Most Read: Local

Posted at 02:56 PM ET, 04/26/2012

Rain fail not rainfall: why today’s wet forecast went wrong

Left - rain simulated by NAM model for today last night. Right - doppler estimated rainfall so far today. ( and NOAA)
It happens once in a great while. We get a forecast wrong. From the look of things, today’s going to be one of those days. That quarter-to-half inch of rain we told you was coming - unless some late day showers intervene (slight chance) - simply isn’t.

So what happened? The warm front that we expected to come over the region with an area of rain remained about 100 miles too far south.

The image above illustrates how we were led astray. The NAM model from last night showed a swath of steady rain from southern Virginia to the south side of Baltimore today. It correctly simulated heavy rain over central Virginia, but overdid the northward extent - significantly. The GFS model, not shown, actually did a better job simulating the rain shield, but also overdid it.

Top - warm front position forecast at 2 p.m. by National Weather Service. Bottom - actual warm front position. (National Weather Service)
We probably expressed too much confidence in the morning to midday rain scenario. Trying to predict the exact location of a front is risky business. That’s why the weekend forecast is so tricky. It’s another situation where we’re trying to figure out exactly where a front is going to set up.

The difficulty in trying to get a 12 to 24 hour forecast right under these circumstances demonstrates the lunacy of AccuWeather issuing 25-day forecasts.

Let me use this analogy: getting today’s forecast wrong was like a football kicker missing a 35 yard field goal - unusual but expected from time to time. Trying to forecast weather systems like this 25 days out is the equivalent of trying to kick a 99-yard field goal - virtually impossible...

By  |  02:56 PM ET, 04/26/2012

Categories:  Latest, Weather Checker, Science

Read what others are saying

    © 2011 The Washington Post Company