Wednesday, 28 March 2012

The Classic Windstorm: Perhaps More Likely This Autumn?

The track of the 1962 Columbus Day Storm is an excellent example of the "Classic Path."

A classic path windstorm is one that develops in the lower midlatitudes, say south of approximately 42.5ºN. Maybe this number could be pinned at 40ºN, or perhaps 45ºN. The exact location is not particularly set. As the low deepens and crosses the 130ºW line, it then tracks nearly due north off the Pacific Coast of North America. Ultimately, the extratropical cyclone lands on Vancouver Island. These northward-trending lows, staying offshore, tend to retain their strength since the centers do not strongly interact with rough terrain until the final encounter with The Island. As the weather system sweeps northward, strong and damaging winds, primarily in the southeast quadrant of the storm, sweep the entire coast and inland region from Northern California to Southwest British Columbia, causing widespread damage and affecting millions of people.

As of the time of this writing, the last true "classic path" windstorm occurred on 16 Jan 2000. This cyclone seems relegated to a lesser position—in other words is often forgotten—than events like 14 Nov 1981 and 12 Oct 1962 (AKA Freda or the Columbus Day Storm). I believe the tendency to forget the tempest of 2000 is mainly because the central pressure did not get as fantastically deep, roughly 98 kPa vs 96 kPa, and strong winds did not quite cover as large a region. However, peak wind speed and gust during the 16 Jan 2000 storm at many Willamette Valley locations from about Salem north were comparable to 14 Nov 1981.

Therefore, it has been about twelve years since the last significant classic path event.

Since 16 Jan 2000, big Cascadia windstorms have often tracked more ENE to NE than NNE to N. This includes the major Hanukkah Eve storm of 2006. That event produced the strongest wind and gust speeds yet recorded at Sea-Tac (records go back to 1944), and caused phenomenal tree damage to parts of Vancouver, BC. The late-season 02 Apr 2010 windstorm ultimately did hook on a nearly due north track, but after it had tracked NE, landed on central Vancouver Island and then filled rapidly. The strong 03 Mar 1999 windstorm also fits into the NE-tracking category of event. And the 14 Dec 2001 cyclone tracked just north of due east. NE tracks appear to be far more common than N tracks.

It seems like what is required for a strong classic path event is a narrow upper-level (50 kPa, or approximately 5 km up in the atmosphere) trough situated between 130ºW and 140ºW. Ideally, the trough is slow moving. If the upper trough moves rapidly, developing cyclones are often thrown ashore before they can hook on a more northward track up the trough's east side--in other words, faster upper-level troughs tend to support E to NE tracks, whereas slow-moving troughs tend to support lows that eventually move on a more northward trajectory. There are some other variables at play, such as the breadth of the trough, but motion is a key one.

We had just such a slow-moving trough on 12 Mar 2012. Over the course of 24 hours, as the surface cyclone spun up and deepened at a rate in excess of 1 Bergeron (central pressure dropping ~24 hPa in 24 hours, a phenomenon known as explosive cyclogenesis), the upper trough only shifted 5º further east. The cyclone quickly ran up the east side while still over the ocean, with decent jet support up to nearly the point of landfall at Solander Island on the north shore of Vancouver Island. This low, incidentally, took on numerous characteristics of the Columbus Day Storm, including extreme pressure gradients near the center and highly unusual wind speeds at locations near the track.

Because slow-moving troughs are an important ingredient, major global blocking events have the potential to support the development of classic windstorms. We had a significant blocking event earlier this month, one that contributed to extremely unusual warmth in parts of Eastern North America, and appears to have influenced the slow-moving trough that supported the northward-trending 12 Mar 2012 windstorm on Vancouver Island. If a similar blocking event returns this fall, in essence a long-term persistence forecast, then, indeed Cascadia could be under the gun for a major classic event.

Another phenomenon supporting the possibility of a classic windstorm in the fall of 2012 is that the El Nino Southern Oscillation (ENSO) appears to be shifting from a La Nina (cold-phase) pattern to an El Nino (warm-phase). Models suggest that ENSO may be in a neutral state by the fall. It has been noted by some meteorologists, such as Cliff Mass in his book "The Weather of the Pacific Northwest" (2008), that available evidence suggests that many of the biggest Cascadia windstorms have occurred during an ENSO neutral state.

It appears that conditions may be in place this fall that supports a greater likelihood of a major classic windstorm event. An extreme windstorm, of course, is not certain, but conditions may make one somewhat more likely.

The period from October to December is a key time for major windstorms. Pacific sea-surface temperatures still tend to be warm, relatively speaking. The Columbus Day Storm, for example, developed in the vicinity of 21ºC (70ºF) water off of Northern California. Extratropical cyclones, being storms formed via baroclinicity (strong temperature contrasts), do "feed" off the latent heat in our oceans, and a good source of warm water can help significantly energize a storm. Even in December, with the shortest sun-up times in the northern hemisphere, the Pacific still retains some of the heat from the high sun of the previous summer. On land, people experience this Pacific warmth in part via strong cyclonic warm sectors, with temps climbing to and exceeding 15ºC (60ºF) at times after the leading warm front has moved through. Given such warm air, static stability tends to be low, supporting the presence of downdrafts. These downward currents help mix down momentum from the upper-level airflow, making warm sectors ripe for strong wind readings. The rare Autumn windstorm can bring temps in excess of 18ºC (64ºF) while a gale is raging outside. Later in the storm season, say January through April, Pacific sea-surface temperatures have cooled considerably, and surface warm advection fields also tend to be cooler in response, though a strong storm with deep tropical feed can still bring some strikingly warm air. Colder temps in the north help provide continued baroclinicity to keep cyclones going, a much colder form of storm, with warm sectors that may not have the same degree of instability exhibited during Autumn storms, potentially reducing high-wind potential.

