Pacific Temperature Patterns

On this blog there has often been discussion of the influence of Pacific Ocean temperature patterns on Alaska climate, and rightly so, as the evolving oceanic thermal patterns are known to create first-order impacts on seasonal (and decadal to multi-decadal) climate in this part of the world.  In this connection we usually talk about the El Niño - Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) as the two big players.

A few weeks ago reader Eric suggested that we could do an objective multi-variate analysis of the impacts of ENSO and the PDO to decipher which is more important at different times of the year.  The challenge here is that the two phenomena, as they are usually defined, are not at all independent, so it is impossible to obtain a clean separation of the impacts.  However, if we could re-define the different Pacific temperature patterns to be independent of each other, then we could proceed.  This is the approach followed by Dennis Hartmann of the University of Washington in a paper published this year:

http://www.atmos.washington.edu/~dennis/Hartmann-2015-Geophysical_Research_Letters.pdf

Hartmann looks at temperature patterns over a large domain from 30°S to 65°N across the Pacific and calculates the top 3 EOFs, which are the independent spatial patterns that dominate the variability over time.  The top pattern/EOF largely represents ENSO but also includes significant variability that we normally think of as the PDO (see Hartmann's figure 1, reproduced below).  The second pattern/EOF also resembles the PDO pattern, so we can think of the PDO as contributing to both EOF1 and EOF2.  Hartmann's third mode is what he calls the North Pacific Mode, and he focuses on this pattern as being implicated in the extreme circulation anomaly over North America in winter 2013-2014.

The independence of the three temperature patterns allows us to proceed with Eric's idea of a multi-variate linear regression, and so I've done this for Fairbanks monthly and seasonal temperatures since 1950.  The first chart below shows the estimated coefficients of regression between the 3 Pacific modes and Fairbanks monthly temperature.  This can be interpreted as the temperature anomaly that typically occurs in Fairbanks when each of the modes has an anomaly of 1 standard deviation; for example, when EOF1 is one standard deviation above normal in January, Fairbanks is typically 3 °F above normal.  Note that the coefficients are very small in summer for two reasons: first, because there is a less robust impact of the Pacific modes, and second because the summer temperature variability is much smaller.

It's interesting - to say the least - that Hartmann's North Pacific Mode (EOF3) has a small impact compared to ENSO and the PDO, and it is negatively correlated with temperature in most months.  I did not expect to see this, because the North Pacific Mode is closely related to "the Blob" of warm northeast Pacific SSTs that has recently gained notoriety, and I have speculated that this anomaly was partly responsible for the warmth of the past two winters in Alaska.  This analysis would suggest that we can't pin the recent warmth on "the Blob".

Here's the corresponding chart for seasonal (3-month) mean temperatures in Fairbanks.

If we look at the statistical significance of the 3 predictors in the multiple regression model, we find that the impacts of the Pacific modes are generally more robust for seasonal temperatures than for monthly temperatures, as we would expect.  The charts below show the t-values of the individual regression coefficients; the t-values directly indicate the probability that the coefficients could have been obtained by random chance.  On a seasonal basis, the ENSO/PDO impacts are highly statistically significant from autumn through late spring.  The NPM also shows up as significant at certain times of the year, but with a negative coefficient as noted above.

The overall correlation coefficients for the 3-predictor models (monthly and seasonal) are shown below - again generally higher for seasonal than monthly anomalies, and especially in the autumn.

From the perspective of seasonal forecasting it is interesting also to look at the statistical impacts on a lagged basis, because of course we don't know ahead of time how the Pacific temperature patterns will evolve in the months ahead.  The chart below provides a quick look at this; here I've repeated the procedure using seasonal Fairbanks temperatures and the Pacific modes one month ahead of the first month of the season - for example, the columns for DJF show the significance of the regression between November Pacific patterns and December-February Fairbanks temperatures.  The results indicate that the predictability associated with EOF1, the ENSO/PDO mode, is largely retained at a one-month lag, but EOF2 and EOF3 lose most of their predictive power.  This is not too surprising, as the coupled ENSO/PDO anomalies evolve slowly, but the 2nd and 3rd modes can change significantly over the course of a few months.

Last 80 Degree Day?

Yesterday may well have been the last 80+ °F day in Fairbanks for the season, and if so, the summer's total number of such days will end at 11, which is exactly average (1930-present and 1981-present).  Both June and July were slightly cooler than normal in the mean, but May was much warmer than normal and added 2 days to the count.

