The Northern California Fires: Driven by the Diablo Winds That Were Predicted Days Before

A large area north of San Francisco was devastated Sunday night/Monday morning by explosive wildfires that have killed at least 13 individuals, destroyed over 1500 structures, and burned over a hundred thousand acres. Over one-hundred people are missing.

Picture courtesy of KRDO.

This sudden catastrophic event eclipses the damage of the highly publicized Hurricane Nate, during which no person is known to have lost their life.

As we shall see, the northern California wildfires were produced by the rapid development of strong winds, which gusted to 50-70 mph in places.   Importantly, the winds were highly predictable, being forecast by current operational weather models days in advance.

Here are the max winds (in knots)  north of San Francisco during the event (provided by UW grad student Conor McNicholas). Some locations had winds over 55 knots (orange and red) and lots of places reached 30-45 knots (green colors)

High-resolution NASA MODIS satellite imagery show the explosive development of the fires.  Around noon on Sunday, October 8th, California is clear, with little evidence of smoke.

 One day later, massive smoke plumes are moving westward from a series of fires north of California.

The key element of this event was the rapid development of very strong offshore (northeasterly) winds, with gusts to 40-70 mph, and rapidly declining humidity during Sunday evening and the early morning hours of Monday. 

To illustrate, here are the observations at Napa Valley Airport (KAPC) from 1154 PDT (1854 UTC) Sunday through 2:24 AM PDT on Monday.  Temperatures early in the day were in the 70s F, with southerly winds and moderate dew point (upper 40s). But subsequently the winds switch to northeasterly, the dew point dropped into the teens (very dry) and the winds gusted to 35-40 knots (40-46 mph).

Why did dry winds pick up so rapidly?  Ironically, it was due high pressure, associated with cold air, passing to the north and east of northern California.

Looking at National Weather Service large-scale pressure analyses, one can view the changes.  The lines are isobars of constant sea level pressure.  On Saturday at 5 AM, there was high pressure over the Pacific and a cold front was moving into the Northwest.  Everything was fine in California.

By 5 AM on Sunday, the front had reached northern CA and high pressure was pushing eastward over Oregon, Idaho, and northern Nevada.  A trough of lower pressure was beginning to develop over coastal CA.

By 2 AM Monday morning, the world had changed, a very strong pressure difference had formed over northern CA, as high pressure pushed east and southward to the east of the Sierra Nevada and a trough of lower pressure intensified along the coast.  There was a big change of pressure with distance, which meant strong winds.

The UW WRF 12-km forecast of wind gusts and sea level pressure for 2 AM Monday, shows a huge pressure gradient over northern CA, with areas of strong winds. Very strong winds over the eastern Pacific as well.

And the humidity forecast for the same time shows very dry conditions are northern CA (dark brown color).

A higher resolution (4-km grid spacing) model run by the CA CANSAC group at the same time shows powerful northeasterly sustained winds (not gusts) north of San Francisco.

So the set up was the following.   Northern California was at the climatologically driest point of the year, after a summer of little rain (which is normal).  In fact, the latest official drought U.S. monitor graphic did not show particularly unusual dry conditions (see below).

High pressure then build in north and east of northern California, forcing strong  offshore (easterly) flow.  As the flow descended the western slopes of the regional terrain it was compressed and warmed (see schematic).  This warming resulted in reduced relative humidity and pressure falls at the base of the terrain (warm air is less dense than cool air).

In fact, the strong downslope wind over northern California has a name:  the Diablo or Devil's Wind. 

The pressure falls associated with the Diablo wind helped rev up the horizontal pressure gradients, and thus the surface wind speeds.      So you had antecedent dry conditions, strong winds, warm temperatures and low humidity--all the ingredients needed for explosive fire growth.

And then we have the other issues:  human initiation of fire, mismanaged local forests and grasslands, and folks living too close to fire-prone vegetation.

This event should be considered a severe storm situation, driven by a well-forecast weather phenomenon.   In fact, the forecast models were predicting this event many days before.  To show this, here are forecasts of sea level pressure and temperature made for 2 AM on Monday started 9 hr before (Sunday at 5PM) and 117 hr before (5 AM Wednesday).   VERY similar and both were predicting the strong winds.  The strong winds and dry conditions should have surprised no one.

Could we have warned people better of the upcoming wind event and the potential for fire blow up?  Could we have saved lives if we had done so?

Should the powerlines have been de-energized when the winds exceeded some threshold?  Was poor maintenance of powerlines (e.g., lack of trimming vegetation) a major issue?

I will let others answer these important questions.