Friday, June 20, 2008
You Can’t Get There From Here
OK. THIRD INSTALLMENT. Already, this is getting to be a chore, so I may cut this short—er, I mean—wrap this up, soon. This has been written in fragments and glued together like Frankenstein’s monster. I’ve tried to sand down the seams and sutures and cover them with literary Bond-O, but I’m sure I missed a few. Sorry for the bumpiness of the finish.
Last time, I posed what is for me the key question. When it is asserted that temperatures have risen x amount over t time, I want to know: How do we know that?
When the question first occurred to me, I started out with a probably provincial mental picture of the world of the Nineteenth Century (not to mention earlier times) as being pretty much benighted outside of Europe, North America, and the scattered bits of the British Empire. (Not so much, actually, though not as far off as you might think.) I thought that, although the thermometer was developed in the Middle Ages, it wasn’t until the middle of the 18th Century that we finally had what might be recognized as a modern thermometer, complete with a centigrade scale based on the freezing and boiling points of water.
Call it 1750, in round years. Coincidentally, also a convenient, round-numbers starting point for the Industrial Revolution.
Although all that was the case, I didn’t figure that you’d find decent temperature records of a long enough time-scale, with decent coverage, until late in the Victorian era, and even then there’d have to be significant gaps.
But I also figured I was probably underestimating the intelligence and creativity of our forebears. We moderns frequently do.
I had no idea.
("Moderns" we call ourselves. How amusing that conceit will doubtless seem in another hundred years.)
It turns out the earliest easily-accessible records are in the USHCN (United States Historical Climatological Network), and they go back to the early Nineteenth Century. New York has records dating to 1813, Ohio to 1826—just to give two early examples. But the earliest an apparent “full set” of national records is sometime in the last quarter of that century—about 1880 or so. The GHCN (Global Historical Climatological Network) starts in 1880. Since we’re addressing a phenomenon called Global Warming, I will take it that the actual record of observed temperatures starts then. There are earlier records. CO2 was first recognized as such in the 17th Century, and curious souls have been messing with it ever since.
And they’ve discovered some interesting facts. For example, while CO2 concentrations worldwide are pretty consistent, they do vary, sometimes by startling amounts. However, early assumptions about concentrations of trace gasses were made on the basis of remarkably few actual instrumental observations. A fact which gives this skeptic some pause.
There have also been attempts to extrapolate temperature patterns from objects called proxies, such as the isotope content of layers in ice cores, or the width of growth rings in trees. While the former may be perfectly accurate, the latter appears to be of most dubious value. In any case, I remain skeptical of proxies for many reasons not yet germane to this discussion, so we’ll set them aside and call 1880 the start date of the temperature record.
128 years.
Cue Charlton Heston: “Earth is four-and-a-half billion years old...”
128 years.
That’s point one, really. Life has been around on the planet for damned near four billion years. We have a pretty fair notion of what conditions have been for a long time past (higher and lower concentrations of carbon dioxide, higher and lower temperatures), and yet the warmistas are trying to tell us the putative current trend is unique. Unprecedented. Alarming.
And they’re basing this on a temperature record that’s 128 years long on the time scale.
However, I believe an examination of the spatial resolution of the array proves even more Disturbing.
Last time, I asked you to consider what temperature is. Now, I ask you to consider what it means. Why does a particular temperature reading—taken, say, at the nearest international airport—matter? Why does it matter to an individual member of the local population? Why does it matter to a traveler from another city arriving at said airport? Why does it matter to someone living a hundred miles downwind? A hundred miles upwind? Farther away in any direction?
Why does it matter to you, who live (perhaps) ten miles away as the crow flies and have your own thermometer in your back yard?
So what does temperature mean? What does a particular temperature mean?
Well.
In one sense, it means exactly what it says—“Right here, right now, the temperature is X.” And that’s all it means. It tells you nothing about where the temperature is even five feet away. It tells you nothing about the direction the temperature may be trending—up, down, or steady. It says nothing about what the temperature was yesterday or will be tomorrow.
Where you have looked, when you have looked, the temperature is / was / has been X.
Reading more than that into it seems a little foolish, doesn’t it? For example, would it be wise to assume the temperature five miles away, in a sheltered ravine, at a lower altitude, near a running stream, would also be X?
If you assumed a box one mile wide and tall and ten miles long between these two points, what do you think the conditions would be in the intervening volume—ten cubic miles of air—if you knew that the second point’s temperature at the time of our reading was ten degrees cooler than X, and that the cooler point was downwind?
