Fairbanks Records

Reader Eric asked recently about the rate of breaking weather records in Fairbanks, and specifically whether the pace of record-breaking has picked up in the past several years.  The official weather records in Fairbanks date back to observations taken in the early part of the 20th century, but with various site changes involved in the earlier years I decided to look only at the GHCN data for Fairbanks airport since 1949.  The series of charts below shows the annual number of records, either broken or tied, relative to the 1949-2013 history, for several temperature and precipitation variables.

First, the daily maximum and minimum temperature records: 2013 saw a large number of daily records for both warmth and cold; the total for the year was 32, only slightly behind 1969 and 1992 (33 records each).  [Note that here I am looking only at high maximum and low minimum temperatures, not low maxima and high minima.]  The long-term temperature trend is evident, with more warm records broken in the last two decades, and far fewer cold records broken since the Pacific climate shift of the mid-1970s.  One other interesting feature: only one year (1972) saw no high temperature records, while 8 years saw no low temperature records.  So it seems that cold temperature records are slightly more clustered in time.

Next, daily precipitation and snowfall records.  There is no obvious trend in the precipitation records, but it seems that fewer snowfall records have been broken since 1976.  Note that 5 daily precipitation records have been set so far in 2014, with today's event not quite adding to this count.

Finally, I also looked at the records for monthly mean temperature and monthly total precipitation and snowfall; see the charts below.  Again the temperature records are the most interesting, with a cluster of record warm months from 1975-1981 and none prior to 1975.  Two monthly temperature records were broken in 2013 (cold April, warm October); 1979 was the only other year with both a record warm month and a record cold month.  For snowfall, the period 1991-1995 saw record snowfall for 5 of the 10 months, although the 1995 record was for the only August snowfall in the history (0.1 inches).

In conclusion, 2013 was certainly an extreme year for breaking temperature records, and was a little more extreme than "usual" for precipitation and snowfall; but other years of the past decade do not stand out as especially unusual.

Update July 9: I've added charts below for the daily low maximum records and high minimum records, as well as the total of all four daily temperature record types.  1992 really stands out here for the low maxima, and 2013 tied with 1981 for most high minimum records.  In all four categories combined, 2013 was second only to 1992 in terms of the number of temperature records.  However, 2013 was very unremarkable in terms of the number of low maximum records.

Heavy Rain Again

Heavy rainfall is occurring again in Fairbanks this morning - the fourth time in less than three weeks, and this time accompanied by thunder in the early morning hours.  As of 8:11am AKDT, 0.84 inches of rain had fallen.

At Fairbanks airport between 1998 (when the change to ASOS occurred) and 2013, an observation of "heavy rain" was reported on only 27 days.  Seven of these 16 years did not see any reports of heavy rain, while the record in this period was 8 heavy rain days in 2008.  We have now seen 4 events in the past few weeks (June 19, June 26, June 30, and now today).

The culprit for today is a near-stationary east-west band of rainfall with embedded convection.  Let's hope it doesn't remain in place too long, or it will cause renewed flooding concerns.

Wet But Not Thundery

With the June rainfall record having been broken this year in Fairbanks, and with July already in the top 10 wettest on record, it is interesting to consider that thunder has been reported only twice this year at Fairbanks airport (June 8 and 23); and these two events contributed zero precipitation to the June total.  This led me to wonder if very wet summer months are typically lacking in thunder in Fairbanks.  If so, this suggests a very non-linear relationship of thunder to total precipitation, because very dry weather would also tend to be associated with a lack of thunder.

To answer the question, below are scatter plots of the total precipitation versus number of days reporting thunder, for June and July.  Indeed we see that in both months, the highest rainfall totals are not associated with a high frequency of thunderstorms.  Instead, the highest thunderstorm frequency tends to be associated with moderately, but not excessively, above-normal precipitation.  In June it is possible to have quite high thunderstorm frequency with well below-normal precipitation, but it seems that in July the very driest months tend to have reduced thunderstorm activity.

