Extreme Weather in McGrath

The heatwave is on the way out, and Fairbanks didn't reach 90°F after all, but McGrath did - and it's a much more unusual event down there.  McGrath tends to have a more maritime influence, being closer to the sea and to the west of the Alaska Range.

With data back to 1941, McGrath has only reached 90°F on 5 days: once in 1969, on 3 consecutive days in 2013 (reaching an amazing 94°F), and once in 2019.  All were in June.  However, July and August have both seen 89°F, most recently on July 8, 2019.

To dispel doubts about the accuracy of yesterday's measurement, the AWOS instruments measured 88°F upstream (on the Kuskokwim River) at Nikolai and 86/87°F downstream at Sleetmute and Aniak.  As usual, RAWS thermometers in the area were higher, but they should be disregarded.

However, the heat is only half the story.  Shortly after the temperature peaked at 6pm, a very intense rain shower developed, dropping over an inch of rain in a very short time.  Brian Brettschneider notes that the rain rates exceeded the estimated threshold for a 1000-year recurrence interval!  Here are the precipitation frequency intervals from NOAA's Atlas 14 (click to enlarge):

Between 7:23 and 7:53pm, the ASOS reported 0.78" of rain (compare to 0.73" for a 30-minute event with a 1000-year recurrence interval), and 1.22" fell in 1 hour 24 minutes.

The notion of return interval is based on historical statistics and the assumption of a stable climate, which is of course never quite true, and we do expect more frequent heavy precipitation events in a warming climate with more atmospheric water content.  Nevertheless, this was clearly an extremely rare event for McGrath.

Putting the two aspects together, the combination of rare heat and rare heavy rainfall was obviously way outside the historical climate record for McGrath.  For instance, the warmest day in the past 84 years with a daily precipitation total of 1" or higher was only 72°F.  There have only been 3 days with a high temperature above 80°F and a daily rainfall above 0.5" (most recently - again - in 2019).

Second Summer

It's all change again in Alaska's weather, with summer returning this week for many places.  The temperature rose into the 80s F yesterday for much of the interior, and the next couple of days will be the hottest of the year for many.  The NWS expects Fairbanks to see 90°F, which has only happened 6 times this century so far (2009, 2010, two days in 2013, 2017, and last year).

Here is yesterday afternoon's 500mb analysis, courtesy of Environment Canada.  Instead of a strong Bering Sea trough dominating the circulation over Alaska, we now have a strong west-east ridge with a connection to western Canada and the western lower 48, where heat and wildfires have been running rampant lately.

This has the look of a classic early/mid summer heat wave; it's not as typical of late July (and would be very unusual in August).  It's certainly interesting to see this unfold after a very "westerly" first half of July.  Here's how this year compares to previous years in terms of the Seward-Deadhorse westerly wind speed index that I described a few weeks ago (when the previous pattern change occurred).

By this metric, July 1-15 this year fell in the top 10 for westerly winds aloft.  Not surprisingly, very strong westerly flow at 700mb (approximately 10,000' above sea level) corresponds to suppressed daytime temperatures in Fairbanks, owing to increased cloud and rainfall, as well as the predominance of cool North Pacific air masses.

Is it unusual for such a strong westerly regime to be followed by a heat wave later in summer?  No: the historical data shows no relationship between early July westerlies and the highest subsequent weekly temperature (looking at daily high temps):

And if we look at the highest daily temperature in the rest of summer, it's interesting to see that a strong westerly pattern in early July tends to have at least one subsequent episode with temperatures into the 80s in Fairbanks - unlike some years.

It's possible that the high-amplitude pattern that produces a strong westerly regime in early July is also conducive to subsequent hot weather, at least for a day or two.

In other words, don't write off summer in early July.

Finally, for completeness, here's the July 16 - August 31 total precipitation in Fairbanks, in relation to early July westerly flow.  Strong westerly patterns tend not to be very dry for the rest of summer, but neither are they typically very wet.

