WUWT readers may recall: Upside-Side Down Mann and the “peerreviewedliterature” at Climate Audit. Steve McIntyre wrote then:
“…there isn’t a shred of doubt that Mann et al 2008 used these proxies upside down from the Tiljander interpretation. “
It seems the use of “upside down data interpretation” has leaked into a White House official report. WUWT reader “Jimmy” says in Tips and Notes: Check out the interesting temperature graph on this economic post from the White House today, “Deviation from Normal Temperature”.
Excerpt:
3. The first quarter of 2014 was marked by unusually severe winter weather, including record cold temperatures and snowstorms, which explains part of the difference in GDP growth relative to previous quarters. The left chart shows the quarterly deviation in heating degree days from its average for the same quarter over the previous five years. By this measure, the first quarter of 2014 was the third most unusually cold quarter over the last sixty years, behind only the first quarter of 1978 and the fourth quarter of 1976. In addition, there were four storms in the first quarter that rated on the Northeast Snowfall Impact Scale (NESIS). The right chart shows that no quarter going back to 1956 had more than three such storms.
Yes, while technically correct, showing heating degree days, it is upside down to the normal human interpretation of temperature, especially when the title says “Deviation from Normal Temperatures” while presenting degree days rather than a temperature plot. The other two largest positive spikes are the brutal winters of 1977 and 1978.
Source: http://www.whitehouse.gov/blog/2014/05/29/second-estimate-gdp-first-quarter-2014
UPDATE: here is how I would have presented this graph. Simply changing the title removes the inverted thinking about “Deviation from Normal Temperatures” and leaves it technically correct without unnecessarily confusing the reader.
Most people looking at that graph don’t have a clue what a heating or cooling degree day is. In case you don’t, here is a definition from NOAA/NWS
Q: What are degree days?
Heating engineers who wanted a way to relate each day’s temperatures to the demand for fuel to heat buildings developed the concept of heating degree days.
To calculate the heating degree days for a particular day, find the day’s average temperature by adding the day’s high and low temperatures and dividing by two. If the number is above 65, there are no heating degree days that day. If the number is less than 65, subtract it from 65 to find the number of heating degree days.
For example, if the day’s high temperature is 60 and the low is 40, the average temperature is 50 degrees. 65 minus 50 is 15 heating degree days.
Cooling degree days are also based on the day’s average minus 65. They relate the day’s temperature to the energy demands of air conditioning. For example, if the day’s high is 90 and the day’s low is 70, the day’s average is 80. 80 minus 65 is 15 cooling degree days.
Heating and cooling degree days can be used to relate how much more or less you might spend on heating or air conditioning if you move from one part of the country to another. Of course you’d have to take into account how well insulated your new home will be in comparison to your old one and the different costs of electricity, gas or heating oil. You could also use records of past heating degree days to see if the money you’ve spent on insulation, or a newer furnace or air conditioner is paying off. To do this, you’d also need records of past energy use.
The heating degree season begins July 1st and the cooling degree day season begins January 1st.
Source: http://www.erh.noaa.gov/cle/climate/info/degreedays.html
But also of interest in the same report is this graph and summary, which does make sense. It seems the winter of 2013/2014 set a new record for snowstorms.
4. Within the first quarter, several key indicators were lower in January and/or February before rebounding strongly in March, suggesting that the severe weather had a disruptive effect that only began to abate at the end of the quarter. Light vehicle sales, average weekly hours, core retail and food service sales, and core capital goods shipments dipped starting in December and/or January before bouncing back in March, and so were left little changed for the quarter as a whole. One outside group has estimated that the elevated snowfall in the first quarter slowed the annual rate of GDP growth by 1.4 percentage points, with all of that lost activity to be made up in the second quarter.
With this severe winter behind us, I have to wonder if any similar WH economic report (or any U.S. government report) exists that shows anything close to “slowed the annual rate of GDP growth by 1.4 percentage points” for a warmer than normal period. The summer of 2010 would be a good candidate for such a report.
If readers know of one, leave a note in comments.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
This was done to explain a downturn in economics….right?
Statistics Canada says the economy grew by 1.2% in the first quarter of this year, after posting a 2.7% expansion in the last three months of 2013. For the last month of the quarter, March, the economy edged ahead by 0.1 per cent over February.
Anthony,
This article was blindingly clear to me.
We see the Obama Adminsitration grasping at straws to explain away their economic failure.
Inadvertantly they are showing that the heating needs of Americans has increased over the last ten years or so, which is odd whn one think of how ‘global warming’ was impacting America so severely, according to Mr. Obama, during this same period of time.
