Monday, February 29, 2016

Warmth in Context

Fairbanks has reached 40°F or higher on each of the last 4 days, which is pretty remarkable for the time of year.  Only February 1942 had a higher number of 40+ degree days (6 of the last 7 days of that month).  In the long-term history since 1930, the highest temperature normally observed in December through February turns out to be 40°F, so half of all winters never get out of the 30s during the 3 coldest months (and a few do not get out of the 20s).

Unusual warmth has been a persistent feature of Alaska's weather for over two and a half years now, ever since the dramatic flip in 2013 from a very cold spring to an extremely warm June.  (Check out this blog's archives for the commentary at the time.)  Interestingly if we look at the Alaska climate division data that became available last year, that first warmth month (June 2013) was one of the most anomalous months in the warm spell, as the statewide (13-division) average of the standardized monthly temperature anomalies was +1.96 standard deviations.  This compares to +1.94 SD in October 2013, +1.88 SD in January 2014, and +1.96 SD in May 2015.  For comparison, January of this year came in at +1.71 SD relative to the 1981-2010 normals.  (Note that this is a simple average; the land areas of the divisions are not equal.)

The chart below shows the long-term history of the statewide average standardized anomaly, with a 12-month running mean applied (1981-2010 normal used throughout).  It's interesting to note that the current warm spell appears comparable to the warm period that extended from late 2002 through 2005; so far, Alaska's latest warm spell is not warmer than that previous period on a 12-month average basis.  This is confirmed by looking at 12-month running means of the statewide area-average temperature in the same climate division data: the 12 months ending August 2003 had a mean temperature of 30.95F, and the 12 months ending January 2016 had a mean temperature of 30.93F.  (However, the 12 months ending in February 2016 will presumably be warmer.)


If we look at only winter months, we get a rather different picture, as previous warm spells did not produce back-to-back winters with such anomalous warmth.  (This winter's data is complete only through January, but we know February will be very warm, and March seems unlikely to be very cold.)  The most comparable event in the past may be the back-to-back warm winters of 1943-44 and 1944-45.


We can get a different perspective on the extent of the warmth across the state by looking at the fraction of the 13 divisions that had temperatures in each of the 3 climatological terciles (above-normal, near-normal, and below-normal).  After doing this calculation for each month, and again applying a 12-month running mean, we get the chart below:


Again we see that the current warm spell looks rather comparable to that of 2002-2005.  There were a couple of other periods (1993 and 1997-98) that showed similarly widespread warmth, but for shorter periods.

Here's the chart for just the winter averages of the monthly tercile frequencies.  The absence of any significant cold in the past two winters is remarkable; only one climate division (North Slope) was colder than normal for only one month in December 2015.


Thursday, February 25, 2016

More on Barrow Variance

As a follow-up to my recent post about Barrow's remarkable drop in temperature variance, I thought it would be useful to look at a map of variance changes to see how widespread the phenomenon has been.  Has the annual mean temperature become much more stable in recent years over most of the Arctic or even farther afield?  Or is it a localized change near Barrow?  Using the 2m temperature data from the NCEP/NCAR reanalysis, here's the answer (click to enlarge):


According to reanalysis data, Barrow's reduction in temperature variance is not representative of the Arctic region as a whole, as strongly contrasting changes have occurred from place to place.  However, neither is Barrow's change a hyper-local phenomenon; there is a zone of much diminished variance that extends east along the south coast of the Beaufort Sea.  Looking farther afield, there are regions of strongly decreased variance near Russia's New Siberian Islands and in the Greenland Sea, but variance has increased strongly over the Kara Sea (and the North Pacific!).  It seems likely that at least some of the changes in the Arctic are related to the pronounced sea ice extent changes of recent years.  It's not as simple as saying that variance has decreased because ice cover is reduced, but it is notable that much more of the Arctic has seen reduced variance than increased variance.

The "before" and "after" periods in this comparison look quite different - see below.  The zone of low variance from eastern Alaska through the Arctic to the North Atlantic is rather striking in the analysis for the past 15 years.



In the earlier post I mentioned that Barrow's decrease in annual temperature variance was consistent with reduced autocorrelation, i.e. reduced persistence of temperature anomalies.  The chart below illustrates this change; it shows the evolution of the 10-year running autocorrelation of monthly mean temperature anomalies, at a 1 month lag (blue) and 3 month lag (red).  Note that a correlation coefficient greater than +/- 0.18 is statistically significant at the 95% level for a 120-month sample.  Most of the time prior to the mid-2000s, the lag-1 autocorrelation was significantly positive, meaning that consecutive months tended to have the same sign of temperature anomaly.  But in recent years the autocorrelation has collapsed to about zero.  The lag-3 autocorrelation has also decreased, changing from marginally positive to slightly negative.


The chart below shows the results for lags of 6, 9, and 12 months.  The lag-6 autocorrelation has recently decreased to a statistically significant -0.18, so temperature anomalies are tending to cancel each other out at opposite ends of the year.  This is consistent with the overall decrease in annual variance, but the fundamental reason for the change is not yet clear.


To conclude for today, here is a chart of annual mean temperatures and running variance at Cold Bay, where the reanalysis suggests that variance has increased in recent years.  The observed temperatures from Cold Bay confirm that variance has increased, and in this case the cause is clear: the North Pacific temperatures have fluctuated widely as the PDO phase has undergone dramatic shifts from year to year.  The extraordinary warmth of the last 3 years in Cold Bay has been particularly unusual in light of the long-term history.


Saturday, February 20, 2016

Fairbanks Update - Low Variance

Fairbanks airport managed to drop to -21°F yesterday morning, which was 9°F below normal for the date and just inside the "below-normal" category (lower tercile) for daily minimum temperatures.  The noteworthy aspect of this is that it's only the second time that -20°F has been reached this year so far.  In the history since 1930, only 1977 started off with so few -20°F days in January and February.  It seems unlikely that Fairbanks will see any more such days this month, as the forecast is very warm indeed.

Despite the lack of cold, however, 2016 is by no means the warmest year on record for the first 50 days of the year; it's actually only running as the 8th warmest on record.  1977 started off more than 5°F warmer than 2016, and 1981 was almost 10°F warmer through this date.

The profound lack of cold reflects the extremely low variance of temperature thus far, as we discussed before.  As of yesterday, the year-to-date standard deviation of daily mean temperature anomalies is the lowest on record at only 8.5°F and barely over half the usual value of 16.4°F.  For context, the 1981-2010 normal standard deviation for this time of year is close to 8.5°F in Valdez and also down in my neck of the woods in Athens, GA.

Fairbanks is also tying 2010's record pace for driest start to the year, with only 0.06" of liquid-equivalent precipitation thus far.  So with the combination of non-variable temperatures and lack of precipitation, this might be considered the most boring start to the weather year ever in Fairbanks.


Wednesday, February 17, 2016

Seasonal Forecast

In my day job I spend a good deal of time working with seasonal weather forecasts, and recently I have focused on developing a new scheme to calibrate seasonal forecasts from climate models such as NOAA's CFSv2 model.  The idea here is to transform the model ensemble forecasts into reliable probabilities so that users can make confident decisions.  The word "reliable" has a technical meaning: if a probabilistic forecast is reliable, then the predicted probabilities correspond to the observed frequency of occurrence over a long period of time.  For example, if the forecast says there is a 70% chance of above-normal temperatures, then over the long haul above-normal temperatures will be observed 70% of the time.  It's important for users to have confidence that forecast probabilities are reliable in this sense; if they're not, then the forecast isn't suitable for use in quantitative decision systems.

There is a lot more that could be said on this topic, but I'll just illustrate the probabilistic forecasts by showing the latest CFSv2 seasonal forecast for Alaska.  The maps below show forecasts for temperature, liquid-equivalent precipitation, and 10m wind speed, for March.  It's traditional with seasonal forecasting to divide the historical data into three equally-likely categories of below-normal, near-normal, and above-normal, and then predict the probabilities of each category.  In the forecasts shown here, all three probabilities are predicted at each point, but the map shading indicates the highest of the 3 tercile probabilities.  According to the CFSv2 forecast, significantly warmer than normal conditions are quite likely over all of interior and northern Alaska, with the probability exceeding 60% in the eastern half.  The forecast also favors drier than normal and less windy than normal conditions, but with much lower probability.



The rather high probability of very warm conditions in March reflects two facts: the raw model output is showing a large departure from normal, and the model has some skill at predicting March temperatures at this lead time.  The first point is illustrated by showing a comparison of this year's forecast for Fairbanks with forecasts from previous years (which are run in retrospective mode) - see below.  The blue dots represent the individual ensemble member forecasts made in February, and the red markers show each year's ensemble mean.  The bold horizontal lines indicate the tercile boundaries for the raw model forecasts.


Remarkably, the current forecast for March shows warmer conditions than in any of the retrospective forecasts from 1982-2010.  In the forecast history, the highest temperature predicted for March was -5.4°C, but 4 of the 40 latest members are showing a March temperature higher than that. (One caveat: the temperatures are not bias-corrected here; they are simply taken from the nearest model grid point.)

So the model has a strong warm signal this year.  But does that mean a warm March is actually more likely?  Obviously if the model has no skill, then it doesn't mean anything.  This is where the historical calibration comes in: the retrospective forecasts tell us how good the model is, and the calibration scheme transforms the model signal into a reliable probability.  Here's a chart showing the performance of all forecasts at 1-month lead time from January through March, i.e. January forecasts for February, February for March, and March for April.  The performance isn't great, but it's enough to do something with.

In a few days NOAA's Climate Prediction Center will release an official forecast for March, so it will be interesting to make a comparison.  In the meantime, here are the tercile probabilities for Fairbanks and Anchorage (closest grid points):

Fairbanks March temperature: 12% below-normal, 27% near-normal, 61% above-normal
Fairbanks March precipitation: 35% below-normal, 43% near-normal, 22% above-normal

Anchorage March temperature: 6% below-normal, 23% near-normal, 71% above-normal
Anchorage March precipitation: 28% below-normal, 31% near-normal, 41% above-normal

Finally, here's the forecast for the 3-month period March through May.  The probabilities of warmth are even higher, exceeding 70% in parts of the interior.  It probably goes without saying that this may bode ill for the early part of the fire season.


Update Feb 18: here's the performance chart using actual temperature observations from Fairbanks airport rather than the CFS Reanalysis for verification.


And here's the CPC forecast, hot off the press today.  The CPC has similar high probabilities of warmth for March, but somewhat lower probabilities for the March-May season.  Also, they expect relatively wet conditions in southeast Alaska during March, whereas the CFSv2 is showing dry.



Update April 20: here's the verification for March.  The CFSv2 did better than the CPC for precipitation in southeast Alaska, as it turned out to be relatively dry.




Saturday, February 13, 2016

Barrow Climate Mystery

A couple of days ago Rick Thoman sent me a very interesting plot of Barrow annual mean temperatures, showing not only the dramatic warming of recent years but also a remarkable collapse of interannual variance.  I've reproduced the essence of the chart below.  Until the early 2000s, the standard deviation of annual temperatures was fairly steady, hovering around 2°F, but in the past decade or so it has dropped precipitously as cold outliers have ceased to occur and every year falls within a much-reduced range of temperatures.



Long-time readers will recognize that the chart is rather similar to that for October mean temperatures, which have similarly become radically less variable; Rick pointed this out some years ago.  Here's an updated chart for October mean temperatures.



It's obvious, I think, that the reduction in annual temperature variance partly reflects the October change, but curiosity led me to examine all months of the year to see which other months might have contributed to the annual change.  The chart below shows the standard deviation (SD) of monthly mean temperatures for each month, both for the 1930-1990 period and for the past 10 years.  October stands out as having the greatest drop in SD, but January and February have also become much less variable.


The interesting part of this analysis is that the average of the SD decrease in all 12 months is only about 24%, whereas the annual SD has dropped by 64%.  At first glance this doesn't make sense - how can the annual variance drop so much when the individual months are, on average, only modestly less variable in the modern climate?  Part of the answer is presumably that the winter temperature variance contributes more to the annual variance, and the variance has dropped much more in winter than in summer; but it's still puzzling, as no other month besides October comes close to the percentage variance reduction that has occurred on an annual basis.

To explore this question in more detail, I did some simulations of monthly mean temperature variability and combined the months into annual temperature values to see what we would expect from random chance.  Specifically, for each year in the history I calculated the observed mean and variance of monthly mean temperatures within a +/- 10-year window and then created 1000 instances of 10-year periods by taking random samples of each month from an assumed Gaussian distribution.  Annual mean temperatures were calculated from the 12 monthly values in each instance, and I assumed that each month is independent of the next, i.e. zero autocorrelation.  Finally, for each 1000-member sample of 10-year periods I obtained the 90% confidence interval on the annual SD, and this was all repeated for each year from 1930-2015.

Don't worry if you didn't follow the details; the end result is a statistical estimate of what the annual standard deviation should be given a Gaussian distribution of monthly temperatures and assuming no correlation from month to month.  The chart below shows the result.  For the majority of the history the observed decadal variance has been considerably higher than the expected value, and for much of the time it even lay above the 95th percentile of the synthetic distribution.  This makes sense, because month-to-month temperatures are not independent and the positive autocorrelation increases the variance.  In other words, a very warm month is more likely than not to be followed by another warm month, so annual temperatures vary more than you would expect if you don't account for that.



But look at what has happened in the past few years; the observed variance has dropped well below the expected value and is not far above the 5th percentile.  This means that over the past 10-15 years, annual temperatures have varied less than we would expect from the monthly temperature variability, even if there was zero autocorrelation!  Of course this could happen just by random chance - that's the point of the 90% confidence interval - but it's quite surprising.  Does it mean that the autocorrelation is now negative?  Yes - it appears that temperature anomalies separated by 6 months have been negatively correlated in Barrow in the past decade; so the monthly anomalies have tended to cancel each other out, allowing the annual temperatures to become remarkably stable.

In conclusion, there appears to be quite strong evidence that the climate in Barrow has shifted towards a more stable temperature regime in which departures from the new (warm) normal are much less persistent than in earlier decades.  It used to be that some years would see persistent unusual cold or warmth for many months, leading to larger annual anomalies, but this doesn't happen any more.  In the past 8 years, every single year has had either 5 warmer than normal months and 7 colder than normal months, or vice versa, or 6 of both (relative to the decadal normal).  This kind of behavior used to be abnormal.

I'll leave it to another post to speculate about causes of this change in Barrow's climate.  For now, it's a bit of a mystery, but an interesting one, because it exposes an aspect of climate change that I've not seen discussed before.  It would certainly also be interesting to see if the same thing has happened elsewhere in Alaska or farther afield.

Saturday, February 6, 2016

January Arctic Warmth

A blog post by Weather Underground's Bob Henson caught my eye on Thursday, as he discussed the extraordinary warmth that occurred in the Arctic basin during January.  To put this in context, I calculated the 60-90°N area-average mean temperature at 925mb (about 2200' elevation) for each January since 1950, according to the NCEP/NCAR reanalysis; the results are shown below.  The temperatures last month set a new record for January warmth in the reanalysis era, by a large margin.


It's interesting to see that a similar kind of outlier (for the time) occurred in January 1977, which is the winter that we've recently discussed as being very warm in Fairbanks.  January 1977 occurred just after the "great Pacific climate shift" and the onset of what turned out to be an enduring positive phase of the PDO.  One wonders if this means that the next decade or two will see a continuation of the positive PDO phase that we've observed in the past 2 years: have we just passed through another major Pacific climate shift?  From an Alaska-centric viewpoint, have we shifted into a new normal of winter warmth, as happened in the decades following 1976?  Let's hope not.

Looking at daily 925 mb temperatures since the beginning of 2015, the Arctic (and sub-Arctic) area has seen persistently above-normal temperatures, but since late December the area-average temperatures have been near or above previous record levels for the time of year.


On the chart below I've added the daily temperatures for the past two very strong El Niño events, 1997-98 and 1982-83, and we can see that neither of those winters produced unusually warm conditions over the Arctic, at least through January.  This suggests that we shouldn't pin the blame for the recent warmth on El Niño.  It has more to do with the transition to a strongly negative Arctic Oscillation at the beginning of January, which allowed cold air to spill south into the mid-latitudes and, conversely, warm air to invade the Arctic from the south.  I believe there is also a close connection to the stratospheric polar vortex, which was extremely intense in November and December, but began to undergo a weakening trend in January.


The following maps show the January temperature's departure from normal beginning at the surface, moving up through the troposphere (up to 300mb) and into the stratosphere.  There is a dramatic transition from extremely warm conditions in the lower-to-mid troposphere to unusually cold conditions in the stratosphere; note that I've doubled the range of the scale on the last two maps to accommodate the magnitude of the cold anomaly aloft.  The cold in the stratosphere was associated with the strength of the polar vortex, which remained much greater than normal in January.











Wednesday, February 3, 2016

Lack of Cold and Low Variance

There have been a few mentions recently of the lack of 30-below temperatures so far this winter in Fairbanks; the coldest temperature observed at the airport was -29°F on Christmas Day.  If this remains the coldest temperature of the winter, it would be almost unprecedented for lack of cold: in the past, only the winter of 1976-77 failed to reach -30°F (lowest temperature -28°F).  The lack of cold is similarly unusual if we look at the 850mb level aloft, where this winter's coldest temperature as measured by balloon sounding was only -14°F on November 18.  Only once before, in 2000-2001, did the Fairbanks soundings fail to measure a colder 850mb temperature at some point in the winter.

How likely is it that -30°F will be reached at the airport in the remainder of winter?  Based on the last 40 years of historical data, more than 70% of years reach -30°F on or after February 5, and the odds don't drop below 50% until February 22.  Even in March it's not too uncommon - in fact, 4 of the last 5 years have seen -30°F or colder in March.  However, the medium-range forecast is quite warm (see below), and of course the ongoing El Niño and positive PDO phase suggest that warmth will continue to dominate.

 



Despite the overall very warm pattern and lack of cold conditions this winter, the number of very warm days has not been particularly unusual.  Since November 1, the temperature has risen above freezing on only 5 days in Fairbanks, which is only slightly above the long-term normal of 4 days (in the period Nov 1 - Feb 2), and far below the record of 19 days in 1936-37 (Nov 1 - Feb 2).  In other words, the variance of temperature has been low, which is a characteristic of strong El Niño winters.

The chart below shows how the January-March standard deviation of daily mean temperature anomalies (departure from normal) varies with an index of El Niño/La Niña behavior.  The standard deviation is quite noticeably reduced when the ENSO index is above +1, so there's little doubt that the El Niño episode is contributing to the lack of variability in Fairbanks this winter.


Fairbanks winter temperature variance is similarly affected by the PDO phase (which is of course correlated with ENSO), and if we create a combined index of PDO+ENSO behavior, the overall correlation with the variance is slightly greater than for ENSO alone.

Here's a chart showing how the ENSO/PDO effects on temperature variance change through the year; each column represents the correlation coefficient for a three-month period.  Interestingly the variance reduction for El Niño is slightly greater in late winter than early winter, which probably reflects the fact that El Niño's impacts on the atmospheric circulation reach their peak in late winter.  The opposite effect is observed in late summer and autumn, with temperature variance being somewhat enhanced during El Niño conditions and reduced during La Niña.