We take pride in getting the forecast right most of the time. Occasionally, we’ll mess up a forecast on a given day and then recover. But this week, starting Monday - more or less - we’ve hit an extended slump.
Of course, we’re not alone. The TV guys/gals have fared similarly and so has the National Weather Service. It’s simply been a complex weather pattern to predict and no one has gotten a handle on it.
Monday through Wednesday, we (and everyone else) predicted high chances of rain, with the possibility of heavy rain and even flash flooding. Reagan National’s airports total rainfall during that stretch? 0.02”. Dulles? 0.02” BWI? A not quite ark demanding 0.46”.
Then today, we forecast just a slight chance of afternoon rain, and got soaked in the morning.
What’s the problem?
It all starts with Isaac’s remnants. The idea behind the heavy rain potential was that Isaac’s moisture was going to hook-up with a meandering front over the region. Well, the moisture came in - as evident from the offensive humidity levels - but the rain didn’t as the front drifted just far enough away (to the north and northeast).
CWG’s David Streit put it this way in response to a comment this morning:
The main problem was that the true circulation of Isaac never quite got to us as it headed south after covering Ohio and West Virginia with general rains of an inch or better. The moist air from the system did reach us but didn’t get much upper air support to set off thunderstorms until it reached the Delmarva and New Jersey where 1 inch plus was also nearly general.
Essentially, Isaac’s remnants stiffed us. The main circulation split apart, passing to our north and south. At the same time the front it was supposed to “hook up with” jogged far enough to the north/northeast that there was no real focus for Isaac’s moisture locally.
Today’s forecast problem was that the cold front that was supposed to push all of Isaac’s moisture out of here stalled. We recognized that possibility and forecast a chance of thundershowers in the afternoon. But it turned out the air was so thick that it decided to rain itself out this morning with the front hanging around, rather than this afternoon (although some isolated showers/storms remain possible). We should’ve included the possibility of rain at any time of the day.
Why couldn’t we overcome these problems?
For one, the models stunk. The majority of models simulated moderate to heavy precipitation totals Monday-Wednesday in the region (and very little today). Granted, by Tuesday maybe we should have caught on to the model errors and lowered rainfall forecasts. That was a mistake.
Also, forecasting decaying tropical weather systems is difficult. Over time, they become increasingly less defined, making it hard to predict exactly where they’re going to produce rain.
Finally, forecasting the timing and placement of convective, summer-time rain is always challenging. Think about heating a pot of water on your stove and trying to isolate where the bubbles are going to pop up. It’s a head-spinning conundrum.
Let’s hope today marks the end of this streak of forecasting futility and frustation.