Yikes it’s
cold out there! This winter has brought record cold temperatures to Ithaca [1,2]
with continuing waves of Arctic air [3] making life pretty uncomfortable. All
the Finger Lakes are frozen except for Cayuga and Seneca [4] but with
temperatures expected to barely top freezing for the foreseeable future we may
yet see a frozen Cayuga Lake for the first time since 1978-1979 [2].
The past
several weeks are undeniably colder than normal, but what’s normal? Is this simply
cold weather or does it imply a cold climate? What can we say about how unusual
this winter is, and does it have anything to do with climate change? To answer
these questions, let’s conduct some rudimentary data exploration. This is the
first step when trying to understand scientific data.
Expanding the analysis done in a previous post, here are the 10-year average maximum, average, and minimum daily temperatures (˚F) for this December, January, and February (so far). The purple line marks the climatological long-term average:
Ithaca 10-Year Average
Temperature (˚F) from [5]
Looking at
the long-term climate we can expect winter temperatures to drop from 30˚F in
December to a bottom of around 20˚F in January and then a slight increase to
around 25˚F by the end of February. This 10-year running average matches this
climatological trend with some noise (a ten year average of weather is not yet
climate so we expect this noise).
Compared to
these 10-year averages, how does this year’s December-January-February compare?
Ithaca 2015 Daily Temperature
(˚F) from [5]
Yikes! The
average climatological January is 28˚F while this January averaged 17˚F. The
average climatological February is 26˚F with this February averaging only 11˚F.
This previous December, however, was slightly warm: the average December is
28˚F while this December averaged 32˚F. Taken as a whole, however, this is
indeed an unusually cold winter.
Besides the
cold temperatures, are there other aspects of this winter that are unusual? Is
there more or less variability? Do other years show similar anomalies, either
low or high? Here’s an animated comparison of the past 10 Ithaca winters:
Animation of Daily
Temperatures (˚F) for Ithaca from the Past 10 Years from [5]
Take a look
and see if you can find anything unusual. Check out this previous post for good
ways of examining these types of data.
It’s
difficult to look at the current weather and draw conclusions about climate, so
let’s look at these data from different perspectives to see what we can see. My
goal here is to tune your baloney detectors when being presented with weather and climate data and to do some data
exploration. Data itself can’t lie, but certain interpretations or
presentations of the underlying data can lie, especially if it is being
presented out of context. Always be skeptical when you’re presented with data!
If portions of data or methods are being hidden, there’s likely a hidden agenda
in the presentation.
For
example, here is the average February temperature for the past four years.
Based solely on this graph, what can you conclude about Ithaca winters?
Average February
Temperature (˚F) for Ithaca from 2012 – 2015 from [5]
I can hear
people shouting in the background: “Look! No such thing as global warming!” And
sure, based only on the average temperature of the past four Februaries,
completely removed from any larger time or regional context, that might make
sense. But what are we missing when we leave out the context? Let’s zoom out to
the average temperature of the past ten Februaries:
Average February
Temperature (˚F) for Ithaca from 2006 – 2015 from [5]
Looking at
ten years of data we can see that there isn’t much of a trend. The fact that
the past four Februaries line up in a nice straight decreasing trend seems to
be more of a coincidence than a statement about climate.
Let’s zoom
out a little more. What’s the past ten years worth of December-January-February
averages look like?
Average December-January-February
Temperature (˚F) for Ithaca from 2006 – 2015 from [5]
The unusual
February anomaly in Ithaca is much less apparent here. There appears to be a
slight decrease in the winter temperature in the past ten years, but it’s not
really a robust trend (or in more technical speak, it doesn’t seem to be
statistically significantly different from no trend at all). Let’s zoom out
some more. What do the previous 100 winters look like in Ithaca?
(NOTE: The data in in this figure is taken at a different
site than the data from the previous figure, thus the Ithaca winter
temperatures are not an exact match)
We
definitely get a different perspective here. The year 2012 was the warmest
winter of the decade, so it’s not an ideal place to start looking at
temperature trends. What we do see is some slow (over multiple decades)
increases and decreases, with a maximum in 1932, after which winter
temperatures generally decreased until 1978, after which they increased again
until around 2000, after which we see no real trend in winter temperatures. These
changes, however are small compared to the “noisiness” of the data.
This
100-year perspective puts the 10-year perspective into a broader climate
context. Among some slow variations in temperature from decade-do-decade we see
a lot of year-to-year variability. By looking only at 10 years of data we’re
missing a lot and we have to be careful about what we conclude from only 10
years of data if we’re talking about climate.
While we’re
looking at the previous 100 years of Ithaca winters, have you ever heard
someone older than you talking about how the weather or climate was clearly
different when they were young? Did you trust their interpretation and their
memory? Let’s take a look. Here’s the average winter temperature for each
decade during the past 100 years:
Average Ithaca Winter
Temperature (˚F) for Each Decade from 1900 – 2015 from [6]
If they
were born in the 1950’s, 1960’s, or 1970’s, then it was indeed colder when they
were young, although only by 3 – 4 ˚F. If they were born in the first half of
the century, then these current temperatures are close to what they remember
when they were young. Looking at the century as a whole, temperatures have
certainly fluctuated but there is no evidence of a clear, unambiguous trend.
What if
Ithaca is unique? What do the same data look like for the greater Northeastern
region? Here’s the December-January-February temperature trend for the entire
Northeast for the past 100 years:
Average Winter
Temperature Anomaly (˚F) for the Northeastern US Region from 1900 – 2011 from
[7]
Here we see
that there is a moderate trend showing increasing temperatures. Over the past
100 years, winters in the Northeast have warmed by around 2˚F, although there’s
a lot of variability in this data (i.e. the noise is large compared to the
signal). Ithaca’s trends don’t quite match this regional trend, but Ithaca is
just a single city in this broad region.
In Conclusion
This
broader climate context deepens our understanding of Ithaca’s unusual winter
weather we’re experiencing this year and provides us a long-term perspective.
Looking at the data like this is usually the first step in trying to understand
a scientific phenomenon since it helps to understand the larger picture.
Jumping right into a smaller portion of the data and drawing conclusions (e.g.
the previous four Februaries) leads to distortions and misunderstanding, and we
all need to be wary of data that is presented without this greater context.
[3] http://www.weather.com/storms/winter/news/another-cold-arctic-blast-great-plains-midwest-northeast
[4] http://ithacafingerlakes.com/2015/02/12/finger-lakes-ice-from-above/,
image from http://www.nasa.gov/content/snow-covered-northeastern-united-states/#.VO_DebPF_wH
