring location may look really unstable when using the forecast high for the day, but since that sounding was only a snapshot of the atmosphere very early in the morning, you have to know evaluate 'advection' (is warmer or colder air going to move into the area and how will that change conditions).
-- and then a one-hour lecture three times a week which presented a lot of detail on the challenges of forecasting and also the various computer models. Hard to summarize this lecture but it was fantastic. It's where we learned how difficult it has been historically to forecast precipitation, especially heavy precip that goes with strong convective events (thunderstorms, etc). And we learned the limitations of each computer model. Another great feature of this lecture is that we learned the particulars of our area, and the typical kinds of weather events that dominate, and what the forecaster looks for in trying to decide whether such events will occur. For example, we have a convergence zone in Western Washington that can occur, and it usually is good for creating thermals, but knowing whether it will occur or not is the challenge. If you soar downwind of a mountain range, then you're always wanting to know whether a 'push' of warmer, moister area is going to move into the soaring area.

Overall, this class taught me a lot, but primarily that forecasters use a step-by-step approach to arrive at a forecast. To be really successful then, the forecaster has to use this 'standard' approach and practice it and then examine the actual results to understand what she/he missed. As noted in other posts to this topic, a forecaster's discussion is posted typically at each city's NWS web site and is called the Forecast Discussion. They can be challenging to read/decode at times but they do provide a summary of what the forecaster is seeing, and this discussion forms the basis for each of their forecasts (whether aviation related, agricultural, fire weather and public weather).

One good way to get started is to keep track of what conditions exist when soaring is good and when it isn't, and you will definitely notice certain patterns. For example, when a fresh supply of cold air moves into the area. This means you need to interpret the upper-air plots at 850 and 700 millibars to look for this 'cold air advection'.

Following on what I posted, I wanted to point out that -- other than all their other tools, like computer models, etc.. -- forecasters have long had a set of documents used locally (and more of them getting on the Internet) to help evaluate certain local forecasting challenges AND big events that sometimes catch them by surprise.

so, first ...http://www.wrh.noaa.gov/ and look at the Technical Lites and Technical Attachments ... there are some interesting articles there on how forecasters evaluate events, like what we are interested in, although their docs are not focused on soaring ... but, want to learn about how forecasters evaluate convergence zones in SoCal, and you can find papers there...

Dennis Eckert