D.J. Hawkins says:
May 30, 2014 at 8:50 am
“Second, the anomolies aren’t calculated against the entire record, they are calculated against 5-year bins. For Q1 for 1980, it’s calculated against the average for Q1 for 1976-1980.”
3. The first quarter of 2014 was marked by unusually severe winter weather, including record cold temperatures and snowstorms, which explains part of the difference in GDP growth relative to previous quarters. The left chart shows the quarterly deviation in heating degree days from its average for the same quarter over the previous five years. By this measure, the first quarter of 2014 was the third most unusually cold quarter over the last sixty years, behind only the first quarter of 1978 and the fourth quarter of 1976. In addition, there were four storms in the first quarter that rated on the Northeast Snowfall Impact Scale (NESIS). The right chart shows that no quarter going back to 1956 had more than three such storms.
OH! So the graph is totally misleading. Not only is it not “deviation from normal temperatures” it’s the short term anomaly of the temperature integral.
So they used an integral as a sort of low-pass filter , subtracted this (a difference being a high pass filter) from a 5 running mean of itself (a different distorting low-pass filter ) and then WTF ? Typical economists data processing.
What does the mangled data actually tell us? Does anybody know? I doubt it.
Now forgive me for not going to the “explanatory text” that explains that what the graph shows is not at all what the graph claims to show. Silly me.
I should add my thanks for pointing that out. It finally makes sense of this graph which shows neither “deviation from normal” nor “heating degree.days” but some mangled process data whose significance we can only guess at.
This processing reminds me somewhat of the MACD indicator explained here;
http://wattsupwiththat.com/2013/10/01/if-climate-data-were-a-stock-now-would-be-the-time-to-sell/
It seems to be some kind of short term variability indicator for detecting a change in direction. The choice of 5 years giving the scale of change considered.
This also explains the slight lag of a couple of years w.r.t the tornado data of the mid 70’s that I linked above.
The recent swing may well indicate we are moving into another period of tornado activity like the mid 70’s. Maybe those living in tornado alley need to know this. Climate change that MATTERS.
Excel data processing at it’s best LOL
Seriously the 2014 point looks like a BUY indicator for tornado insurance.
Jesse Ventura just lost my vote
http://youtu.be/99p55moV8kw
Now we know what it is about , Anthony’s new title would not be correct either since heating requirements are determined by ACTUAL temperatures not filtered,integrated short-deviations from a 5 year mean.
Anthony, what you make of MACD thing and it’s possible link to a up turn in US tornadoes?
Maybe the original series was called rednajit.
Perhaps something more basic could be added like “Furnace is running” and “Furnace is not running” along with cost per BTU average running through it. Show the impact to the pocket book.
“Second, the anomalies aren’t calculated against the entire record, they are calculated against 5-year bins. For Q1 for 1980, it’s calculated against the average for Q1 for 1976-1980.”
It escapes me to understand why the data is massacred in such a way? Why in the world did they just not present the degree days that are universally understood.
I would like to believe it is incompetence rather than some malicious intent.
As someone else noted the actual data on degree days would be an indicator of long term warming or cooling (assuming the data has not been manipulated). I assume that is available somewhere.
There seems to be a staggering amount of misunderstanding about the not that complicated concept of degree days. As the definition in the post amply illustrates degree days are an anomaly datum. 65 degrees is the baseline for that anomaly. It is a purely arbitrary number selected by those who created the concept and represents neither a mean, median, average or any other statistical concept. It could be any number that the creators thought made sense at the time with the only imperative for the integrity of the data set being that it stay the same from beginning to end.
Another difficulty which hasn’t been resolved is that the graph heading claims to show degree days for the “same quarter over the previous five years” when the graph seems to cover more than five decades. That error wasn’t addressed in the graph with the corrected title.
What is most amazing, as has been pointed out by others above, is that an administration which is one day out ballyhooing their Climate Assessment about how we we will all die agonizing deaths from Global Warming if we don’t bend over and spread ’em immediately, can, in the blink of an eye, declare that the profound failures of their economic policies were caused by too much cold weather. Calling upon Orwell is quite a cliche at this point in time but Big Brother Barry and his gang of thought police thugs have raised Doublespeak to levels that even old George never seemed to have envisioned.
Whoops! I had an Emily Litella moment in the previous comment! Nevermind!
The graph under discussion doesn’t seem to be a standard NOAA product. I’ve gone to their “Degree Day” page and can’t find it. I also can’t seem to find any data on the site prior to 1981. Anyone else know where to find the graph or the complete data it’s constructed from?