The chart below shows the climatological frequency of reaching 80 °F on or after the date shown on the horizontal axis.  As of today there is still nearly a 50% chance of reaching 80 °F again, but the forecast for the next week suggests little chance of it happening in the near-term, and after August 12 the frequency drops to under 30%.

Seasonal Transition

The weather in Fairbanks turned unusually wet in the last week of July, taking the month's rainfall total to 2.78 inches and into the top 20% of the historical distribution for July.  More than a quarter-inch of precipitation was observed on 4 consecutive days ending on the 30th, and remarkably this is only the 4th time this has happened in the history since 1930.  The previous occasions were in September 1942, April 2002, and August 1967, the latter being the occasion of the great Fairbanks flood.

Looking at a slightly broader timescale, Fairbanks recorded significant rainfall on 6 of the 8 days ending August 1st.  This is of course much more common, especially at this time of year.  Historical data show that 6 of 8 consecutive days have seen at least 0.05" of precipitation the following number of times since 1930:

8-day periods ending in May: 3 times
June: 17
July: 32
August: 59
September: 22

Nearly one-third of all Augusts see a wet spell of this kind, and of course the shift to wetter weather is a symptom of the seasonal transition to late summer.  Recently I've been wondering about the causes of this seasonal shift, so I created a few maps of climatological upper-level flow patterns to illustrate how the circulation changes over the course of the summer.

First we can look at the mean 500 mb height for June, July, and August: see below.  The upper-level height increases universally between June and July, as the atmosphere heats up in Northern Hemisphere summer, but the increase is especially pronounced over the North Pacific and Bering Sea as the Bering Sea trough is eroded.  A related change is that there's less of a tendency for upper-level riding over Alaska in July.

In August the upper-level height drops almost everywhere on the domain shown here, but the decline is greatest at high latitudes, and the 500 mb height actually increases over the Pacific south of about 50-55 °N.  This is a reflection of the changing distribution of heat input: in the northernmost areas, solar heating is much less in August than in July, and the atmosphere cools off accordingly, but farther to the south there is relatively little decrease in solar input and the atmosphere and oceans remain warm or even continue to heat up.

The effect of the changing height pattern is to increase the north-south gradient of height, especially over the Bering Sea, and this in turn causes the normal westerly flow to accelerate.  The following maps show the normal 500 mb wind speed in July (top) and August (bottom), and we can see that there is a widespread increase in wind speed over Russia, the Bering Sea, and Alaska.

The difference in normal 500 mb wind speed between July and August (below) clearly highlights the zone of energized flow circling the entire Northern Hemisphere, with the maximum change extending across southwest Alaska.

What does this transition mean for interior Alaska weather?  It implies a more progressive flow with a lower chance of persistent ridging and a higher frequency of upper-level disturbances that bring cloud and rainfall.  On the synoptic (large) scale the north-south temperature gradient strengthens from July into August, and it is this gradient that generates and feeds cyclonic weather systems.

To wrap up I'll show the frequency of 500 mb blocking highs and lows.  First, blocking high pressure systems (see below) become less common from July to August across most northerly regions, reflecting the strengthened jet stream; a stronger circumpolar circulation is less prone to developing "kinks" that produce blocking weather patterns.  Alaska in particular sees a much reduced frequency of cut-off high pressure systems as the summer advances.

The frequency of cut-off low pressure systems is not as easy to interpret (see below), although the frequency increases over southwestern mainland Alaska over the course of the summer.

In summary, it's the seasonal intensification of the north-south temperature gradient in the atmosphere that drives the circulation changes in late summer across Alaska and other high-latitude parts of the Northern Hemisphere.  The North Pacific region is especially prone to this seasonal change, presumably because the Pacific Ocean, being the world's largest water body, creates an especially large lag in seasonal warming at middle and lower latitudes.

El Niño Update

It's been almost three months since I discussed the strengthening El Niño episode in the tropical Pacific Ocean, so it's time for an update.  There have been no big surprises, as El Niño has continued to strengthen, and sea surface temperatures have reached levels characteristic of classical "strong" El Niño events, i.e. at least 1.5 °C above normal in the Niño 3.4 region.  See here for NOAA's weekly technical discussion, with lots of graphics.  Here's a recent SST anomaly map showing the El Niño anomaly in the eastern tropical Pacific and of course ongoing warmth in the eastern North Pacific.

