Over the past few decades, I've been following the debate on global warming with an interest in the data and applied mathematics. As engineer with a professional background in RF engineering and research and development, all of which are math intensive, I feel the time has come to expose some of the false premises that I keep seeing crop up in alarmist claims regarding global temperatures being at an unprecidented upward trend.
Two of the most widely used methods for supporting such outlandish claims are: 1) Cherry Picking - selecting a date range that exaggerates one's point of view and 2) what period of time constitutes climate vs. weather. I intend to present why these two issues greatly cloud discussions on climate change and in fact, amount to rank dishonesty in much of the graphs and charts presented in favor of global warming.
The climate naturally warms and cools in response to several short term, multi-decadal, multi-millinial, and longer term cycles that coincide with periodic changes in our planet's oceanic thermohaline circulation, orbital variations and inclinations called Milankovitch cycles, solar activity cycles, and even cosmic ray galactic cycles. These cycles intertwine with each other to modulate the temperatures of our planet up and down on various scales. Specifically, the following cycles which affect earth's climate have, to date, been identified:
Solar Cycles:
·11 year Schwabe Cycle
·22 year Hale Cycle
·87 year Gleissberg Cycle
·176 year King Hale solar cycle[1]
·210 year Suess Cycle (aka. de Vries Cycle)[2]
·2,300 year Hallstatt cycle[3]
·6,000 year (as of yet unnamed) cycle[4]
Thermohaline Oceanic Cycles:
·65 year Atlantic Multidecadal Oscillation (AMO)[5]
·80 year Pacific Decadal Oscillation (PDO)[5]
·Variable short timescale El Niño-Southern Oscillation (ENSO)[5]
·Short timescale North Atlantic Oscillation (NAO)[5]
·1,400 year Bond Event Cycles[6].
Milankovitch Orbital Cycles:
·22,000 year procession (wobble)[7]
·41,000 year axial tilt[7]
·110,000 and 400,000 year eccentricity[7]
Cosmic Ray Cycles:
·Decadal cosmic ray fluctuations (following solar activity)[8]
·140 million year cycles traversing galaxy spiral arms.[8]
It is outside the scope of this blog post to discuss the science behind the above identified cycles other than to establish the fact that a broad body of scientific literature and paleoclimate research demonstrate a strong correlation between the earth's climate changes and regimes to these cycles. It suffices to say that our climate is cyclical.
It is important to understand that because the nature of climate change is cyclical, it opens the door for fraudulent statistical representations of climate trends. Benjamin Disraeli was quoted as saying "There are three types of lies - lies, damn lies, and statistics." It has also been said "If you torture data enough, it will confess to anything." I'll show you why most climate trend analysis presented in media and the blogosphere today is worthless and how the scale and extent of climate change is grossly misrepresented through statistical tricks.
Cherry Picking
In statistical terms, cherry picking occurs when one purposely chooses a start point and an end point in a data set that biases the results towards one's own preconceived conclusion. In cyclical data, this is actually quite easy to do in a way that actually looks quite convincing with seemingly truthful claims that the result is based solely on the data. Let me show you how it works visually. Keeping in mind that climate change follows cycles, we can neatly represent climate change as peaks and troughs (increasing and decreasing temperatures) on a sine wave. Consider the following graph of a pure sine wave:
Notice that when we use a measuring stick that is exactly one wave length (the length from one peak to the next peak of the wave) we get a zero trend line. In other words, there is no net positive or negative change in the slope of the trend line with respect to where it intercepts the wave. As long as we keep the measuring stick the same length, we can slide it in any direction and no matter where we position it, we will always get a zero trend line. The figure below demonstrates this point.
What happens when you shorten the measuring stick to, say, 3/4 of the wave length? The measuring stick is no longer suitable to be used to measure the wave. Depending on where you slide the measuring stick to, you will change the net trend measured. The next two figures demonstrate this rather strikingly:
Here, we are measuring from zero crossing to the peak of the wave. The gray line is the zero crossing. If we draw a red line from intercept point on one end of the green measuring stick to the intercept point on the other end, the result is a net trend that is positive.