Speaking of being unfair to Mia Tiljander… did we spell her name right in the title? (sorry to pick nits)
D.J.Hawkins@1:29
http://www.eia.gov/totalenergy/data/annual/showtext.cfm?t=ptb0107
D.J.Hawkins@1:29
http://www.eia.gov/totalenergy/data/annual/showtext.cfm?t=ptb0107
ftp://ftp.cpc.ncep.noaa.gov/htdocs/products/analysis_monitoring/cdus/degree_days/archives/Heating%20degree%20Days/monthly%20states/
Sorry but I don’t buy the excuse that the authors were trying to be “real”. Come on you guys. You know the message: Anything up is bad, anything down is good. That is the message. So we need to be consistent and show how “up” things are bad, even if the “up” is really cold. So it looks like to me they tried to find a statistic that had that basic “consistent message”. Up is bad, down is good. So I still call it FUBAR on climate scientists!
@Brian says:
May 30, 2014 at 4:01 pm
You da man, Brian! Thanks.
Was on with Cavuto highlighting Weatherbell.com cold winter forecast.. that will pick your pocket and slap your face. Since Mid Jan 2013… the weather has had a cold look to it. Joe D and I already are out with the idea that a modiki enso event is on the way, not a super nino and that the two closest winters to analog the past 15 years are 02/03 and 09/10. Right or wrong, Co2 has nothing to do with it. Perhaps the central planners for the government should decentralize and get other opinions on what the winter will do besides widespread equal chances..or warm
To follow up on Anthony’s final question, no, I have never seen a warm quarter credited with a drop in GDP. Even hurricanes show up as (local) spikes, due to the economic activity involved in preparing beforehand and clearing up afterwards. That a high number of snowstorms in a single winter can reduce GDP by 1.4% shows how much more damaging cold weather is to the economy than warm weather.
The HDD and CDD concepts have been based on 65 F since their inception. However, better insulated homes and homes maintained at less than ~72 F during the winter would begin requiring heat at lower outside temperatures. Actual heating requirements are reduced by ~1% per degree F per 8 hour period when the temperature of the heated space is kept below 72 F, in a region with a winter average temperature of ~47 F (eg. Columbus, OH). For example, a home maintained at 66 F for 16 hours per day and using night setback to 55 F for 8 hours at night would have a heating energy consumption approximately 24% lower as the result of the lower normal set point (8 F lower for three 8 hour periods per day) and an additional 11% lower as the result of the night setback (an additional 11 F lower for one 8 hour period per day). The percentage savings would be higher in warmer regions, though the magnitude of the energy savings would be lower.
There is a lot of carping about this thread being about very little if anything, but I think Mr. Watts’ point is valid. I teach Mechanical Engineering at a university, and one of the courses I teach is a thermal/fluids laboratory. I demand that students write clear, brief and technically accurate reports. The fact that this graph was mislabeled as to the variable it plots would have been a fatal error, and I would have lowered the student’s grade over it. As a practicing scientist and engineer over many years I am certain I would have suffered some sort of penalty for having done something like this or less on my reports to clients and government agencies–I once had a proposal rejected for merely having not provided the units of a particular variable which allowed a cranky or possibly incompetent reviewer to supply his own units and then claim my work was wrong. Graphs, in particular, are supposed to make difficult material more clear not more difficult to comprehend.
Ed Reid says:
May 31, 2014 at 1:19 pm
You miss the point of this thread completely – whether deliberately or because you yourself failed to see that propaganda and salesmanship, I cannot say. A plot or graphic can be “accurate” but completely misleading at the same time. We have been told for seen years now that “The “message” of global warming catastrophe” has not been presented right” and “The way that scientists present the message” has to be “changed.”
This is salesmanship, gaming the impression, and propagandizing into t a willing and gullible mass media that itself desperately wants to make Obama’s message look good, viable, and uplifting.
Again, look at this temperature plot from the same government bureaucracies. What was the temperature before 1975? Where is that little blue line hiding? Where is “zero”? Why is “temperature” rising so suddenly and alarmingly recently?
http://www.ncdc.noaa.gov/sotc/service/global/global-land-ocean-mntp-anom/201001-201012.gif
And this plot. Again, look at the FIRST IMPRESSION that it makes.
http://wattsupwiththat.com/reference-pages/global-weather-climate/global-climate/#comment-1650125
Now, compare to this “NO RED BARS” plot of similar temperature data over the same period of time.
http://www.metoffice.gov.uk/hadobs/crutem4/data/web_figures/global_n+s_monthly.png
Different impression entirely, isn’t it?