The long-range computer model forecasts expect that El Niño will continue to strengthen, and indeed they foresee an acceleration to very intense levels in the relatively near future: see below for an example.  However, it is interesting to observe that earlier forecasts (e.g. from May, see below) were expecting El Niño to be already considerably stronger than it is now, so the evolution is lagging what was expected.  To my mind this is a hint that the episode may peak at lower intensity than is currently predicted and it may be "merely" a strong El Niño rather than a super or record-strength El Niño.

Another interesting aspect of this year's ENSO (El Niño - Southern Oscillation) behavior is that it's rare to have moderate or strong El Niño conditions in two consecutive winters.  The scatter chart below compares ENSO index values for consecutive years in the October-December season.  The top highlighted rectangle draws attention to the top 10 El Niño events since 1900, and we see that the prior year ENSO index was always between about -0.8 and +0.7.  Last year's October-December ENSO index was +0.92, so we'll fall outside the top rectangle this year unless El Niño suddenly weakens.

Looking at the bottom rectangle, we see that the strongest La Niña events tend to follow either a significant (but weaker) La Niña event or a moderate to strong El Niño in the previous year.  These features of the chart may seem like only sampling variability, but it is generally accepted that La Niña episodes tend to be longer lived than El Niño events.  In 115 years of data plotted here, only twice was the ENSO index above +0.75 in two consecutive years, whereas on 8 occasions it was below -0.75 in two consecutive years.

Fire Weather Climatology

A couple of weeks ago I wrote about Alaska's daily fire acreage numbers from the last couple of decades and noted that the median burn rate drops off precipitously after about the first week of July.  Since then I've been wondering if we can identify any sharp changes in the Fairbanks weather climatology that would help explain the sudden shift to less favorable fire conditions at that particular point in the summer.

The first thing to note is that the normal temperature and rainfall don't indicate any sudden changes; the normal daily maximum temperature passes through its smooth seasonal peak at the beginning of July, and rainfall frequency and amounts generally increase into early August.  There's actually a downward blip in rainfall frequency in early July, although this is presumably an artifact of the sample size.

The cloud cover and humidity also increase as the summer advances.  The chart below shows the daily mean cloud cover and the normal daily minimum relative humidity as reported from hourly observations.  The upward trend in daily minimum humidity is quite striking, and since we know (see e.g. here) that low humidity is a key component of fire danger, this seems like part of the puzzle.  However, the seasonal change in humidity is gradual with no sudden change in early July.  The cloud cover does tick up a bit in early July, but this is a small shift in the climatology.

There are other variables that could be looked at (e.g. lightning frequency or wind), but as with the examples above, it seems likely that the normal curve for any other individual variable will also show rather gradual changes over time.  However, when we combine variables and look at the joint distribution of two or more variables, some different behavior can show up in the tails of the distribution.

For example, the chart below shows the frequency of warm and dry conditions that might be considered very favorable for fire growth.  In late June and very early July, 80F temperatures with sub-30% humidity occurs on more than 15% of all days, but the frequency drops rapidly to under 10% by July 15 and under 5% by July 21.  Comparing this to the normal fire growth rate curve (also below) and recalling that the second chart shows week ending dates on the horizontal axis, there is a rather good correspondence.  Within just a few days of the start of July, favorable fire growth conditions become much less common in Alaska's interior, and as a result the median fire growth rate comes almost to a halt.

A note to regular readers: I'm traveling at present, so posting and commenting activity on the blog will be light for the next several days.

Chilly and Damp

Sunday's weather in Fairbanks was unusually chilly, and not just in comparison to hot conditions earlier this summer.  The high temperature of 55°F was the lowest high temperature in the month of July since July 31, 2008.  It wasn't quite a record, though: July 19, 1965 saw a high of 54°F.  The July record for daytime coolness is 48°F in 1934 and 2003.

Thick cloud and light rain were of course the direct cause of the temperature suppression yesterday.  The 500 mb height analysis from 4pm AKDT (below) reveals a trough along the 140th meridian and a weak ridge over southwest Alaska.  In the northwesterly flow between the two features, some subtle vorticity (cyclonic flow) features and warm advection appear to have caused the rain and cloud cover.

Here are the 500mb and 850 mb vorticity maps from 10am yesterday; yellow and orange shading shows enhanced vorticity, which when transported across a region typically brings cloud and precipitation.