And here we slid the measuring stick 1/4 wavelength to the left and you can see that it radically changed the net trend to negative.
Perhaps you're starting to see the problem. If you attempt to measure the net trend using a measuring stick that is either too short (or too long), the results can be completely opposite, depending on where you pick your starting point and ending point along the cycle. Unless you know the frequency of the cycle and use a measuring stick that coincides with the length of the cycle, almost any trend line you attempt to measure is going to be wrong. Herein lies the fallacy of most graphs depicting climate change - the measuring stick, or measurement period, is always way too short and positioned on the most recent positive slope of the climate cycle. Such graphs show us only where we are in a narrow range of a much longer cycle. They tell us nothing about net warming or cooling across the complete cycle. They're completely useless in telling us if the planet is warming or cooling on the larger climate cycle scale.
Weather is not Climate and Climate is not Temperature!
There is a raging debate in the blogosphere regarding what time scale differentiates weather from climate. Simply put, weather is the short term variability or noise, if you will, that when averaged across time, yields a smoothed trend referred to as climate. In other words, weather is the daily and monthly changes in temperature and other aspects of our environment. Climate trending is a way to try to look at averaged weather over longer periods. Most climatologists agree that the smallest unit of climate measurement is 30 years. Hang on to this number because I'm going to come back to later on.
The notion that our planet has an average temperature is completely baseless. The temperature of the planet has constantly changed across the geologic scale such that there is no discernible mean temperature - no number you can point to and say this is the baseline to which we should compare today's temperatures to. Geologists and paleoclimatologists both agree that there have been periods in the recent past where temperatures have been considerably higher and lower than today's present values. For example, the previous interglacial epoch was on average six to eight degrees hotter than today's temperatures. Likewise, nobody argues that the previous glacial period was much colder than it is today.
The best we can arrive at pinning down average world temperature is to say that planet earth is an icebox most of the time and we are fortunate to be living in a comparatively short interglacial period (the warm period between ice ages) where temperatures are much more to our liking. Viewed another way, our planet goes through one complete glacial cycle every approximately 140,000 years with the glacial period (ice age) lasting from 110,000 to 130,000 years and the interglacial period (the warm period we know today) lasting approximately 12,000 to 16,000 years. To put things into perspective, our present interglacial period, called the Holocene Epoch, started approximately 12,000 to 14,000 years ago. We're very near the end of our present interglacial period and on the doorstep to the next ice age if mother nature keeps on schedule as depicted below.
Image courtesy freedomkeys.com
Above, I drew in a horizontal blue / red bar into the graph. The blue depicts each glacial period (ice age) and the red depicts each interglacial period (warm age) across the 450,000 year timeline. As you can see, the warm period we presently live in is nearing its end. Also notice that the temperatures we've experienced in our current interglacial period are cooler than previous interglacial periods. You can also see a fairly typical error in the zero line drawn in this graph. It doesn't accurately represent the "average" temperature. If it did, it would be positioned much lower on the graph. It represents the "threshold between comfortable and uncomfortable climate as seen by humans."
Image courtesy University of Wisconsin
I believe the above graph, produced from the EPICA and Vostek ice cores better illustrates the timeline and temperature extents. If you look closely at our present and past interglacial periods, you can see temperature varies up and down on the shorter scale, as seen by short excursions up and down through the warm periods. These up and down cycles are driven mostly by the solar and thermohaline cycles I enumerated above. Lets zoom in on just our current Holocene interglacial period to get a better picture of these intermediate cycles:
Image courtesy climate.geologist-1011.net
Our interglacial temperatures follow roughly an 900 to 1200 year cycle. This overall cycle is formed by the additive and subtractive effect of the various solar and thermohaline cycles coming together. Now, to place things into proper perspective, notice at the very bottom right corner of the timeline, I drew in a purple and green horizontal bar. The green portion of the bar, which is barely even noticeable represents 30 years. The purple portion of the bar represents our entire instrument record - the period of time in which man has been measuring and recording temperatures. What do you notice about the temperature trend across the timeline covered by my bar? The climate is most definitely warming!
