By now you’ve probably read about Wind speed trends over the contiguous USA by Pryor et al. (2009, in press, JGR)
There is also an Associated Press story which cites this as a “first-of-its-kind study, and suggests that average and peak wind speeds have been noticeably slowing since 1973, especially in the Midwest and the East.”
Steve McIntyre of Climate Audit takes the study to task, not only for its data and conclusions, but for the Shenanigans of Michael Mann and Gavin Schmidt when they try to backtrack post facto after giving independent interview to AP’s Seth Borenstein and suddenly finding themselves in disagreement over whether “climate change” is involved or not. Go team.
I’ll point out that measuring wind accurately over a long period is not easy, particularly because surface anemometers tend to be problematic by the nature of their mechanical design. Aerovanes are better, but still have some of the same issues.
There’s also issues with land use change around the sensor long term, which I’ll get to in a moment. But first, the instrumentation.
For the traditional anemometer, here are some issues I’ve identified:
- Cup anemometer accuracy is typically +/- 5% when new some are even less. It depends on cup size and how linear the detector is. Some are very non-linear at low speeds.
- Anemometers, being mechanical, age, just like automobile engines, they are only good for so many revs of the engine before stuff wears down, affecting accuracy.
- The most common aging problem of anemometers is dust/dirt accumulation in the bearings, which tends to slow them down. I used to have a collection of dead anemometers from the former California State Climatologist…the vast majority had sticky bearings. I’d use them as a prop then give them away to classrooms when I went out to give talks to kids.
- Aging tends to affect low wind speeds more, by virtue of friction and by the fact that the lions share of wind measurements are at low speeds below 10mph. 40+ mph winds are not an everyday occurrence in many parts of the world. Easy to see in in plot of wind speed distribution. Of course there are exceptions.
USHCN Climate Station, New England, North Dakota. Photo by Eric Gamberg
So, depending on how good the instrumentation is to start with, and how well it is maintained, we might see the slowdown in wind caused partly by long term sensor issues.
If maintenance of NWS/NOAA wind equipment used in this study is anything like what we’ve seen for the USHCN network, I think the answer will be self evident.
But there’s another issue- urbanization and the associated land use change.
On the Climate Audit thread, one of the commenters pointed out the wind record from the Great Blue Hill Observatory near Boston.
For those who don’t know, according to commenter MarkB “the Great Blue Hill observatory web site. GBH, is just south of Boston, USA, and supposedly has the longest continuous meteorology record in the country (1885-present). The observatory is at the top of the hill – 635 ft. – with a large woodland park all around. Wind speed is certainly down since 1980.”
Our own Evan Jones recently visited the Blue Hill Observatory for a TV interview on the surfacestations.org project by Mish Michaels at WBZ-TV in Boston. He pronounced it a CRN 3 by the way.
Here is a look at the observatory USHCN station. Click for large images direct from the Blue Hill website
The wind instruments are mounted on the old observatory tower:
See the graph of wind speed from the observatory below:
The graph brings some issues about urbanization to mind. The Boston area has been highly urbanized.
Urbanization could be roughing up the boundary layer.
Clearings, buildings, shopping malls, subdivisions, etc. create more friction and turbulence in lower level winds than say, open fields or vast swaths of forest top. More turbulence tends to inhibit smooth flow of wind, this decreasing the wind speed.
Granted Blue Hill is a bit higher than surrounding surface, but I think the the effect of increased turbulence and drag in the highly built up northeast will make an impact even at higher levels.
see this report by Zhang et al on urbanization impacts in China:
http://www.lapc.ac.cn/UpLoadFiles/File/2008meeting/dahui/zhangning.pdf
Urbanization leads to “heat island”, “dry island” and wind speed loss over urban area …
See the wind speed plots comparing urban/non urban on page 11
The wind speed slowdown due to urbanization has been known for quite some time. Here’s an older book on the subject:
page 1267:
Yet, the surface roughness of the city serves to reduce wind speeds…
and others:
http://ieeexplore.ieee.org/iel5/4534879/4534880/04535390.pdf?arnumber=4535390
Due to the friction and drag of buildings, there are wind speed losses over urban area in all cases (Fig.4). In summer, the reduce of wind speed can be 1.0 …
It seems then, since Blue Hill Observatory is in the middle of a large area of growth for the past 50 years, that some of this wind speed reduction related to land use change and urbanization would apply?
Have a look at this Google Earth 3D image showing the patchwork quilt of the terrain around the observatory. Note the may holes in the forest canopy around the observatory and the signs of ubanization. How much of that affects wind speed measured on the hill? How much of it was there 30, 40, 50 years ago?
I’d also like to find out when they last replaced/recalibrated their wind sensors. It is one of the sensors where detecting a problem is not always obvious, especially when they clog up with dust in the bearings.