The diurnal temperature range yesterday was only 3°F (high of 55, low of 52), which is only the second time in Fairbanks' history that the temperature has remained so steady for an entire day in July.  The previous occasion was July 11, 1990, when 1.24" of rain fell.

It is rare for the temperature to remain so steady in the height of summer, because the long daylight hours and large solar radiation nearly always allow the temperature to jump up by at least a few more degrees.  So it's even more remarkable that yesterday's small diurnal range occurred in conjunction with only 0.08" of rain.  This has actually never happened before from May through August; the previous smallest amount of rain associated with such a small diurnal range in these months was 0.23" on August 21, 1978.  So in this sense yesterday was a record-breaking day.

A Change of Scene

A very welcome change in the weather pattern has developed in the past several days, with western and southern Alaska coming under increasing influence from low pressure in the Bering Sea and Aleutians.  The contrast from the second half of June is great: see the 500 mb height anomaly maps for June 16-30 and July 1-15, below:

Significant rain has fallen or is now falling in parts of the western interior, with McGrath - highlighted here for its drought conditions just 3 days ago - recording 1.23" of rain in the past 3 days.  More is to come in association with the current cyclonic system that is moving over western Alaska.  The 500 mb height analysis below, courtesy of Environment Canada, shows the situation at 4am AKDT this morning.

Today's upper-level low is fairly strong for the time of year: Bethel observes a 500 mb height this low in July in fewer than 1 in 4 years (although interestingly it happened in each of 2011, 2012, and 2013).  In terms of standardized anomaly, it is the most negative 500 mb height anomaly in more than 2 years in Bethel; in McGrath it's the most negative since February of 2014.  The chart below shows the observed 500 mb height anomalies at McGrath since the beginning of 2014.  Note the persistence of above-normal heights throughout last winter and in the past couple of months.

Here are some radar loops from Bethel today, showing the cyclonic circulation associated with the system.  The animations show consecutive 2-hour periods.

Another symptom of the change is that - according to my data - nowhere in Alaska has exceeded 80 °F since July 10.  Let's hope the new pattern sticks around: the state's fire acreage growth has completely stalled in the past few days.

Warm and Dry Extremes

In pondering this year's record-setting fire season in Alaska (up to 4.4 million acres burned as of yesterday), I found myself wondering "just how dry and warm has it been?"  The extreme fire behavior suggests that the accumulated climate anomalies over the past several months must have been equally extreme, so it's of interest to see if the data reflect this.

First I'll show results for Fairbanks.  The chart below shows the temperature and precipitation ranks for periods of varying length ending with June.  The ranks are calculated with respect to the history since 1930.  So for example, June temperature by itself was very close to the 1930-2014 median temperature, but June precipitation was ranked 31 out of 85 previous years (i.e. percentile rank of 36%).  If we do the calculation for May and June combined, the temperature was ranked 93% and the precipitation was ranked 25%, so the May-June period was much warmer than normal and considerably drier than normal.

As we include more months in the average, going farther back into spring and winter, the temperature ranks generally keep growing and the precipitation ranks keep dropping, until we find that the most recent October-June period (9 months ending in June) was ranked 96% for temperature and only 14% for precipitation.  Going even farther back, however, we see the effects of last year's very wet summer in Fairbanks, which raises the precipitation ranks as we include those months.

Now take a look at the chart below for McGrath.  All running averages from 1-12 months are above the 94th percentile for temperature (with several being records), and periods from 4-12 months are below the 10th percentile for precipitation.  In other words, it has been both extremely warm AND extremely dry in McGrath over seasonal timescales leading up to this summer.  In view of these numbers, the extreme fire season is anything but a surprise.

The figures below show another way of looking at this year's temperature and precipitation within the historical distribution, for a specific combination of averaging periods (6 months for precipitation, 3 months for temperature).  I picked this combination because McGrath had the driest January-June period on record and the warmest April-June period.  I've also colored the years that had strong El Niño or La Niña conditions in April through June, just to see how strong the El Niño connection is.  It seems that recent warmth is consistent with the El Niño influence in both Fairbanks and McGrath, and there's a slight tilt towards drier conditions in Fairbanks during strong El Niño conditions (as we noted here).  However, La Niña does not necessarily bring the opposite anomalies, as some of the strongest La Niña years were also very dry in Fairbanks in late spring.

Here's a graphical depiction of this year's record year-to-date precipitation deficit in McGrath.  The moisture situation has improved slightly in the past couple of weeks, but we're still at the driest year-to-date on record.