These are all general thoughts. Exceptions abound--one of the key reasons why Nature is so interesting.

To me, the Fall storms are often the most spectacular. This is in part because deciduous trees are still leaved, providing much more drag against the wind. Trees have also had a spring and summer of lighter wind speeds in which to accumulate new dead and weakened material. When an Autumn gale strikes, much of the accumulated dead material falls. Leaves, twigs and branches are torn off in droves, making for a spectacular sight, a blizzard of flapping foliage and a hail of dropping woody debris.

Tuesday, 27 March 2012

The Vancouver Island Windstorm of 12 Mar 2012

Visible satellite photo of the 12 Mar 2012 extratropical cyclone taken at 1030 PST, just after peak winds at many locations. The centre is right over the northern tip of Vancouver Island. Image courtesy of the US National Weather Service.
A factor that makes the 12 Mar 2012 extratropical cyclone stand out is that it landed at, or very near, peak intensity. Many cyclones arrive on Vancouver Island in a mature and fading state. These dying lows can still deliver a good gale, but cyclones at peak intensity have a greater capacity to deliver a major windstorm.

Using hourly pressure data from Estevan Point (CWEB), Solander Island (CWRU) and Port Hardy (CYZT), changes in the sea-level pressure gradient can be followed as the storm swept over Solander Island. As the low neared, the gradient rapidly escalated to 18.3 hPa/100 km by 0900 PDT, with a pressure slope orientation of 76º.

This magnitude of gradient is phenomenal. Many storms of history--such as 03 Mar 1999, 14 Dec 2001, 15 Nov 2006, 15 Dec 2006 and 02 Apr 2010--have not produced a gradient at this level. The steep pressure slope of the 12 Mar 2012 storm is very reminiscent of Freda (AKA the 1962 Columbus Day Storm). For regions affected by a given windstorm, most appear to generate maximum gradients in the range of 5-10 hPa/100 km, and a gradient of 4-6 hPa/100 km is often enough to deliver a gale. The next strongest event behind the 12 Mar 2012 storm, 03 Mar 1999, produced a 14.7 hPa/100 km maximum gradient along the North Washington coast; the 15 Dec 2006 event brought a peak 12.2 hPa/100 km gradient to the same region.

An 18.3 hPa/100 km gradient at the latitude of North Vancouver Island corresponds to a powerful geostrophic wind magnitude of 470 km/h. Of course, at the surface, this kind of speed is not likely to be realized due to turbulent drag effects (a key assumption of the geostrophic wind is that it is non-accelerating and largely free from friction). However, if the surface wind speed reaches just 25% of the geostrophic potential, it is still an incredible 118 km/h.

The intense gradient only extended for a short distance, say approximately 100 km from the centre of the extratropical cyclone. Outside of this, a strong gradient existed for a few hundred km more. Using pressure data from Victoria (CYYJ), Comox (CYQQ) and Squamish (CWSK) on 12 Mar 2012, the gradient peaked at 0800 with a value of 7.2 hPa/100 km. Nothing like the intense reading near the storm center, but enough to deliver a memorable windstorm.

Of interest is that the timing of peak winds at a number of locations corresponds well to the time of peak gradient.

In response to the diminishing pressure gradient southward from the low centre, the peak wind magnitude of the storm diminishes relative to the climatology. For example, at some selected stations here is the time interval between the 12 Mar 2012 windstorm and the last time winds of comparable magnitude occurred:


  • Campbell River: ~40 years [1]
  • Comox: ~27 years [2]
  • Victoria: ~5 years
  • Vancouver: ~2 years


[1]
Station operates from 0600-2100 LST, so some big storms may have been missed.

[2] A peak gust of 133 km/h occurred on 23 Mar 1989. Probably measured on an analog U2A anemometer when "instant gusts" (highest value indicated on a direct-reading dial or strip-chart) were recorded, and the indicated speed is probably comparable to the 117 km/h 5-second average measured on 12 Mar 2012.

Here are the corresponding peak wind, a 2-minute average, and gust, a 5-second average, speeds during the 12 Mar 2012 windstorm:



  • Campbell River: 74 km/h (40 knots) gusting 106 km/h (57 knots)
  • Comox: 96 km/h (52 knots) gusting 117 km/h (63 knots)
  • Victoria Int'l: 67 km/h (36 knots) gusting 95 km/h (51 knots)
  • Vancouver Int'l: 65 km/h (35 knots ) gusting 89 km/h (48 knots)


The maximum wind speed at Comox is incredible. Winds of this magnitude can have a terrific effect on trees, with great potential for swathing forests in a phenomenon termed windthrow. Structures can be physically damaged by the raw wind force: shingles peeled from rooves, awnings hammered apart, garage doors punched-in. Interestingly, the ratio of gust/wind is pretty high for most of the listed stations, with the exception of Comox. This kind of response is, like the intense pressure gradient, reminiscent of Freda.