A Disappearing Art in the Science of Meteorology

I was blown away when I realized that this year marks 10 years since I graduated from college with a hard-earned degree in Meteorology. To make me feel even older; I realized that this summer will also mark 15 years with "The Bureau" (unfortunately the first 4 of which apparently don't count towards my retirement date). When I first started back in 1994, the state of The Bureau was considerably different; PCs were just entering the scene, we were still using a computer system known as AFOS that was designed by Ford Aerospace and was VERY finicky, and the NEXRAD radar had just entered the scene. The majority of the weather analysis was done with the MK I eyeball, an assortment of colored pencils, basic printouts on a facsimile printer which could result in a bit of a buzz due to the ink used, and an occasional walk to the window to see what was going on outside right now.

Meteorology is a science but within the science is something that has a bit of artistic value. Our depiction of weather conditions at the surface and at various levels in the atmosphere was and can still be created by doing hand analysis. This involves looking at a "standard plot" of surface data and contouring the important parameters; pressure, temperature, and dewpoint. Other parameters such as 3-hourly pressure change can be plotted as well. Finally frontal positions can be derived from the data and contours. Likewise, analysis of upper-air data reveals information about the structure and layout of parameters at given levels at the atmosphere.

Unfortunately, hand analysis is a dying artform within our science. It has rapidly been supplanted by computerized objective analysis as hand analysis can take considerably longer than a computer chomping through the data. Although quality checks of the data are improving and have aided in the accuracy of an computer analysis, there are still times a bad observation/data point can sneak through and blow the objective analysis out of the water. It boggles my mind to think that there are generations of meteorologists coming out of various programs in the U.S. that are no longer teaching such a basic form of weather forecasting. They are taught to trust the computer analysis and use the "extra" time for outreach, teletraining sessions, or the generic "professional development".

I didn't realize how rampant this problem was until I had an intern walk up to me while I was working on a hand analysis of surface and upper-air data. He asked why I was wasting my time drawing with colored pencils when the computer version was already done and available. My response was with my own series of questions, which I will also now pose to other "young'uns" who might be reading this. Let's say AWIPS systems are down and the only thing you have access to is the internet. I will not pose the most difficult scenario of having to download all the raw observations and make your own surface map, but say that the only sites you can access are those which have the data pre-plotted on their standard models. Would you be able to create a short-term/12 hour forecast based on that data alone? The reply I got was a rambling statement of how the scenario I posed was highly unlikely and that we could always call another office to get the forecast we needed, blah, blah, blah...all to keep from having to admit that he had no clue nor interest on how to do a basic hand analysis.

Call me "old school". Hand analysis is one of the easiest ways for a meteorologist to get a feel for the basic state of the atmosphere and what might happen in the next few hours. A detailed hand analysis can also reveal subtle features that the models might miss by getting their first guess of conditions (known as model initialization) wrong. With almost 15 years of perusing data via hand analysis, certain features start to appear which is known as pattern recognition; current maps reflect a pattern seen with previous weather systems which means the potential exists for the same kind of weather event to unfold. I would venture to say that combining a firm grasp of what is going on currently plus the various forecast models (computer simulations) of what the weather might be can result in a MUCH better forecast in the first 24 hours than looking at a computer model alone.

So, to my mentors from our collective time at the Bureau office in the northern part of the western Metromess; my sincerest thanks for establishing a firm foot in that wonderful art of hand analysis. Al Moller, your retirement is a great loss for the younger meteorologists coming into the Bureau as your guidance in my early stages of learning hand analysis were priceless. My thanks also go to the late Dr. Dusan Djuric, my METR451 instructor who required map analysis and discussion at the start of every lab day and for his textbook which covers the basics of hand analysis as well as more complex items. All those countless days with various colors rubbed onto the palm of my drawing hand and discussion of what those maps showed have helped me perform the most basic part of my job through all these years; forecasting the weather. To all you young'uns who think it is a waste of time - you are only as good of a meteorologist as you make yourself to be. Take the time to learn hand analysis; it will help you out more than you know...