Fire Acreage

There are probably not many residents of interior Alaska who aren't aware that this year's fire season is on pace to exceed the record year of 2004 in terms of total acreage burned.  It still has a long way to go to get there, but the acreage burned so far is well ahead of 2004.  Here's a graphical depiction of that fact: the chart below shows the daily accumulated fire acreage for each year since 1996, based on daily reports archived at the National Interagency Coordination Center (NICC):

I've added the 1996-2014 median and mean year-to-date acreage as solid and dashed black lines respectively.  Obviously the total fire acreage has a very skewed distribution, as a few years are far above the long-term normal while most years have relatively insignificant fire activity.

It's interesting to note that the median acreage basically stops increasing after about July 7, whereas the mean acreage keeps rising into August and even September.  What this means is that fire activity drops to nearly zero after early July in a normal year, but in big fire years, activity continues into late summer.  In other words, some fire activity is normal in Alaska in June and early July, but by mid-July it is unusual to have significant activity; this reflects the seasonal change in weather, with increased humidity and rainfall in July usually being sufficient (under normal fuel moisture conditions) to halt fire growth and stop new fire development.

A closer look at the normal rate of burn (see below) shows that the median and mean burn rates both peak around July 8 (week ending date), and the median rate drops extremely rapidly thereafter.  However, the mean fire acreage growth drops off gradually through late summer.  Given that we are currently in an extreme fire year, we would not expect fire activity to cease suddenly now that we're past early July, but rather we should see a gradual decline whose details depend on the weather.

Looking at the daily fire acreage numbers, this year has the top 2 spots for daily acreage burned.  Below are the top 5 in the history since 1996 (note that I'm using the dates of the NICC reports, which lag the actual fire dates by a day or two).  According to the July 7 report, 0.15% of the state's land area burned in one day.

546321 acres  July 7, 2015
511604 acres  July 1, 2015
455214 acres  July 4, 2004
368823 acres  July 3, 2015
303492 acres  July 15, 2004

One more remarkable statistic: the week ending June 30 this year saw 1.85 million acres burned, which is 46% higher than the worst week in 2004.  It's hard to overstate how extreme the situation has been.

Heat in Historical Context

Residents of the eastern interior suffered through a sweltering summer day yesterday - quite possibly the hottest of the summer.  The temperature reached a high of 89 °F at Fairbanks and 87 °F at both Eagle and Bettles.  In Fairbanks and Bettles, temperatures this high occur in slightly more than 1 in 3 years over the long run.

Reader Gary noted that the temperature in Fairbanks remained high late into the evening yesterday and suggested that this duration of warmth seemed exceedingly unusual.  It's interesting to consider what the historical data reveal about comparable past events.  First, we can note that the temperature at midnight (standard time) was 77 °F, which is indeed very unusual.  Here are the highest temperatures reported in the hourly observations for midnight AKST:

79F  July 12, 1975
78F  July 11, 1975
78F  June 22, 1987
77F  June 22, 1991
77F  June 21, 2013

So last night's warm midnight temperature ties for 4th highest in the historical data since 1950.

As for the duration of excessive warmth, yesterday produced 8 consecutive hourly observations at or above 85 °F at the airport, from 1pm to 8pm AKST inclusive.  Looking at past events, this is again close to but not quite a record.  Since 1950, Fairbanks has recorded 4 events with 9 consecutive hours of 85+ °F, 3 events with 10 hours, and 1 event with 11 hours.  Here are the top durations of 85+ heat:

11 hours:  11am - 9pm June 21, 1991
10 hours:  Noon - 9pm June 15, 1969 (including 94F at 6pm; this was the all-time hottest day in Fairbanks, 96F)
10 hours:  Noon - 9pm July 10, 1975
10 hours:  Noon - 9pm July 11, 1975
10 hours:  Noon - 9pm June 25, 1983

Note that the 1969 event may have been an hour longer in reality because observations were only taken once every 3 hours at that time, and the noon observation was already 88 °F.

Here's a chart showing the number of times that multi-hour periods have exceeded certain durations with the temperature at or above 85 °F.

This gives some additional context for yesterday's hot spell and shows that while very unusual, the airport observations (8 hours at 85+) do not indicate that the heat itself was historic.  However, this doesn't exclude the possibility that the heat was slightly more intense or long-lived in other parts of Fairbanks-land, or that it produced a more oppressive sensation due to other factors.  One possibility is that the dewpoint may have been very high in parts of town; it rose to 60 °F at the airport at 10pm, which is certainly very unusual - I can find only one historical instance of comparable temperature and higher dewpoint at that time in the evening (August 5, 1994).

