Can you tell me if I am reading this graph correctly...?

Question

Can you tell me if I am reading this graph correctly...?

Answer 1

As Jeff points out, 2005 was the warmest year on record. 1998 (not 1997) was an outlier year, due to a strong El Niño event.



You can see this with the NASA anomalies. Starting with 1994, you have anomalies of 23, 37, 28, 39, 56 (for 1998), 31, 32, 47, 55, etc.

http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts+dSST.txt



The 56 clearly does not belong.



The effect is even more pronounced with the satellite data. See how much of a spike the 1998 El Niño event was?

http://www.drroyspencer.com/wp-content/uploads/UAH_LT_1979_thru_June_10.gif



1998 was an outlier which was not caused by the usual forcing. It's like if the tide is rising by 8 inches / hour, and all of a sudden you get a 5 foot wave. The 5 foot wave was not caused by the 8 inch / hour rise in the tide. It would not be logical to claim that, since the sea level is not currently as high as the previous 5 foot wave, that the 8 inch / hour steady rise somehow stopped.

Answer 2

The graph you’re looking at is the GISTemp LOTI GTR (The Goddard Institute for Space Studies Land and Ocean Temperature Index Global Temperature Record).



The peak you’re using as the start point for your trend is the value for 1998, the graph clearly shows that one year exceeds this value and two years are very similar. Below is the same part of the graph expressed in numerals. Values are expressed as anomalies against the 1951-80 base period mean and the figures in parentheses show the ranking across the range.



1998 … 0.5600°C … (05)

1999 … 0.3125°C … (13)

2000 … 0.3250°C … (12)

2001 … 0.4717°C … (10)

2002 … 0.5517°C … (06)

2003 … 0.5417°C … (07)

2004 … 0.4758°C … (09)

2005 … 0.6167°C … (02)

2006 … 0.5308°C … (08)

2007 … 0.5692°C … (03)

2008 … 0.4375°C … (11)

2009 … 0.5658°C … (04)

2010 … 0.7233°C … (01)



So in reality, if we use this temperature record then 1998 ranks 5th, with the temperature for that year having been exceeded on 4 occasions in the last 6 years.



To make life easier I’ve produced a graph showing just the last 13 years, it’s on a larger scale than the NASA one so it’s easier to compare values from one year to the next. The graph uses the LOTI data from 01st January 1998 to 09th July 2010 and is shown as the red line expressed as an anomaly against the 1951-80 base period (ie exactly the same as the graph you linked to). The blue line is the trend expressed as a 2nd order polynomial.

http://www.flickr.com/photos/trevorandclaire/4779286897/sizes/l/in/photostream/



However, in the last question you asked you stated that “Even the most strident AGWer will have to acknowledge that a 13 year trend is not grounds for unequivocal proof of a climate change.” This begs the question why you are now using a 13 year trend and concluding “there appears to be a levelling out or even decline in temperatures”.



As you initially stated, and has been pointed out on countless occasions, 13 years is too short a period of time to deduce anything with respect to the climate as there are too many short-term variables which affect the temperature.



It’s for this reason that both the NASA graph and my graph use 30 year base-periods and express values as anomalies against this.



Using exactly the same data source but plotting a 30 year graph produces this…

http://www.flickr.com/photos/trevorandclaire/4779953718/sizes/l/in/photostream/



And doing the same thing but using the last 100 years of data produces this…

http://www.flickr.com/photos/trevorandclaire/4779964488/sizes/l/in/photostream/



Now, the whole premise of your argument is fundamentally flawed in that you’re trying to identify climatic changes by comparing annual temperatures to a value for a specific year. As I’ve mentioned, there are too many short term variables to be able to do this.



In your example the year you have chosen is 1998. This was the year of the strongest El Nino on record, this is an event that causes changes in the atmospheric conditions over the southern Pacific Ocean and leads to warming on a global scale.



The strength of the ENSO (El Nino Southern Oscillation) is measured on the Oceanic Nino Index (ONI). The 1998 episode was accompanied by ONI values that peaked at 2.5, there were 6 consecutive months where the ONI exceeded 2.0. To put this into perspective, the ONI has exceeded 2.0 on only 6 other occasions in the last 50 years so to get 6 consecutive months with such high values is something quite exceptional.



Current temperatures are being pushed above the norm due to the effects of the ENSO. In this instance the ONI peaked at 1.8 in December 2009, because ONI’s are calculated using rolling averages there is no value yet for June or July; the strength has been declining throughout the year and the likely current value will be around 0.2 or 0.3.



We can produce another set of figures showing how the global temperature has changed, this time by eliminating the effects of the ENSO and other short term influences such a volcanic eruptions and solar storms. The graph can be further enhanced by incorporating not just one temperature record (LOTI in the example you quoted), but all the temperature records.



LOTI shows a warming bias and the inclusion of other temperature records reduces the anomalous values.



The graph below shows the LOTI record in blue, the average of all temperature records in red and the effect on the temperature record by elimination (as best as possible) of the effects of the ENSO and other short term variables.

http://www.flickr.com/photos/trevorandclaire/4779993816/sizes/l/in/photostream/



In conclusion, when the data are correctly interpreted there is always going to be a clear upward trend in temperatures.



PS – You can check the accuracy of all the figures and graphs by downloading and analysing the data for yourself.

Answer 3

No...