Weather Notes

Large parts of Alaska have been much wetter than normal recently; here's a map of estimated 14-day precipitation totals (NOAA data) and the "percent of normal" (click to enlarge):

Several longstanding climate observing sites saw near-record rainfall over the weekend, including:

Bethel 1.98" on Friday-Sunday: the highest 3-day rainfall total in the month of July since 1955.

Anchorage  2.28" on Saturday-Sunday: the second highest 2-day total in the month of July; only 2001 had more (July 4-5).

Sitka  2.35" on Saturday, the highest calendar day total in July since 1948.

Of course all these numbers would be less notable in August, when heavy rain storms are less rare.

Wet and unsettled weather is common across southern Alaska in summer when the PDO (Pacific Decadal Oscillation) is strongly negative, as it is now.  Here's a map showing the percent of years with above-normal sea-level pressure in summers with a bottom-10 PDO index since 1950:

And the PDO is certainly entrenched in the negative phase: here's the daily index for the last several years.

Sea surface temperatures are much lower than normal in the eastern Bering Sea, which is both cause and effect of the atmospheric trough (low pressure) that is tending to recur in that area.

The weekend rainstorm was produced by a strong trough over that very location; here's the 500mb chart from 4pm AKDT on Saturday afternoon (click to enlarge):

The deep, strong southerly flow into south-central Alaska is amply illustrated on the 4pm balloon sounding from Anchorage: observe all the wind barbs pointed out of the south on the right side of the profile.

Also notice the very high freezing level - over 11,500 feet above sea level - and the high water content (precipitable water) of well over an inch (PW=3.02cm).  No wonder it rained a lot.

The combination of cloud, rain, and a chilly ocean nearby also produced unusually low daytime temperatures: Anchorage and Bethel haven't seen back-to-back days with such low daily maximum temperatures in July since 1970 (54-55°F in Anchorage) and 1971 (48-49°F in Bethel).  Guess what: the PDO was significantly negative back then too.

Regional Lightning Data

For my latest data project, I have been looking at the distribution of lightning within Alaska's 21 fire management regions, i.e. the Predictive Service Areas:

Using the 2012-present lightning data from AICC, I've assigned each lightning strike to one of the zones, or to an adjacent region in Canada or over the ocean - I'm still working on the ocean zones.  This will be useful for looking at historical lightning variations by region, and for assessing realtime activity compared to the (modern) historical average.

The greatest density of lightning occurs in the three zones to the north of the Alaska Range, what I'll call the Tanana valley zones (AK01E, AK01W, AK03S).  (There's also a Tanana Zone-North PSA, but that's considerably less active.)  Here's the annual number of ground lightning strikes in these 3 zones, both for the full year and through July 12:

These zones comprise just over 10% of the area of Alaska, but they receive about a third of the lightning strikes in the state.  This year's activity is below normal so far: the 4th lowest in 13 years of data.

Interestingly the last two years were polar opposites in terms of the timing of lightning: in 2022 it was nearly all prior to this date, but last year most of the lightning activity was after July 12.

The region of Alaska with the largest lightning deficit so far this year is the northeast interior, i.e. the large Upper Yukon Valley zone.  This zone is very nearly as large as the 3 Tanana valley zones combined, and it gets the most lightning of any single zone in terms of total number of strikes.  Here's the chart of annual numbers:

As of July 12, the total activity in this zone is above last year, but that will change imminently.  It's interesting to see how active the first half of summer was in 2012-2017 in this zone, whereas since then only 2022 was very active before mid-July.

Now that I have this data set at my disposal, blog readers should brace for a few more posts analyzing the numbers.

June Climate Data

June climate data are in from NOAA/NCEI, so it's time for a recap.  As we all know, it was a very warm and dry June for the state as a whole: the sixth warmest and fifth driest since 1925, according to the NCEI data.  The only years that were both warmer and drier in June were 1936 and 1957, although June 2022 was very similar in temperature and was the driest on record in the NCEI dataset.

The warmth was most anomalous for the northeastern interior, and the ERA5 climate model (reanalysis) data backs this up:

Only June 2004 was warmer in the northeastern interior, and that was also the warmest June statewide, leading into the worst fire season in modern history.  This year's 300,000 or so acres that burned in June pales in comparison with the nearly 1 million acres that burned in June 2004 (followed by another 5 million in July and August).