And the atmosphere is 5-8 miles thick, so that space between your house and the airport is actually 50-80 cubic miles of air. Remember that energy flows from warmer to cooler, and that energy is what drives winds, which create movement in the atmosphere, and that pushes clouds and other gaseous substances around, mixing it all up, and creates weather. So what valuable, significant information is conveyed to you by a single average temperature for that 80mi3 box?
And so, in that volume of 12 billion cubic miles that is the earth’s atmosphere, how many samples would you want to take in order to be comfortable that you knew the average temperature of the whole at any given moment? Over a 24-hour period? Over a week? A month? A year? A century?
Don’t name a figure just yet. Just think about that and read on to find out what the scientists seem to think is adequate. See if you agree.
Here are a couple of descriptions of the data sets from which the data “supporting” global warming theory are taken. You can find these from links at CO2 Science. (Select from the “Data” menu.)
The U.S. Historical Climatological Network (USHCN) data set is a product of the Carbon Dioxide Information Analysis Center (CDIAC) and the National Climatic Data Center (NCDC). It is composed of monthly temperature and precipitation data from 1221 stations located within the conterminous United States. A plot showing the location of all 1221 stations can be found here. The period of record varies for each station. Most stations have at least 80 years of continuous data and extend through December of 2000. Stations with longer records extend back to the early to mid-1800s.
The Global Historical Climatology Network (GHCN-Monthly) data base contains historical temperature, precipitation, and pressure data for thousands of land stations worldwide. The period of record varies from station to station, with several thousand extending back to 1950 and several hundred being updated monthly via CLIMAT reports.
GHCN-Monthly contains mean temperature data for 7,280 stations (Figure 1) and maximum/minimum temperature data for 4,966 stations (Figure 2). All have at least 10 years of data. The archive also contains homogeneity-adjusted data for a subset of this network (5,206 mean temperature stations and 3,647 maximum/minimum temperature stations). The homogeneity-adjusted network is somewhat smaller because at least 20 years of data were required to compute reliable discontinuity adjustments and the homogeneity of some isolated stations could not be adequately assessed. Precipitation data are available for 20,590 stations (Figure 3) and sea level pressure data for 2,668 stations (Figure 4). In general, the best spatial coverage is evident in North America, Europe, Australia, and parts of Asia. Likewise, coverage in the Northern Hemisphere is better than the Southern Hemisphere.
(GHCN description found at NCDC site.)
The gridded data are a blend of the CRUTEM3 land-surface air temperature dataset and the HadSST2 sea-surface temperature dataset. As well as a best-estimate valuue for the surface temperature, a comprehensive set of uncertainty estimates are available. The image below shows near surface temperature anomalies for the most recent available month. ...the dataset starts in January 1850.
--HadCRUT3 dataset, UK Met Office
The MSU Satellite data set is a product of the NASA and the University of Alabama in Huntsville. In contrast to the GHCN and Jones et al. temperature data sets that represent temperatures at about 2 meters above the earth’s surface, the MSU data set represent the temperatures of a layer of the atmosphere that extends from the surface to approximately 8 kilometers (5 miles) above the surface. The data are obtained from microwave sounding units (MSUs) on the National Oceanic and Atmospheric Administration’s TIROS-N satellites, which relate the intensity or brightness of microwaves emitted by oxygen molecules in the atmosphere to temperature.
Spatial coverage of the MSU data set is nearly global, and data are available in a grid-box size of 2.5° of latitude by 2.5° of longitude, which represents the finest spatial scale of the four data sets available on our web site. Temporal coverage, however, is limited, as the MSU data set has only been in existance since 1979. The version of the MSU data set used here is D.
The data of each grid box of the original data base for which the center is located within the user-defined grid box is weighted based upon the cosine of its latitude. The weighted values are then summed to create mean values for the period of interest through which a regression line is fit.
The MSU data set is probably the most-accurate, finest resolution, and broadest coverage data set, but only extends back to 1979. The “here” referenced in the description is the CO2 Science site.
You may remember from our previous installment that a degree of longitude is roughly 69 miles at the equator and 49 miles at 45 degrees north or south. It just so happens that at 39 degrees north or south, one degree of longitude is roughly 54 miles. Which makes that 2.5°-square box, right here in River City, 135 miles on a side.
My home, in just that sheltered ravine, above the banks of the Ohio River, is a bit over 12 miles straight line from the international airport and within 50 feet or so of its elevation. (A relevant factor in this region. Elevations can vary by 500-600 feet in a few hundred yards ground distance.) It is not at all unusual for the thermometer in our back yard to register at a variance of as much as 10°F (generally cooler, but not always) with the “official” temperature reported from the airport. (Which, by the way, is not on the USCN.)
