Thursday, March 19, 2015

Diurnal Cycles

A few days ago reader Gary inquired about the time of daily minimum temperatures in Fairbanks throughout the winter.  The simplest way of addressing this is to examine the climatological mean temperature by hour of the day; the chart below shows the results for the modern ASOS era, 1998-2014.  Each curve shows the mean diurnal cycle for a different month, and the small diamond markers show the maximum and minimum points.

The seasonal changes in the diurnal cycle are very much as expected, with a tiny diurnal cycle in December giving way to a substantial daily oscillation by February.  According to the hourly means, the coldest hour of the day in December is 7 am AKST, which is well before the sun rises; but there is less than 0.5 °F difference from midnight to 11 am.

We could also look at the frequency with which each hour of the day records the lowest or highest temperature, and examine some histograms based on those results, but I'll leave that for another time.

As we're on the topic of short-term temperature fluctuations, it's worth noting that these have been very large in Fairbanks lately.  The plot below shows the 2m temperature trace from UAF's North Campus, near Smith Lake, over the past 4 days (with a period of missing data in the middle).  The 2m temperature rose from -39.4 °F on Sunday to +47.2 °F yesterday, a rise of 86.6 °F in three days.  Looking at the official Fairbanks history since 1930, only two occasions saw a larger 3-day temperature rise, and both of those were in January.

At Fairbanks airport, the recent 3-day rise was 84 °F (-37 to +47), the third largest on record and the largest ever outside of December and January.

The diurnal temperature range has also been very large in recent days: 46 °F yesterday, which is equivalent to the largest that is typically observed in any given year.  March is by far the most common month for observing the year's largest diurnal temperature variation in Fairbanks.

Monday, March 16, 2015

Coldest Airmass of the Winter

The 850 mb temperature observed by balloon sounding in Fairbanks dropped to -29.9 C on Saturday morning, which is the coldest of the entire winter.  Prior to the arrival of this cold airmass, the coldest observed at 850 mb was -27.7 C on January 25.  If the previous low point had stood, the winter's coldest measurement would have been the second warmest on record for an entire winter, but this new cold is closer to (but still above) the long-term normal for coldest of the season.

Is it unusual to see the coldest air (aloft) of the season in March?  Not especially; the chart below shows the month in which the coldest air was observed at 850 mb for each winter since 1948-1949.  It's nearly as common to see the coldest air in March as in December.  The last time the coldest air was observed in March was in 2003, which followed the very warm winter of 2002-2003; obviously the odds of this happening are heightened after persistent warmth earlier in the season.  Over the history since 1948, the earliest that the coldest air was observed was November 16, 1956, and the latest was April 18, 1959.

At the surface, Fairbanks airport observed 6 consecutive days with a daily minimum of -30 F or lower, culminating in -37 yesterday, March 15.  This is fairly impressive for so late in the season; it's the most consecutive days reaching -30 this late in the winter since 1966.  However, it can't compare to the 13 straight days reaching -30 F from March 9-21, 1959.

The -37 F yesterday morning is the coldest this late in the winter since 1995 (-37 F on March 24).  The record for coldest this late or later is -46 F on March 15, 1964.

Saturday, March 14, 2015

Frequency of Extremes

I've spent some time recently working with the NCEP/NCAR reanalysis data to assess patterns of upper-level temperature and pressure extremes over the Northern Hemisphere, and the changing distribution of these extremes over time.  I won't dive into the details of this investigation now, but I did want to put up one figure that I thought was interesting - see below.

The map shows the long-term average of the frequency of +/- 2 standard deviation anomalies in the daily mean 850mb temperature.  The anomalies are calculated with respect to a baseline climatology that includes a long-term trend, so that patterns of extremes do not simply reflect overall warming.

The interesting result is that southern Alaska, western Canada, and the northern Gulf of Alaska see the highest frequency of 2 SD extremes of anywhere in the northern extratropics.  If the distribution of daily anomalies were perfectly Gaussian, then the frequency would be 0.0455, so the 850mb climate over southern Alaska has more frequent extreme temperature events than would be expected.  Most of the northern extratropics have slightly less frequent extremes than expected.

Looking at the distribution by season, it is mainly the winter and spring that contribute to the excess of extremes in the Gulf of Alaska.  The map below shows the mean frequency for December through February.

In contrast to the peak frequency near Kodiak Island, the lowest annual frequency of 2 SD extremes is over the North Atlantic, and the two distributions can be compared as follows:

Category (SD)57.5°N 155°W50°N 35°WExpected Gaussian
-3SD or lower0.00380.00010.0013
-3SD to -2SD0.02260.01350.0214
-2SD to -1SD0.11410.16080.1359
-1SD to +1SD0.72330.64140.6827
+1SD to +2SD0.10690.17010.1359
+2SD to +3SD0.02560.01420.0214
3SD or higher0.00370.00000.0013

The chart below shows how the frequencies in three categories have changed over time near the peak in the Gulf of Alaska.  Relatively high frequencies - higher than expected based on a Gaussian distribution - were observed prior to the PDO shift in 1976, but the frequency of extremes has been generally lower since then.