ERA5 data confirms that unusual dryness was widespread from the Panhandle to the West Coast and across most, though not all, of the interior.

The circulation pattern that produced this outcome involved high pressure over the East Siberian Sea and low pressure in the Gulf of Alaska:

In winter this would be a cold pattern as cold air sinks into Alaska from the high Arctic, but in the height of summer a warm outcome is favored by ample sunshine heating up the continental interior (see below).  It's also worth noting that the exceptional warmth in the last third of June was driven by a trough in the Bering Sea and a big ridge over the interior, i.e. a more amplified pattern than the monthly mean suggests.

Wind was above normal in the Aleutians and western/southwestern Alaska, but mostly lighter than normal elsewhere.

But June is firmly in the rearview mirror now, as the pattern has been completely different in the last week.  McGrath tied a daily record low temperature yesterday, and CPC suggests the generally cool weather will continue for at least the next couple of weeks.  There have only been 4 years since 1925 when July was cooler than June on a statewide basis - will it happen again this year?  It's seems quite possible.

Big Pattern Change

What a change a few days can make.  A dramatic onset of westerly flow has brought much cooler weather to western, interior, and northern Alaska, and substantial rains have provided major relief for the wildfire situation in many areas.  All burn suspensions have been lifted statewide.

Yesterday was really quite chilly in the interior: with cloud and rain, the high temperature was only 60°F in Fairbanks and 55°F in Bettles.  These are distinctly unusual conditions for this early in July, but would be less so in the second half of July.  Rain totals in the past couple of days exceeded 3" in some spots in the hills around Fairbanks:

Earlier in the week, even higher rain amounts fell in northwestern Alaska: as much as 4-5" was measured in the hills to the north of the Kobuk River delta:

Well over 3" fell in 12 hours at the ASIK RAWS site, elevation 1329': it was an onslaught of heavy, wind-driven rain on Tuesday afternoon.

The immediate explanation for this pattern change is the development of a strong trough on Alaska's side of the Arctic, bringing strong westerly flow across central and northern Alaska.  Furthermore, a powerful ridge has developed over the Aleutians, boosting the north-south pressure gradient that's driving the westerly flow.  Here's the 500mb chart from 3am today, courtesy of Environment Canada:

The estimated 500mb height of 597dm just south of the Aleutians is extraordinarily high and may be an all-time record for the location; I'll need to look into this.

It's interesting that a westerly episode of this magnitude has developed so early in July, when we typically think of late July and August as the time for onset of rainy, cool weather across interior Alaska.  The chart below shows estimates of the normal seasonal cycle of 700mb westerly flow across the 150°W meridian between 60 and 70°N, i.e. a north-south line from near Seward in the south to roughly Deadhorse in the north.  Two 30-year periods are shown: 1961-1990 and 1991-2020.

This confirms that westerly flow tends to see a seasonal minimum around late May and ramps up quickly to its seasonal maximum in August.  There seems to be a hint that recent decades have seen weaker August westerlies, but this difference may well be within the random error from a 30-year sample size.

In a subsequent post I'll take a look at the historical variability of "westerly onset" timing.  

Fire Update

A quick update on fire: the weather is now changing significantly, but the hot conditions of the last several days have pushed statewide fire acreage up to 371,000 acres, per the AICC's daily report.  That total is now above the historical median for the date.

Unfortunately for Fairbanks, almost a third of the acreage has accrued in the nearby McDonald fire, and air quality was atrocious in town yesterday, with the visibility reported as low as 1/4 mile at the airport.  The last time the visibility was that low from smoke (not fog) was way back in 2005.

The following chart illustrates the relationship between the 7-day increase in statewide fire acreage and the weekly-mean high temperature in Fairbanks.  For reference, the latest 7-day mean high temperature in Fairbanks was 81.6°F.

Of course there's much more to fire growth than just high temperature - there have been large acreage increases with quite chilly temperatures in the past - but the warmest weeks do tend to produce at least moderately large acreage gains, as we would expect.