Allow me to draw your attention to a few other notable conclusions that can be made from this graph: The current trend in temperature rise is not, in any way, unprecedented. If you compare our current warming trend to that of the second peak of the Holocene Optimum, you can see that temperatures rose much more quickly and to a warmer extent than today. You'll notice also that today's temperatures have been rising, or recovering, from the latest Little Ice Age, which ended roughly 250-300 years ago. There is absolutely nothing unusual going on with respect to today's climate!
So bringing this discussion full circle, one can see that the 30 year measuring stick that is so often used in portraying global climate is entirely too short to say anything about climate at all. All it can tell us is if we are warming or cooling across the 30 year measurement period, nothing more. It is no surprise that the 30 year climate measuring stick says the climate is warming. We are on the positive swing of a roughly 900 year cycle. To get a more descriptive measure of today's climate, we need to use a much longer measuring stick, one approximately 900 years in length. But that measuring stick can only tell us how our temperatures compare to the previous intermediate cycle within the Holocene. It tells us nothing of how our temperatures compare in the longer glacial cycle. To do that, we need a measuring stick approximately 140,000 years in length.
The whole point here is simply this... Global warming alarmists emphatically state that climate should be measured across 30 years and for good reasons - cherry picking and climate trend glossing. Over the past 30 years, our climate has decidedly warmed. The warming is easily correlated to natural cycles and there is nothing about today's temperatures that are in any way unusual. However, the charts and graphs commonly used by alarmists make a rather striking and compelling but false case. To get an accurate and meaningful measure of how our climate changes, a 30 year measurement stick is absolutely wrong as it tells us nothing - period end of report. So the next time you're shown proof of global warming in the way of a graph or in reference to the ususal 30 year measurement period, keep in mind that what you are looking at is either intentionally cherry picked or conveniently ignores the greater climate picture.
I'll close here by presenting a blink animation of climate history that pulls all of the above discussion into visual context:
Animation courtesy of WattsUpWithThat.com
References:
1. R. K. Tiwari and K. N. N. Rao, Solar and tidal reverberations of deglaciation records from the tropical western Pacific: a clean spectral approach, GEOFIZIKA VOL. 16-17 1999-2000
2. Braun, H; Christl, M; Rahmstorf, S; Ganopolski, A; Mangini, A; Kubatzki, C; Roth, K; Kromer, B (10 November 2005). "Possible solar origin of the 1,470-year glacial climate cycle demonstrated in a coupled model". Nature 438 (7065): 208–11
3. "The Sun and Climate"; U.S. Geological Survey.
4. Xapsos, M. A.; Burke, E. A. (July 2009). "Evidence of 6 000-Year Periodicity in Reconstructed Sunspot Numbers". Solar Physics 257 (2): 363–9. doi:10.1007/s11207-009-9380-3
5. Knight, JR er al. “A signature of persistent natural thermocline circulation cycles in observed climate.” Geophysical Research Letters 32 (2005):L20708.
6. Gerard Bond, et al; A Pervasive Millennial-Scale Cycle in North Atlantic Holocene and Glacial Climates, Science 14 November 1997: Vol. 278. no. 5341, pp. 1257 - 1266
7. Milankovitch, M. 1920. Theorie Mathematique des Phenomenes Thermiques produits par la Radiation Solaire. Gauthier-Villars Paris.
8. Henrik Svensmark; Influence of Cosmic Rays on Earth's Climate, Solar-Terrestrial Physics Division. Danish Meteorological Institute, Lyngbyvej 30. DK-2100 Copenhagen, Denmark