In conclusion, the wettest summer extremes in Fairbanks are not associated with much-enhanced thunderstorm activity, but in fact quite the reverse: thunderstorms are less common in the weather patterns that bring the wettest conditions.  This is evident also in the history of heaviest daily rainfall amounts: of the top 20 one-day rainfall events in June and July (for years with hourly weather reports available), only 5 days reported thunder.

Updated Rainfall Totals / Map

Here is quasi-final map of rainfall totals in the Fairbanks portion of interior Alaska. Due to the large number of stations, it was impracticable to ass labels to all of the dots. Thanks to Rick for putting the numbers together and for passing them along. Please remember that these are unofficial tallies.

Figure 1. Map of rainfall totals unofficially tallied by the NWS.

Figure 2. List of rainfall totals unofficially tallied by the NWS.

Preliminary Rainfall Totals

The NWS Fairbanks Office released a Public Information Statement at 1:30 p.m. today with preliminary rainfall totals for the storm. Figure 1 shows those values. In addition, the number of consecutive hours with measurable precipitation in Fairbanks ended at 1 p.m. AKDT at 37. This breaks the previous record of 33 set twice (data only goes back to 1974). Figure 2 shows the hourly precipitation during that time period (Note: the max 24-hour value is slightly different than what the NWS stated).

Figure 1. Storm totals around Fairbanks based on the NWS Public Information Statement.

Figure 2. Hourly precipitation for Fairbanks according to publicly available ASOS.

...STORM TOTAL RAINFALL AMOUNTS FOR CENTRAL AND EASTERN INTERIOR AND 
THE ALASKA RANGE AS OF 11 AM AKDT...

...DELTANA AND TANANA FLATS...

OKLAHOMA RAWS.........................................1.68 INCHES   
DONNELLY RAWS.........................................1.33 INCHES   
RHOADS CREEK..........................................1.20 INCHES  
BLAIR LAKES...........................................1.10 INCHES   
JARVIS CREEK..........................................1.10 INCHES  

...DENALI...

DENALI VISITOR CENTER.................................1.54 INCHES   
MARGUERITE CREEK......................................1.39 INCHES   
EIELSON VISITOR CENTER................................1.13 INCHES   

...EASTERN ALASKA RANGE...

GOLD KING.............................................3.56 INCHES
BOLIO.................................................1.05 INCHES

...MIDDLE TANANA VALLEY...

ESTER.................................................4.24 INCHES
MANCHU................................................3.94 INCHES
LITTLE CHENA RIVER (11 MILE CHSR).....................3.84 INCHES
TWO RIVERS............................................3.67 INCHES
FAIRBANKS 11NE CRN....................................3.64 INCHES   
UNIVERSITY WEST.......................................3.58 INCHES
BIRCH HILL............................................3.45 INCHES
FORT WAINWRIGHT.......................................3.35 INCHES
FAIRBANKS INTERNATIONAL AIRPORT.......................3.31 INCHES
COLLEGE OBSERVITORY (UAF).............................3.11 INCHES
SMALL ARMS RANGE (FORT WAINWRIGHT)....................3.10 INCHES
CHENA RIVER AT GRANITE TORS CAMPGROUND................3.09 INCHES
NORTH BIAS DRIVE FAIRBANKS............................2.93 INCHES
SALCHA RAWS...........................................2.76 INCHES
LITTLE CHENA RIDGE....................................2.66 INCHES
CARIBOU PEAK..........................................2.49 INCHES
BONANZA CREEK.........................................2.49 INCHES
WICKERSHAM DOME.......................................2.45 INCHES
STUART CREEK..........................................2.31 INCHES
ANGEL CREEK...........................................2.14 INCHES
FISH CREEK............................................2.03 INCHES
UPPER NOME CREEK......................................1.91 INCHES
SALCHA RIVER AT JOHNSONS..............................1.78 INCHES

...UPPER TANANA VALLEY AND THE FORTYMILE COUNTRY...