When studies cite surface data from weather stations, it is always a good idea to look at land use/land change around the stations as well as the station instrumentation. Chances are you’ll find issues that may not have been considered.
How about heated wire anemometers? I’ve been researching these as I plan on building one to be able to monitor wind speed in my yard. Worth doing or should I be looking for something more conventional?
REPLY: They don’t hold up in weather very well. – Mostly lab use. – Anthony
I read the AP article on this a couple days ago and was instantly struck by the passage in which Gavin Schmidt exressed skepticism of the results, but only on the basis that the computer models showed no reduction in wind speed. That’s like being skeptical of a report of a quarter inch of water in a rain gauge on the basis that the computer models didn’t forecast any rain. It says a lot when a climate scientist’s knee-jerk reaction is to accord more weight to a virtual simulation of a system than the actual measurements taken of that system.
Granted, trying to confidently identify a real trend in a climate variable is difficult given the uncertainties in the instrumentation, not to mention the chaotic nature of the system itself; hence any study like this should be scrutinized carefully. But it makes little sense to me to have more confidence in a model of a system than you do the real-world measurements of that system. If the real-world measurements of climate are so fraught with uncertainty as to be dismissed that readily, then it stands to reason that the climate models built on those measurements have to be accorded even less certainty, not more.
In other words, if we don’t have the ability to even measure changes in wind speed with any confidence in the real world, how is it that we’ve gained the knowledge with which to imbue in a computer the more-difficult ability to calculate the specific transient response of wind (no change) to a particular input (radiative forcing). It’s not as if the compuer models were handed down to us by some technologically-advanced aliens. The computer model is only as good as the programming that went into it, and the programming is only as good as the ability to measure the actual system.
Now, in all fairness, the AP could have truncated his criticism or taken the comment out of context, but if it did accurately reflect his view of the study, then I really question this guy’s objectivity. What I think is happening is that many climate scientists are accepting the results of computer models as substitutes for real data, not because there is a reasoned basis for doing so, but because the results are clean and unambiguous whereas trying to determine the inner workings of the climate through the more rigorous scientific practice of taking measurements is messy and uncertain. When a computer model says “x” input produces “y” effect in 95% of the trial runs while holding other inputs constant, then the results are black and white. When measurements merely show “y” effect (e.g. a temperature trend) in an uncontrolled system then attributing causation to any particular variable is a very difficult thing. But just because a climate model is more convenient doees not mean that it’s accurate, and is surely isn’t a basis for questioning real data.
Kurt:
AP did not quote out of context or truncate.
See this post:
http://www.climateaudit.org/?p=6244#comment-345845
Or view the relevant thread at RealClimate.org
I will grant you the theory says cube, but from limited experience with small remote power generation, I haven’t seen anywhere near 8x the output for a doubling in wind speed. So I suspect something else is at work. Theory and practice are not always the same.
In John Gault’s post, the article he quotes says: “even in the best locations, the wind will fail at least 5 percent of the time.”
Maybe at 20,000 feet. A bit more at the surface.
“just because a climate model is more convenient doees not mean that it’s accurate”
This is the root of the problem, the GCMs require tractable solutions to a huge problem involving chaotic elements. It would be fine, for advancing knowledge, but betting trillions on what are at best, preliminary findings, and at worse, complete fabrications is going way beyond what the “science” will support. We are at the state in climate science that Lamarck was at in genetics.
“”” Retired Engineer (07:17:59) :
I will grant you the theory says cube, but from limited experience with small remote power generation, I haven’t seen anywhere near 8x the output for a doubling in wind speed. So I suspect something else is at work. Theory and practice are not always the same. “””
Well I have always understood the question to be one of “available” power, which most definitely goes as the cube of wind speed.
That does not mean that a fixed design will actually yield 8 times the power for double the wind speed.
For a start, existing wind turbines that I am familiar with are ALL fixed RPM designs; that being necessary to phase lock the alternators to the grid line frequency.
That means that wind speed changes would have to be accommodated by rotor blade adjustment of angle of attack, to extract greater lift and torque from a fixed airfoil section. And I would be very surprised if high efficiency can be obtained over an 8x lead range at fixed RPM.
In fact I believe that simply fixing the operating RPM takes away one of the doublings, so the available power is now only 4x for a doubled wind speed.
It would seem that to realize the additional power you have to let the rotor free run, and use a torque converter to drive and phase synch the alternator; and I wouldn’t want to have to design that torque converter, and keep it running; not to mention the risks of having the rotor free to run at a variable rpm.
George
George E. Smith (10:39:39) :
Hi George,
From what I read, new turbine/generator designs in Europe don’t use fixed rotor speeds for phase lock. Instead they use an electronically rotated field in the generators. This allows them to vary the rotor speed and still phase lock. Basically an electronic torque converter.