For additional background, the chart below shows the annual number of hours with a temperature of 85 °F or above in Fairbanks.  Note that from 1965-1972 the observations were taken once every 3 hours rather than every hour, but I've accounted for that difference in the calculation.

It's interesting to note that except for the major outlier in 2013, the number of hot hours has been relatively low since 1995.  It will certainly be interesting to see how the remainder of this summer turns out.

Further documentation of yesterday's event can be found below:

Yesterday's 9pm AKST sea-level pressure analysis (note the thermal trough and heat low close to Fairbanks):

Today's 3am 500 mb analysis:

A surface observation map showing the evening heat in the eastern interior (note also an extraordinary 77 °F at Wainwright on the Arctic coast):

A time-height temperature cross-section from a model run initialized yesterday morning, showing the late peak in temperature, especially aloft:

How Good is the Reanalysis?

As regular readers here know, the NCEP/NCAR reanalysis is a very useful tool for examining historical climate conditions across the globe.  The reanalysis contains a complete estimate of the state of the atmosphere every 6 hours all the way back to 1948; here's a link to the original publication that introduced the groundbreaking project.  On this blog we've often used the reanalysis data to explore Alaska climate, for example here and here.

A question that naturally arises is, how well does the reanalysis data correspond to reality?  In the paper linked above, the authors note that variables like upper-air temperature and wind are strongly constrained by the observations, and so are generally very reliable, but other variables are more influenced by the model's internal physics, because there are few or no direct observations of these variables.  Surface temperature is an example that is somewhat constrained by the observations, but precipitation is completely determined by the model (i.e. no precipitation observations are assimilated by the model).

In view of this suggestion that the temperature and precipitation analyses might not be that great, I thought it would be interesting to look at Fairbanks data to assess the fidelity of the reanalysis.  I took the monthly mean temperature and precipitation from the nearest reanalysis grid point to Fairbanks and compared it to the observations from Fairbanks airport; the chart below show the correlations by month, with rank correlation used for precipitation.

We see that the reanalysis temperature is very good (correlation above +0.9) from October through April, but the performance drops off very dramatically in the summer.  Remarkably, the reanalysis temperature in July is almost uncorrelated with what actually happens; but interestingly August is much better.

As expected, the precipitation analysis is less good, with correlation coefficients of only about 0.6-0.7 in winter, and very low correlations in June and July.

The chart below shows the results for 3-month periods, which we might expect to fare a little better because the random daily fluctuations tend to even out.  Unfortunately this isn't true, as the correlation values are similar to those for monthly data.

The reanalysis performance is so poor in summer that I had to look at the data more closely; see below for year-by-year comparisons of the temperature and precipitation in May-June-July, which is the worst-performing season in the chart above.  The temperature chart is fascinating, because it shows that while the reanalysis generally captures the sign of many of the changes from year to year, it has a serious problem with decadal trends.  Remarkably, the reanalysis temperatures became substantially colder from the 1970s to the early 1990s, while the observations showed the reverse, and since 1995 the reanalysis has become much warmer while the observed trend has been small.  I can't say for sure what might cause this, but it probably arises from changing systematic bias in the model either when new observations are introduced (e.g. satellite data from 1969 on) or when the ocean temperature patterns shift (e.g. the Atlantic or Arctic temperatures).  After all, the reanalysis is only a model, with lots of assumptions and physical parameterizations.

The May-July precipitation comparison shows more what we would expect for a system that has very little skill; the model is simply unable to reproduce the precipitation amounts in the Fairbanks area in summer.  This isn't surprising at all, because summer precipitation in interior Alaska mainly occurs from localized convective activity or from rather small upper-level features, which are either not resolved on the coarse grid or not well-represented in the model physics.  An interesting point is that the reanalysis has shown generally wetter summer conditions since 1995, and especially since 2004, but there is less change in the observations.

It would be interesting to look at other locations in Alaska to see what the regional performance is like; and I also want to look at the more modern Climate Forecast System Reanalysis (CFSR) from 1979-present, to see if it does a better job in the warm season.  In the meantime, however, it seems we should be very cautious about using or accepting results based on the reanalysis warm-season temperature and precipitation over the Alaskan interior.