GOODPASTER RIVER......................................1.32 INCHES

...YUKON FLATS AND SURROUNDING UPLANDS...

NOME CREEK (NR BEAVER CREEK)..........................1.54 INCHES
GOLDSTREAM CREEK NR LIVENGOOD.........................1.45 INCHES
EAGLE SUMMIT..........................................1.40 INCHES
LOST CREEK............................................1.17 INCHES

Interior Deluge

[Update 5 pm]  The 2-day total at Fairbanks airport stands at 3.37".  Significant rises are being observed and predicted for area rivers; by this time tomorrow the Tanana River at Nenana is expected to be quite close to a moderate flood stage:

[Update 10 am AKDT: with the event apparently winding down now, the 2-day total is up to 3.31", making this (so far) the 3rd heaviest 2-day event on record in Fairbanks.  According to the NOAA precipitation atlas, the average recurrence interval for this is about 50 years.  Brian supplied more statistics in the comments below.]

The third major rainstorm in a two-week period for the Alaskan interior is proving to be the heaviest, by far, for the Fairbanks area.  Rainfall since midnight Monday night has amounted to 2.67" at Fairbanks airport, and it's still falling heavily.  This is an extraordinarily high amount for interior Alaska; here are the top 5 non-overlapping 2-day precipitation totals for Fairbanks:

Aug 11-12, 1967   4.29"
Jul 26-27, 2003    3.43"
Jul 19-20, 1948   2.81"
Jul 1-2, 2014    2.67"+
Aug 27-28, 1930    2.40"

There will be much more to say when all is said and done, and I expect Rick and Brian will contribute other precipitation totals, but for now here is a map of 24-hour totals from the NWS mesonet page:

Summer Freezing Level

A couple of weeks ago, reader Mike asked about the frequency of snow in June on the mountains around Anchorage and in the interior, and I suggested that a partial answer might be available from the observed freezing level in the balloon soundings.  To look at this, I calculated the freezing level for the Fairbanks balloon observations back to 1948, and the following chart shows the minimum and maximum freezing level observed on each day of the year.  There are some interesting features here, as the freezing level maxima are generally lowest in March and highest in August, which represents a considerable lag relative to the surface temperatures.  Also, the range of extremes widens dramatically in August compared to July; August can bring very deep warm air with very high freezing levels, but can also usher in much cooler air with rather low freezing levels.

Zooming in on the summer months (see below), we see that it is very rare to get a freezing level below 2000 feet in June, and the record minimum for July is close to 3000 feet.  However, by August 10 a freezing level as low as 1400 feet has been observed.

Looking at only days with measurable precipitation at Fairbanks airport (see below), we see that a good many of the cold extremes and the vast majority of the warm extremes occurred on days without precipitation - although of course there may have been precipitation in the mountains on these days.

It may be a bit perilous to draw conclusions about snowfall from this temperature data alone, but presumably we can state when accumulating snow is very unlikely to occur.  It seems that accumulating snow would be extremely unusual at any time in June for elevations of 2000 feet or below near Fairbanks, and it would be unheard of in July; although of course a heavy shower can sometimes bring down cooler air and wet snow for a time, so perhaps we shouldn't rule out this kind of wet snow shower in the hills in June.  The highest elevations of the White Mountains near Fairbanks are somewhat over 3000 feet, so it seems these summits may be prone to seeing occasional snow events in June and August (probably more often in late August), but apparently not in July.  And it appears that for peaks above 4000 feet, such as in the higher terrain of the Yukon-Tanana Uplands, snow can't be ruled out at any time of year.

Rare Severe Thunderstorm Warning

On Monday, June 23, 2014, the Fairbanks National Weather Service Office issued a Severe Thunderstorm Warning for the southern Yukon Flats (AKZ 220).

WUAK59 PAFG 240411
SVRAFG
AKZ220-240430-
/O.NEW.PAFG.SV.W.0001.140624T0411Z-140624T0430Z/

BULLETIN - IMMEDIATE BROADCAST REQUESTED
SEVERE THUNDERSTORM WARNING
NATIONAL WEATHER SERVICE FAIRBANKS AK
807 PM AKDT MON JUN 23 2014

THE NATIONAL WEATHER SERVICE IN FAIRBANKS HAS ISSUED A

* SEVERE THUNDERSTORM WARNING FOR...
  SOUTH CENTRAL YUKON FLATS AND SURROUNDING UPLANDS IN ALASKA 50 
  MILES NORTHWEST OF TWELVEMILE SUMMIT...

* UNTIL 830 PM AKDT

* AT 807 PM AKDT...DOPPLER RADAR INDICATED A SEVERE THUNDERSTORM
  CAPABLE OF PRODUCING QUARTER SIZE HAIL. THIS STORM WAS LOCATED 50
  MILES NORTHWEST OF TWELVEMILE SUMMIT...AND MOVING NORTH AT 5 MPH.

The last time a Severe Thunderstorm Warning was issued anywhere in Alaska was on July 11, 2010 (AKZ 222 near Twelvemile Summit). There are people far more qualified than I am to discuss the atmospherics of this event; however, I did want to show a few maps of the region and some of the radar products. On each map, the trapezoid represents the coordinates of the warned area. Some of the interesting features include the 3.27"/hour maximum precipitation rate and the 90% probability of hail statistic (a text product).

Figure 1. Location map of Severe Thunderstorm Warning.

Figure 2. Base Reflectivity map of Severe Thunderstorm Warning.

Figure 3. Cloud to height map of Severe Thunderstorm Warning.

 

Figure 4. Storm Relative Velocity map of Severe Thunderstorm Warning.

 Figure 5. Instantaneous Precipitation Rate map of Severe Thunderstorm Warning.

Figure 6. Vertical Integrated Liquid map of Severe Thunderstorm Warning.

Wet Rest of Summer?

Fairbanks has now tied the all-time record for June total precipitation (3.55", tied with 1949), so it's natural to ask whether this has any statistical implications for the rest of the summer.  In other words, based on the historical data, does a wet June portend a washout for the rest of the summer?  This is a straightforward analysis to perform: see the scatterplots below for June precipitation versus precipitation in July and in August; the horizontal gray lines show the median value for July and August (1.71" in both months, as it turns out).  There is a hint of a positive correlation for July, as the very driest Junes have generally been followed by dry Julys, and a wet July is slightly favored after a wet June.  The two wettest Junes were in 1949 and 1955, and both saw a wetter than normal July.  August rainfall, on the other hand, shows no discernible connection to June rainfall.

Looking at mean temperatures in July and August (see below), there seems to be a suggestion - more for August than for July - that June precipitation is slightly positively correlated with temperatures later in the season.  1949 and 1955 are obvious exceptions, but these were negative PDO summers, unlike this year.  Here are the June PDO index values for the top 8 wettest Junes in Fairbanks:

1949  -0.70

1955  -2.44

1977  +0.42

1970  +0.06

1989  +0.36

1994  +0.46

1988  +0.74

1962  -1.62

Given this information, we probably shouldn't be too surprised if July is also damp in Fairbanks, but it doesn't seem particularly likely that the recent unusual coolness will continue.  With warmer than normal water in most of the North Pacific this summer, unusual warmth appears to be a more likely outcome for the late summer.

