Weather forecasting is like a game of telephone (!)

This blog entry is part 2 from the INFORMS International Conference in Puerto Rico. Read part 1 here.

Being from the Midwest, I have a natural curiousity about the weather because, frankly, Midwesterners can’t seem to stop talking about the weather. It’s a stereotype that is very, very true. The talk entitled “Doing Something about the Weather” given by Eva Regnier from the Naval Postgraduate School did not disappoint. I would characterize the talk as an autobiographical account of Dr. Regnier’s foray into the wild and crazy world of weather prediction and meteorology. Everyone in the audience was riveted — we had no idea how badly weather forecasting needs us! Looking at weather forecast prediction as a supply chain, a number of components (historical information, observational data, numerical predictions, post-processing, forecasts) interact with each other in not-so-friendly ways to give us the weather report we see on the evening news. Information is handed off along the supply chain, but the bandwidth is very, very low. For example, at one point, all the numbers crunched by the supercomputers and smushed into a single piece if information (literally, the color Red, Yellow, or Green), which is then handed off to the Federal government for weather disaster instructions. As you can imagine, much information is lost along the the other links as well. In fact, the wikipedia entry for weather forecasting explains all of these components, but does not explain how they interact. Weather forecasts do not truly answer the question “Should I take an umbrella with me when I leave my house in the morning?” in a systematic way. In addition, there are two links of information bottlenecks: meteorological information and decision information. Weather forecasting is like a game of Telephone!

What was particularly surprising is that almost no undergraduate degree programs in meteorology require basic statistics, which makes you wonder about the statistics that they bombard us with during every single weather forecast (However, I wonder if this is the case with PhD meteorologists).

Now, I’m not using my blog to badmouth meteorologists (I am actually related to a meteorologist), because the science in each component of the supply chain listed above is certainly first-rate. It’s the connections that make or break the system, which is more motivation of why OR needs to engage the rest of the world. I’d love to hear a meteorologist’s perspective on this topic (if only to tell me if I need an umbrella today).

Weather Forecasting Guide
National Weather Service

One response to “Weather forecasting is like a game of telephone (!)

  • Barrett

    Krzysztofowicz at Uva works on some of these issues. I only know this because he is a few doors down. I would like to hear the names of some of the other OR types who were mentioned at the conference.

    Never thought about the information compression aspect of weather forecasting. Mapping a High dimensional parameter space into three elements..brilliance.

    good post.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: