If you split the data set into 1917-1976 and 1977-2014 data sets (1976-77 being a notable discontinuity the often appears in Alaskan climate records that is commonly attributed to a shift in the PDO), then the long term trends are +0.77 days/century from 1917-1976 (breakup happening later over time), and -5.5 d/c from 1977 to 2014 (breakup happening earlier over time), with a ~-3 day (warming) discontinuity between the two trend lines at the 1976-77 PDO shift.
“it is the trends that are important for judging climate change”.
I would say that that is important for judging global warming.
If that trend is the result of wild fluctuations between very cold and very warm years, [ie where the residuals remaining after detrending are large] then its impact on humanity and wildlife is likely to be greater than if it is merely the result of the climate gradually warming with little variation around the trend [ie with small residuals].
That aspect of climate change, variability, may be a more dangerous aspect of Wally Broecker’s wild beast, at least for the next few decades, than the gradual warming trend.
The anomalously cold winter of 2013/14 in the Great Lakes region may be a one-off, or it may be an indication of greater variability to come.
In the same way you could claim that looking at the global average land/ocean surface temperatures for the last several decades that we have almost always been in a time of declining temperatures! – if you look at only a few years at a time, that is. Yet the overall increasing temperature is undeniable.
The question is: which is statistically significant, the short-term declines or the long-term increase?
On the use of river break-up dates as a climate change indicator, I’d think it’s more the result of a number of factors rather a particularly diagnostic one, like a trend in ice free dates of lakes and ponds. Rivers have more going on with rather short period volatility of flow rates producing substantial affects for the overlying ice crust.
For those with limited or no experience with river break-up, maximum flooding hazard occurs then. Not only is the flow volume very high, but the ice cover doesn’t just melt, rather it’s born downstream often forming a short lived dam of the entire stream. Usually happens quick. From growlers floating by in the channel to floating through the windows in half a day or so.
I don’t understand why you suggested I look at an entirely different graph.
The topic is this graph of the Nanana ice and I posed a query addressing the greater periods before and after the single and shorter period of decline.
The 50 years before 1970-1990 and the 24 years since show no trend while the latest 24 reflects the same pause observed in global warming.
In short, other than the 1970-1990 span my eyes do not see how Gavin concludes “the trends are very much towards a earlier spring.”
I can’t figure out why he dismisses a single years ice breakup date as an “anomaly” but even so there is still no detectable trend since around 1990.
Gavin says, “As we’ve often stressed, it is the trends that are important for judging climate change, not the individual years”
Then why can’t the can the trend over the last 24 years be more important than the previous 20?
[Response: The point is surely that we have 98 years, not 24. - gavin]
Am surprised to read, in contrast to the (warmer) early ice break-up on the Nenana River in Alaska, about “anomalous cold” leading to the still-there ice on the Great Lakes.
By this winter’s weather coverage, these were described as related – Arctic cold moving south into the US, leaving Alaska and the Arctic warmer. Maybe attributed to a polar vortex.
Comment by Same Ordinary Fool — 27 Apr 2014 @ 6:22 PM
I wonder what the TV weatherman will be saying over at WUWT this year? Probably maintaining a discreet silence. http://www.donotlink.com/g0j
I used that graph along with many other graphs, charts, tables, and information to enter the pool this year. I chose April 20, 21, 22, & 25th @ 12:25pm. That is just a 3H 23M difference. We’ll see if anyone else is as close or closer than I am and who knows, maybe I’ll take the cake!
In doing all of my research there are definite variances from one year to the next; however, the climate is definitely on a warmer path. Just look at Greenland. They are now able to prospect areas that have never been accessible before because of glaciers. Now, the once ice covered land is bare and exposed. It is not like a river or lake freeze where they can say “Oh, the weather was warm this year, maybe it will freeze next year”. Glaciers are moving, melting, and calving much faster than history has shown.
Yes, we need to look at trends in variability as well. One of the problems I see in this discussion is a kind of ‘tunnel vision’ as people get attached to and focus on one variable or the other.
It would be also be useful if we clarify, particularly in discussion with ‘skeptics’, what exactly is the question we are answering. Some effects are good indicators that “system energy is increasing”, while others, as you suggest, might be more in the category “these are the kinds of negative effects we are likely to experience.”
When we fail to make those distinctions, we allow ‘skeptics’ to be the ones framing the debate.
The village of Gelana is 230 mi/370 km as the raven flies WNW of the small town of Nenana where the break-up tripod is set on the Tanana River’s ice every winter. The Tanana joints the Yukon 140 mi/230 km above Gelana. Eagle’s ~85 souls are about 10 mi west of the Alaska-Yukon boarder and Circle’s 100 or so residents are another 110 mi/180 km further down river.
In general, the core regions of Alaska had an exceeding cool ending to winter then transitioned very rapidly to summer to conditions which it sustained at above norms in 2013. April at Nenana was more than 10 ºF below norms, and that location only had nights above freezing from May 23rd.
Can we ignore medium sized city (Fairbanks) sitting just 50 mi upriver? It certainly has grown during last century and who knows what its power and water heating facilities discharge into Tanana river. Wouldn’t global sea ice cover be a more robust proxy of climate change (or the lack of thereof)?
You don’t use “your eyes”, you do a statistical regression analysis – and in any regression analysis, more data points is always better for the reliability of any inferences. There will always be short term fluctuations – and 20 years is most decidedly “short term”.
Well, early yesterday I provided a fairly comprehensive and hopefully helpful response to #11 Edwardio, including links to some Tamino ‘Open Mind’ articles that dealt with understanding time series (eg “Don’t get Fooled Again”). The result: I was automatically informed that my post was regarded as spam and had been deleted. I emailed direct to Realclimate to ask for this to be remedied, but so far my post remains undisplayed.
I can appreciate that a site like Realclimate may get bombarded with all manner of dross, but if its systems and operators are unable to distinguish between dross and genuine posts, leading to the latter being deleted, then the site’s discussions becomes greatly degraded and hardly worth contributing to.
My eyes think the trend looks fairly level to maybe slightly increasing since roughly 1990. I would be suprised if the PDO did not have an impact on Alaska temperatures. The PDO has been in a cooling trend since roughly 1990, and I’d expect the trend to go more sharply down again whenever the PDO gets around to its next warm phase.
If this isn’t a sign of changing climate, than I don’t know what else it would take to prove it. It’s affecting the seasons and migratory patterns of animals which in turn effects the ecosystem. It throws the whole cycle out of balance. We need more scientific studies like this to get the point across to a wider audience. The recent climate problems addressed by the President and Congress last week is a great start. But we can do so much more to spread the word.
Comment by Alexis Crawford — 11 May 2014 @ 8:54 PM
I was the lead author on the Nenana paper in Science. Last year with the late breakup, Tucker Carlson’s “news” site was all over me trying to get me to comment – I sent them a very carefully worded explanation of why a single late year does not violate an overall trend. Rather than report it honestly, they just decided not to run the story at all.
Learned about the Ice Classic while doing an intertidal research trip in AK, got the records and wrote it up with Fio Micheli in about a week. But a sharp reviewer noticed an apparent mistake in our figure which was originally Julian Date of Ice Break up vs. Year. Our leap year data on the chart showed up differently than an ordinal table of earliest years of breakup. That led me on an intense investigation of how we record time (thanks to Duncan Steele’s excellent book, “Marking Time” for background). First I learned that Julian Date doesn’t mean days since January 1 as is often reported. More importantly, I learned that leap years and the Gregorian calendar actually lead to a bias in reporting phenological events because the first day of spring comes a little earlier each year of a century on our calendar – see my 2001 paper in Nature on this. That’s why the plot now is corrected for days since the vernal equinox in the year of record. There are even better ways to make this correction, like perhaps time since the sun was at a particular angle at a particular time at the location of the ice melt contest, but I haven’t calculated those.
[Response: Thanks for stopping by. The calendar issues are interesting, though it doesn't make a huge difference in cases where the trends are measured in multiple days/century. If you are still interested in these kinds of indices it would be fun to do a collation of other long time series of break up dates - I know there are a few long European records for instance. - gavin]