I work with this data for a living, and I can tell you that NOAA/NCDC data have serious problems, especially at low wind speeds. The conversion to ASOS practically eliminated reporting of wind speeds less than 3 knots (rounding these down to 0). Even higher wind speeds often report as variable direction (often rounded to zero wind speed), and “calm” conditions are exceedingly frequent at many sites where only light breezes commonly occur.
Maintenance on the wind sensors, especially in low-wind areas and away from NWS offices and climate stations, is if anything even worse than that for temperature data, which frequently deviates noticeably from reality undetected for years on end. Unfortunately, wind speeds appear highly dependent on station siting and exposure. Moreover, high-wind events generally occur on localized spatial scales (and also frequently knock out power to the ASOS during extreme events, rendering the system entirely useless). The local nature of high-wind events inhibits detection of erroneous instruments, especially in areas generally unserved by other networks.
Another factor that you didn’t mention: trees. The more numerous and taller they get, even some distance from the site, the more effectively they block and obstruct wind flow. I wouldn’t be surprised if tree growth alone accounted for a significant proportion of the apparent slowing of the wind at Blue Hill, Massachusetts–a site chosen apparently for its great exposure that trees now perhaps partially block. Trees unfortunately tend to increase in height gradually on their own, so detection, adjustment, and remediation aren’t easy.
I mentioned above that some people are not fond of ASOS weather stations. If we are to believe the one on Martha’s Vinyard, an island off the coast of Massachusetts, then http://adds.aviationweather.gov/metars/index.php for KMVY includes this morning’s report:
METAR text: KMVY 170808Z AUTO 00000KT 3/4SM R24/P6000FT -SN BR SCT001 SCT007 03/02 A3025 RMK AO2 P0000 TSNO
Conditions at: KMVY (MARTHAS VINEYARD, MA, US) observed 0808 UTC 17 June 2009
Temperature: 3.0°C (37°F)
Dewpoint: 2.0°C (36°F) [RH = 93%]
Pressure (altimeter): 30.25 inches Hg (1024.5 mb)
Winds: calm
Visibility: 0.75 miles (1.21 km)
Ceiling: at least 12,000 feet AGL
Clouds: scattered clouds at 100 feet AGL
scattered clouds at 700 feet AGL
Weather: -SN BR (light snow, mist)
Snow! Well, maybe not. It’s certainly cold enough for snow, but there was no precip in the forecast. What most likely happened is that site is well known for extreme radiational cooling despite being surrounded by ocean. The snow was probably just fog, that’s something ASOS has a little trouble telling apart.
Sometimes a pair of operational human eyeballs trumps a non-sleeping semi-blind heap of electronics.
It is believable that KMVY did get into the 30s, other stations nearby including the NWS office in Taunton reported temperatures in the 30s. Blue Hill, well above the inversion, got down to 50F.
See also my source
http://www.jotsweather.com/JotsWeather/JotsArticles/Entries/2009/6/17_Snow_at_Martha%E2%80%99s_Vineyard_this_morning.html
The power from the sky plans show up every few years. The first question to ask is
always, how much does the tether weigh? I can imagine improvements in materials
that will back it possible to put such a power plant in the sky. The next question to
ask is how big a space must be cleared in case the tether breaks?
“”” KLA (16:58:43) :
George E. Smith (10:39:39) :
Hi George,
From what I read, new turbine/generator designs in Europe don’t use fixed rotor speeds for phase lock. Instead they use an electronically rotated field in the generators. This allows them to vary the rotor speed and still phase lock. Basically an electronic torque converter. “””
Hey Thanks KLA, I guess that makes a lot of sense. Trying to phase lock a mechanical contraption that all kinds of rotational moment of inertia; given that it is being powered by a chaotic driving force; with gusts and holes and all kinds of sudden impulses; I can see why they shake themselves to death. They already have a sychrounous vibration mode due to the vertical wind shear so that all three blades are not producing the same thrust at any time so the blades tend to flutter with the lift variation each revolution.
Do you happen to know if Hydroelectric Turbines use similar techniques. I took a look at those monster Siemens alternators at Grand Coulee up in Washington; and I though about some water shock wave being sent up the tunnel if the got some sort of load dump problem; that’s a lot of inertia to have intermittent loading on.
At least with hydros, the dam and lake let them supply a very constant water drive; but how do you low pass filter the wind to a big fan blade.
George
Hi,
We have just added your latest post “A few thoughts on the “climate change lowers wind speeds” study ” to our Directory of Environment . You can check the inclusion of the post here . We are delighted to invite you to submit all your future posts to the directory for getting a huge base of visitors to your website and gaining a valuable backlink to your site.
Warm Regards
greenatmos.com Team
http://www.greenatmos.com