Wednesday, March 11, 2015

Inversion Season Dates

A late-season cold spell has descended upon Alaska, bringing -30 °F temperatures to Fairbanks airport on the past two mornings.  Temperatures this low are a bit unusual for this late in the season, occurring in fewer than 3 out of 10 years after March 10.  Rapidly strengthening sunshine is making a big difference, however, as daytime temperatures would be much lower with an air mass this cold in the depths of winter.

With solar insolation rising quickly, Fairbanks will soon reach the date at which mean daily surface temperatures rise above mean daily temperatures at 850 mb.  According to the 1981-2010 normals, that date is March 19.  Based on the history of radiosonde data from several stations around the state, we can look at how the surface-850 mb temperature difference varies throughout the year; the charts below show the results.  Note that I would prefer to look at the 925 mb level as a measure of low-level inversion characteristics, but the 925 mb was not regularly reported in radiosonde data until 1992.

It's not surprising to observe that more northerly locations in Alaska generally spend a greater fraction of the year with surface temperatures lower than 850 mb temperatures, i.e. with an inversion in place on average; Anchorage sees this condition only during a mid-winter period of about 6 weeks in length.  An interesting feature of the charts is that the surface-850 mb temperature difference is nearly constant during most of the warm season at each of the locations outside the Arctic, so the low-level vertical stability of the atmosphere is little changed for a period of several months during summer.

It's also interesting to note that the Fairbanks and McGrath curves follow each other very closely indeed, despite a fairly large distance separating the two locations; this illustrates the relatively uniform nature of some aspects of the climate in interior Alaska, as opposed to the strong spatial variation near and along the coasts.

Saturday, March 7, 2015

Arctic Ice Update

Arctic sea ice is now close to its seasonal maximum extent, so it's a good time for an update.  The total area of ice cover has seen little growth since early February, and the extent is now running more than 2 standard deviations below the 1981-2010 normal according to the NSIDC.  The warm winter and lack of ice cover in the Bering Sea is contributing to the ice area shortfall, along with a large ice deficit in the Sea of Okhotsk and a smaller deficit in the Barents Sea.  The latest graphics from the NSIDC are shown below.

NSIDC commentary notes that if the ice extent does not advance to a new peak in the next few weeks, then this year will see the lowest maximum extent on record for the satellite era.  At first glance this would seem to bode ill for the upcoming melt season.  However, the University of Washington's ice volume analysis - last updated for the month of January - showed that ice thickness and ice volume were higher than in recent years.  First, the thickness: the PIOMAS model's estimate of average thickness at the end of January was the highest since 2006 - see below.

The estimated Arctic sea ice volume was close to that in 2009 and just over 1 standard deviation below the 1979-2014 mean (see below).  The volume was notably higher than last year in January, which reflects a persistence of the year-over-year gain that was seen during last year's melt season.  It will be most interesting to see if the volume recovery continues this year and if the ice extent and ice volume anomalies continue to show contrasting trends.

According to NOAA's CFS reanalysis, the mean air temperature has been well above normal across most of the Arctic Ocean this winter, with the highest anomalies located near Wrangel Island and the Chukchi Sea.  The extent and magnitude of the warm anomalies can be compared to the previous 10 years in the series of images below.  The previous 10 winters showed a persistent tendency for warmth over the Barents and Kara Seas and with less unusual warmth on Alaska's side of the ocean.  Last winter (2013-2014) saw warm conditions across the entire ocean, but this winter has been notably cooler between 0 and 90 °E, with the warmest conditions now prevailing on our side of the basin.

Monday, March 2, 2015

Kaktovik Wind

The severe blizzard that affected the eastern North Slope on Saturday night was notable for its ferocity in the Barter Island community of Kaktovik, as documented in a news article here.  The highest sustained wind measured at the Barter Island ASOS was 49 knots (56 mph), and the highest gust was 58 knots (67 mph), but the anemometer did not report for several hours during the height of the storm.  It seems very likely that winds were significantly stronger for a time.

How does this storm compare to past high wind events on Barter Island?  The history of hourly observations contains quite a number of bogus wind reports in recent years, but after removing these I found that there have been 18 separate dates (in 13 separate winters) since 1950 on which sustained wind speeds reached 60 knots (69 mph).  All but one of these occurred in November through March.  Without question, Barter Island is a very windy place in winter, and near-hurricane force winds are not all that uncommon.  The top three wind events since 1950 are as follows:

January 6, 1974    70 kt sustained
December 27, 1951    69 kt sustained
September 20, 1957    68 kt sustained

It's interesting to note that most of the 60+ knot events were in the 1950's and 60's, and there do not appear to be any reliable reports of 60 kt or higher since 1995.  This could be due to different wind speed measurement parameters with the introduction of ASOS - but it is surprising nonetheless.

The highest winds at Barter Island are nearly always from a westerly direction, as in this weekend's storm.  The climatological distribution of wind speed also shows a pronounced maximum for east-southeasterly flow (see below), but easterly flow does not bring the same extremes in wind speed.  Winds from the north and south are typically light and are also much less frequent.