Look closer. 2005 is the warmest year on record on that graph. That was an El Nino year. 2007 and 2009 are close seconds while 1998 is almost equal. 1998, not 1997, was the year of the strongest El Nino to date. As you can see it increased temperatures exponentially. However, as the trend increased, other years have exceeded those measurements. How could you even miss 2005? It's the big black highest dot past the 2000 line.

Answer 4

you would've said the same thing in 1990-94... and you would have been wrong then too.

Answer 5

No you are not reading it correctly.



I note that 1997 is hotter by far than any other year in the 110+ years preceeding it. I also see that at least one, if not two years, following 1997 are even hotter, and a third year looks so close that without a better graph I cannot tell which temperature was slightly higher.



I also note that even the coolest year after 1997 is hotter than the hottest year in the entire 110 year period previous to 1990.



I observe a continuous, unabated long term trend towards higher temperatures over the time frame measured.



I don't have any background in climate studies, so I don't know if this 100+ year trend of upward climbing temperature represents stochastic noise in a unchanging baseline, or represents a change in driving forces. If a change in driving forces, I do not know whether that driving force is "natural" or "manmade".



What I do know is that if that temperature continues to increase, for whatever reason, the changing climate will be rather "unpleasant" for most species, including humans. Given that I know that increased temperatures will lead to climate changes that are catastrophic to my preferred indolent lifestyle, I suggest we all think of ways to minimize said temperature increases.



I hear one idea is that we could start by attempting to minimize the levels of green house gases, such as carbon dioxide, that human activity produces. Slightly inconvenient. Perhaps useless. But better than just saying "oh well, nothing we can do."



Because you know - if my car goes over a cliff at 80 miles an hour - I'm still going to step on the brake as hard as I can. Even if it's not my fault the car went over the road, it was just a slippery road. Even if it doesn't actually stop the car, at least I tried to save my ***. Better than just saying "it's gravity, I can't do anything about it."

Answer 6

No. You are not reading the graph correctly.



You are using the graph to suggest that the warming tend is not real. The fallacy is obvious. If the reasoning is your own, you know what's wrong with it. If you've copycatted it, you've been had.

Answer 7

"Does this graph indicate that from a date in late 1990s, dare I say that landmark year 1997, that there has only been one year that average global temperatures have exceeded that year"



No. 2005 was significantly hotter than 1998, and 2009 was slightly hotter (although within the margin of error they were basically tied, as were several other years). Additionally, 2010 thus far has been significantly hotter than 1998.



"....and there appears to be a levelling out or even decline in temperatures?"



No, that is completely wrong. As Trevor shows in his excellent data analysis, the trend continues upward, even using 1998 as a starting point. Also explained here:



"the most recent [10-year trend] (1999-2008) equal to 0.19 ºC per decade"

http://www.realclimate.org/index.php/archives/2009/10/a-warming-pause/

Answer 8

<>



No, and if oiu were allowed to graduate from high school it indicates that something is very sick with our educational system.

Answer 9

The almost constant denier claim is cooling for the last 15 years, hard to fathom as even if you do go by 1998 as the warmest year that is only 12 years ago.

But as you yourself admit by GISS's figures 2005 is the warmest year, which is only 5 years ago. There is certainly no cooling noticeable in the climate data through the last decade (the warmest decade on record).

Denier claim to be skeptics yet want to ignore to glaringly obvious fact that 1998 is a very strong anomaly well above the increase in global temperature (you can see that in the graph you posted) caused by the strongest El Nino of the last century that occurred through 1997 and into 1998 (which are the last two years from the 90's in the 10 warmest years 97 & 98)

http://www.ncdc.noaa.gov/sotc/?report=global&year=2009&month=13&submitted=Get+Report#gtemp

http://www.pmel.noaa.gov/tao/elnino/faq.html#lanina99

What a real skeptic would ask themselves is why are global temperatures so close to the 1998 anomaly without the help of such a strong El Nino and in the presence of an active cooling PDO. As 2009 was the 5th warmest year on record and 2010 is, so far, even warmer. But there are no skeptics in denier ranks so they just continue to ignore the facts are make up stories about global conspiracies or when all else fails, rant about Al Gore.



Hardly surprising as you guys also constantly ask "why did it cool in the 40s and 60s" yet also ignore the PDO data for the period which easily shows why those cooling periods occurred.

http://en.wikipedia.org/wiki/File:Pdoindex_1900_present.png

Lesser PDOs also key in to the smaller cooling noted around 1999-2001 and 2008.

Answer 10

This is only my opinion.



It appears to me that on a small scale, you are correct in saying that there appears to be a leveling out. However the decline in temperatures is more difficult to say since in the large scale (particularly the second half of the century) it has been increasing over time.

Answer 11

What that graph says is that despite co2 emissions rising at a faster rate than predicted, temperatures have remained flat for the past 12 years.



If one looks at the CRU data, it shows a decline in temperatures since 2002 (Source Phil Jones)

Answer 12

Yeah, looks like about 2005. But I think the intent of the "5 year running mean" is to filter out some of the measurement noise.

Answer 13

i particular can, im a freshman taking a junior classification and that i take advantage of one too. ALL YOU DO IS press y= on the acceptable of your calculator... that is going to look like : y1= y2= y3= sort on your fist line in y1= then your 2d line in y2= then press GRAPH

Answer 14

You're right.



It's also a poster child for "how to lie with statistics."



While the graph looks like a huge difference, it represents less than 1 degree. Well within the margin of error for the time considered. It may look dramatic, but it's meaningless.