The airport is atop a windswept plateau, exposed to sunlight. I don’t know how the weather station is sited on the airport grounds. But its not too much of a stretch to imagine that the temperatures found there are unique for the same reasons the site uniquely meets the qualifications of an international jetport. Not that a reading of—say—50°F might not be found anywhere, but that, at a given time of day, the only temperature of our old friend X in the region might be found at the airport, while the rest of the region beyond that plateau shows temps of X + 5 (or minus, one).
Which would be fine, so long as the “average” derived from these measurements also included records for all those other myriad places in the earth. But the distance between my home and the airport is between one fourth and one fifth of one degree at this latitude (39°N). The best resolution data set—the MSU data—has 20,000 or so data points (360 degrees divided by 2.5 is 144. 144 squared is 20,736). The entire metropolitan Cincinnati area would fit inside that 2.5° x 2.5° box and rattle if shaken. In fact…
A few Saturdays ago, Indianapolis was hit with major thunderstorm activity. They had whamdidgious flooding. Tam and Roberta wrote about it—as did others from the area. That same weekend, we here in Cincinnati—roughly 100 miles away straightline and about 37 minutes of latitude (a bit over half a degree) south—were predicted to get something of the same, but we got a few showers and not much more.
Now, Indy is 100 miles off, and actually in a different weather system than Cincinnati. There’s a line that runs about 25-50 miles north of Cincinnati that defines an apparent boundary between Ohio River valley and Mid South systems and Great Lakes systems. Indy’s weather is driven a lot more than ours by what comes in off the northern plains. Ours is driven more by what comes off the Gulf of Mexico and out of the Southwest.
But, remember, that 2.5° square box at this latitude is a bit over 135 miles. So both metropolitan areas and all the intervening countryside could fit inside of it. With one. Average. Temperature. For. The. Whole. Shooting. Match.
More. If the box were 135 miles north-to-south, centered on a line about 17 minutes of latitude north of Cincinnati, it would reach 52 miles south of Cincinnati and roughly that far north of Indy. In otherwords, it would be a swath covering half of eastern Indiana, up I-75 as far as Lima and as far south as damned near to Lexington. Since the distance city-to-city is Fountain Square to the War Memorial, the leftover 17.5 miles at each end would reach easily into the next county on this end, and would likely encompass most of Marion County on the other end.
If you’ve ever watch a storm traveling on radar, you might get the idea of how silly it is to have a single temperature represent that entire area. (Map this grid onto your hood and see where it leads you.)
Now, obviously, it’s not that simple. But I think you can see that a grid with a one-mile, or even a ten-mile resolution might not be sufficiently accurate to get a good handle on the state of Earth’s atmosphere at any given moment.
But the whole global warming “crisis” (warming is probably happening, but certain advocates are trying to push it on the public as a crisis, demanding massive coercive government action to stave off) is founded in the temperature record from about 7,200 stations, some of which have been operating for some part or all of the period since 1880, but most of which have not, a little more than half of which have operated continuously for as much as half of that period, and no more than about 5,000 of which have ever been operating at one time. How in the world can that be considered an accurate and reasonable measurement of the temperature of the atmosphere?
And very few of which have any reasonable long timescale records over open oceans.
Not that there aren’t records of temperatures of ocean water. It’s just that the grid resolution is not very fine, the observation conditions are ... spotty and therefore of diminshed value.
Yet, the oceans cover 2/3 of the planet’s surface.
Yet, some of us have the hubris to imagine that they can or have determined a true global average temperature, that they can be certain it is rising, that the reason it’s rising is due to increased carbon dioxide in the atmosphere, that the reason that gas is growing in concentration is principally anthropogenic, and that you should give up your lifestyle, your prosperity, and your freedom in order that they can mount a fool’s quest to stop it.
But wait! There’s more!
Next time, I’ll detail for you a little more of what I’ve discovered about this whole question—about how the warmistas have attempted to increase the resolution of their data set with dubious statistical tricks, how they’ve fudged the data, and why, even if you accept that the extant record could be sufficiently well-resolved to tell us anything, you couldn’t trust the damned thing because of where the recording stations are.
Next time. Be sure to join us for, When You Get There, There You Are.
Cross-posted at BabyTrollBlog.
Part Two, Temperature, Global Warming, and How Far is a Degree?
Part One, Air Is Free