With the PDO phase apparently not closely related to June precipitation in Fairbanks, what else can we identify as a possible cause?  I would suggest (as I did last summer) that the Quasi-Biennial Oscillation bears some responsibility for climate variations at this time of year.  The chart below shows the June QBO index value plotted against June precipitation, and suggests that precipitation is enhanced when the QBO is strongly negative.  Monthly QBO index values from the past 3 years are shown in the last chart below; we can see that the QBO was strongly positive last summer, and continued positive through the winter, but finally turned negative for the May average and has probably been significantly negative this month.  While this by no means guarantees a wet June, it does seem to raise the odds; and I would also suggest that if the QBO becomes strongly negative by August, then it will favor a relatively warm month then.

Annual Normal Temperature Visualization

If you haven't figured out by now, I enjoy tinkering with data and seeing how it looks on a map. As the saying goes, when you have a hammer (GIS), you look for nails to hit (climate data). In this case I wanted to see how the NCDC climate normals look when put into motion. In the past I made date specific maps of climate normals of monthly normals but not for every day of the year. The YouTube video below shows a side-by-side pannel of all 365 daily normals (high temperature on left and low temperature on right). The colors (categories) are based on the 189 stations in Alaska that have published normals (some using as little as 10 years worth of data). An inverse distance weighted (IDW) surfacing algorithm was used to interpolate values between points.




For a clearer animation, HERE is the standalone high temperature animation and HERE is the standalone low temperature animation.

If you are interested in seeing what the entire U.S. looks like, here is the companion video that focuses on the Lower 48.

Precipitation Update

Earlier in the month we noted that Fairbanks precipitation had amounted to only about half of normal for the year to date, but the recent heavy rainfall has now completely erased this deficit; so 2014 is back to near-normal based on the mean climatology.  The median precipitation by this date is a little less - just under 3 inches.

Heavy rain has also occurred in Barrow: 0.49 inches yesterday, which is the second highest daily amount ever reported there in the month of June.  This continues a pattern of above-normal precipitation in Barrow that has persisted since March of last year; in the nearly 18 months since 2013 began, Barrow has seen nearly twice its normal precipitation.  The wettest periods relative to normal were March through September of last year, November and December of last year, and May 2014 through the present.  I've plotted the sea level pressure anomaly in each of these periods below; note the strong similarity in terms of the above-normal pressure in the Bering Sea.  It appears that this recurring anomaly has helped create a favored path for weather disturbances across the North Slope.

Update: the map below shows the correlation of annual mean sea surface temperature with annual mean sea level pressure at St Paul Island, which is located close to the center of the high pressure zone in the three maps above.  This is a negative PDO pattern, and indeed the annual mean PDO index values are correlated at -0.47 with the annual mean pressure at St Paul Island.  A negative PDO phase favors high pressure in the Bering Sea, although the correlation is only found in November through April.

The PDO was negative throughout 2013, so this may explain the high pressure in the Bering Sea last year, but it doesn't explain the recent persistence of the pattern, because (a) the PDO is now positive, and (b) the correlation goes away in summer.  So I don't have an explanation for the remarkable persistence of the pressure anomaly; and I would suggest that the profound seasonal changes in climate dynamics make it unlikely that a single cause can be identified.

Lightning Frequency

I'll beat the lightning theme to death one last time today. In the last few weeks, I have added lightning-themed posts, here, here, and here. In today's post, I took the 1,884,764 lightning strikes in the BLM lightning archive (1986-2010) and computed a density of strikes. To do this, I created a raster grid of Alaska where each cell is 10 square miles. The GIS software then counted the number of lightning strikes that occurred in each cell and divided by 27 (years) and then by 10 (square miles per grid cell). The resultant map shows the frequency of cloud-to-ground lightning strikes. Figure 1 shows the raw computation of lightning frequency. For visual purposes, I ran a 3x3 mean filter across the data to remove a lot of the noisiness in the map. Figure 2 shows a smoother, more aesthetically pleasing representation of lightning frequency.

The lightning data is fairly similar to the thunderstorm climatology map that NOAA produced in their 1976 report on thunderstorm climatology in Alaska. Their report showed the greatest density in the Fortymile Country whereas the BLM's lightning data indicates the area immediately east of Fairbanks has the highest density of lightning strikes.

