Heavy Rain, Strong Warm Front, Flooding, and Model Problems

The last 24-h has brought some record-breaking rain to parts of Washington, with flooding on some rivers.   And it reveals a problem with our best weather forecasting models.  The cause of all the weather action?  A strong warm followed by an atmospheric river with lots of moisture and strong winds.

First, let's consider some of the amazing rainfall totals (from 1 AM Monday to 5 AM this morning).  Click on the  image to enlarge.

More than five inches at the cost at Quillayute, over DOUBLE the previous record for the date.   6.37 inches near Lake Quinault.  With strong westerly (form the west) winds aloft, much of the Puget Sound lowlands were in the Olympic rainshadow, with some locations only getting about a tenth of an inch.  But as the moist, warm air approached the Cascades, it was forced to rise, resulting in HUGE amounts on the western slopes of the Cascades, with one location getting nearly 8 inches.  Wow.

The action began yesterday afternoon with the passage of a potent warm front that pushed the freezing level to well over 10,000 ft.    If you were in western WA yesterday the warming during the day was pretty obvious--temperatures rising and staying around 60F. 

Below is a plot of the temperatures (red lines) and wind above Seattle Tacoma Airport.  Height is in pressure (700 is about 10,000 ft) and time increases to the left (this is called a time-height cross section).  Look at the 0C line--freezing level.   Heads straight up around 21/18-- 11 AM yesterday.  And if you can read the wind barbs, you can see the winds got very strong from the west--50 knots.

 So behind the warm front, there was a current of strong flow associated with lots of moisture and warm air--an atmospheric river.  To see this graphically, here is a model plot of the total moisture in the atmosphere at 11 PM last night.  Blue and white indicates the most moisture, followed by red.  You can clearly see the current of moisture heading into our region--that is the atmospheric river.

With such huge amounts of precipitation falling in a short period, the rivers had to respond. The latest information from the NWS River Forecast Center in Portland is predicting moderate flooding on several rivers on the west side of the Cascades, like the Snoqualmie.

To illustrate, here is the hydrograph, the prediction of river level, for the Snoqualmie near Carnation.  The river is already at flood stage and soon at moderate flood.  Make sure you don't drive on flooded roads if you live in the area.  But since this a basically a pulse of heavy rain, the river will rapidly settle down.   If you want a show today, head to Snoqualmie Falls.

The National Weather Service has flood watches and warnings out for a number of rivers of our region:

Finally, now the model problem part of the blog.    As many of you know, I am an atmospheric modeler, using computer simulation for understanding both weather and climate phenomena.  I, of course, like to see the models do well, but my most important job and that of UW graduate students and research scientists, is to note model failures and try to fix them.   The model forecast of accumulated precipitation for the event started at 5 PM Sunday (see below) got the pattern quite correct, particularly the huge amounts on the western side of the Cascades and the rainshadow over Puget Sound.  But it underplayed the amounts at Quillayute and the western side of the Olympics.

What is wrong?    A graduate student (Robert Conrick) and I are working on this problem, and we believe there is a deficiency in the physics of warm rain in the model (WRF).   This is the topic of his Ph.D. thesis and I am hopeful that we will make important improvements.  This is where science is both fun and satisfying.