Weekend forecast flub. . . or example of just how tricky Houston weather can be?

Heavy rain Saturday takes area residents -- and forecasters -- by surprise

HOUSTON – It looks like the familiar summertime pattern of isolated thunderstorms, missing for the past few summers, is back.

What happened this past weekend might be an indication of things to come.  After a forecast of 20 to 40 percent rain chances for Saturday turned into a Saturday afternoon wash out from Sugar Land to southeast Houston, we can only hope that more frequent rain continues to pepper the landscape -- without the flash flooding that we experienced, of course!

Recommended Videos



It didn't take long for the social media comments to come flooding in last weekend.  "Sure doesn't look like a 20% rain chance to me!" wrote one person, who also posted a picture of one of the many streets that flooded.

There were many more similar comments.  All were aimed at the meteorologists who offered forecasts for the all-important weekend and suggested that Sunday would be the wetter day.

That's not exactly what happened.  Parts of Houston easily received two inches of rain.  One area, between Sugar Land and Missouri City, received more than 5 to 8 inches!  Why didn't we see it coming?

First, regarding that 20 to 40 percent chance for rain:  Offering a probability, or percentage, of rain has nothing to do with how much rain you're likely to receive.  Bill Read is KPRC's hurricane expert.  He is the former director of the National Hurricane Center and former meteorologist in charge of the Houston/Galveston office of the National Weather Service.  He says, "Probability of precipitation says NOTHING about how much, only the likelihood of occurrence at your location.  Over many years of forecasting, a 30 percent chance of rain for your yard should yield 3 out of 10 rain events and 7 out of 10 non-rain events."

Friday's computer model forecasts didn't provide any indication that severe weather was probable.  It wasn't until late Saturday morning that atmospheric indicators suggested the likelihood of strong storms that afternoon.

I did the forecast for Friday's midday newscast.  Our computer models showed the clear probability of rain in our area at 3pm on Saturday.  It did not show a potential for flooding.  All of the computer models indicated a probability of rain forming along a stalling frontal boundary.  But the question was, "Exactly where, and how much?"

The moisture in the lower levels of the atmosphere provided fuel for the rain.  Upper-atmospheric divergence provided an aid in sparking the storms.  As the afternoon progressed, storms began to "train" over southern Harris County.  That is, a series of storms developed and moved -- slowly -- over the same area.

In Saturday's case it was a period of five to six hours training thunderstorms that produced the 5 to 8 inches of rain that we saw on the southern side of metro Houston.

With the drought that we've been enduring over the past several years, rain events like this are very uncommon.

A rainfall event like Saturday's is a good reminder to us:  When there is a 20 or 30 percent chance for rain in any seven day forecast, there is a good chance many of us won't see any rain.  But that when we do, we could get A LOT!

Remember that forecasting is an inexact science.  The set of data we have to work with is not a complete and perfectly accurate representation of atmospheric conditions, and there are a huge number of variables that can affect short-term weather patterns.  These deficiencies that forecasters have to work with can sometimes mean the difference between isolated showers and flooding downpours.