Figure 1. Raw number of lightning strikes per square mile in Alaska (1986-2012). Grid cells are 10 square miles.

Figure 2. Smoothed number of lightning strikes per square mile in Alaska (1986-2012). Grid cells are 10 square miles and the data were smoothed using a 3x3 mean filter.

Figure 3. Figure 7 from 1976 NOAA report on thunderstorm climatology in Alaska.

A Little Cool in Barrow

It's not often there is a cool anomaly to point out in Barrow, but the past month has actually been a little cooler than the 1981-2010 normal at Alaska's northernmost extremity; the chart below shows the daily mean temperature compared to normal since April 1.  This cool anomaly comes immediately after Barrow had its warmest year on record for the period June 1 - May 31.

Barrow has yet to break the 40 °F barrier this year, which is rather unusual by the time the summer solstice arrives; the last time 40 °F was first exceeded this late was in 1974.  The culprit appears to be an absence of southerly flow, due to higher than normal pressure over the ocean to the northwest, and lower than normal pressure over land to the southeast (see map below).

The third figure below shows the annual sum of thawing degree days up to June 21, and the total so far this year is a little lower than in most recent years (though by no means exceptional).  It will be interesting to see if unusual warmth soon returns or if - as the CFSv2 ice forecast seems to think - the summer might remain relatively cool.

Evaporation Calculations

As a follow-up to my recent post about potential evaporation rates in Fairbanks last summer, I decided to go one step further and include the effect of wind speed and solar radiation.  Higher wind speeds and higher solar radiation both increase the rate of evaporation from land, water, and vegetative surfaces, and so it's of interest to see whether these factors also contributed to high moisture loss last summer.  Hourly radiation, wind, temperature, and humidity data are recorded (among other variables) at the Fairbanks Climate Reference Network (CRN) site, which is located at 1140' elevation about 10 miles northeast of Fairbanks.  Based on this data, and with a requirement for no missing hourly reports in each 24-hour average, I calculated the theoretical daily evaporation rate from an evaporation pan (formula supplied on page 11-12 of this document - thanks Brian).

The chart below shows the monthly mean calculated evaporation rate during the growing season months since 2011.  Unfortunately the humidity measurements did not come online at the Fairbanks CRN until August 2010, so it was not possible to include a larger number of years.  Nevertheless, the results confirm the very high evaporation rate in June 2013, and interestingly the evaporation was higher in each month from May through August 2013 than in the same months of the previous two years.  The theoretical evaporation from June 2013 amounted to well over 8 inches for the month, which is not much less than the precipitation total for the entire year in Fairbanks.

Looking at the solar radiation and wind speed in isolation (see below), the solar radiation was higher last year in May and June than in the same months of any of the previous 10 years.  The wind speed was also a little higher than normal in June and July.  (The reported wind speed looks much too low for May and June of 2011, but I double-checked the data and this is what the supposedly high-quality CRN platform reported.  There is no evidence of lower wind speeds in these months at Fairbanks airport.)

In conclusion, the CRN data confirm that Fairbanks-area land-surface moisture loss in summer 2013 was caused not only by dry air but also by abnormally sunny conditions.  Of course we would expect these two anomalies to go together; both were quite extreme last summer, leading to highly elevated moisture removal rates.  Fortunately, fire activity was not catastrophic as in 2004 - probably because of lack of lightning.  Here's what the state 2013 fire report had to say: "2013 will be remembered as one of the shortest, but hottest summers on record... Fuel conditions reached near record dryness at some locations... Despite many record-setting hot temperatures, the stable high pressure kept thunderstorms at bay, and lightning to a minimum... With many comparisons between the hot, dry summers of 2004 and 2013, it seems that the lack of lightning, and therefore the lack of natural fire starts, is what kept the 2013 fire season from becoming catastrophic."