Home/BlogUPDATE: I’ve appended the results for the U.S. only, which shows evidence that CRUTem3 has overstated U.S. warming trends during 1973-2011 by at least 50%.
The supposed gold standard in surface temperature data is that produced by Univ. of East Anglia, the so-called CRUTem3 dataset. There has always been a lingering suspicion among skeptics that some portion of this IPCC official temperature record contains some level of residual spurious warming due to the urban heat island effect. Several published papers over the years have supported that suspicion.
The Urban Heat Island (UHI) effect is familiar to most people: towns and cities are typically warmer than surrounding rural areas due to the replacement of natural vegetation with manmade structures. If that effect increases over time at thermometer sites, there will be a spurious warming component to regional or global temperature trends computed from the data.
Here I will show based upon unadjusted International Surface Hourly (ISH) data archived at NCDC that the warming trend over the Northern Hemisphere, where virtually all of the thermometer data exist, is a function of population density at the thermometer site.
Depending upon how low in population density one extends the results, the level of spurious warming in the CRUTem3 dataset ranges from 14% to 30% when 3 population density classes are considered, and even 60% with 5 population classes.
DATA & METHOD
Analysis of the raw station data is not for the faint of heart. For the period 1973 through 2011, there are hundreds of thousands of data files in the NCDC ISH archive, each file representing one station of data from one year. The data volume is many gigabytes.
From these files I computed daily average temperatures at each station which had records extending back at least to 1973, the year of a large increase in the number of global stations included in the ISH database. The daily average temperature was computed from the 4 standard synoptic times (00, 06, 12, 18 UTC) which are the most commonly reported times from stations around the world.
At least 20 days of complete data were required for a monthly average temperature to be computed, and the 1973-2011 period of record had to be at least 80% complete for a station to be included in the analysis.
I then stratified the stations based upon the 2000 census population density at each station; the population dataset I used has a spatial resolution of 1 km.
I then accepted all 5×5 deg lat/lon grid boxes (the same ones that Phil Jones uses in constructing the CRUTem3 dataset) which had all of the following present: a CRUTem3 temperature, and at least 1 station from each of 3 population classes, with class boundaries at 0, 15, 500, and 30,000 persons per sq. km.
By requiring all three population classes to be present for grids to be used in the analysis, we get the best ‘apples-to-apples’ comparison between stations of different population densities. The downside is that there is less geographic coverage than that provided in the Jones dataset, since relatively few grids meet such a requirement.
But the intent here is not to get a best estimate of temperature trends for the 1973-2011 period; it is instead to get an estimate of the level of spurious warming in the CRUTem3 dataset. The resulting number of 5×5 deg grids with stations from all three population classes averaged around 100 per month during 1973 through 2011.
RESULTS
The results are shown in the following figure, which indicates that the lower the population density surrounding a temperature station, the lower the average linear warming trend for the 1973-2011 period. Note that the CRUTem3 trend is a little higher than simply averaging all of the accepted ISH stations together, but not as high as when only the highest population stations were used.
The CRUTem3 and lowest population density temperature anomaly time series which go into computing these trends are shown in the next plot, along with polynomial fits to the data:
Again, the above plot is not meant to necessarily be estimates for the entire Northern Hemispheric land area, but only those 5×5 deg grids where there are temperature reporting stations representing all three population classes.
The difference between these two temperature traces is shown next:
From this last plot, we see in recent years there appears to be a growing bias in the CRUTem3 temperatures versus the temperatures from the lowest population class.
The CRUTem3 temperature linear trend is about 15% warmer than the lowest population class temperature trend. But if we extrapolate the results in the first plot above to near-zero population density (0.1 persons per sq. km), we get a 30% overestimate of temperature trends from CRUTem3.
If I increase the number of population classes from 3 to 5, the CRUTem3 trend is overestimated by 60% at 0.1 persons per sq. km, but the number of grids which have stations representing all 5 population classes averages only 10 to 15 per month, instead of 100 per month. So, I suspect those results are less reliable.
I find the above results to be quite compelling evidence for what Anthony Watts, Pat Michaels, Ross McKitrick, et al., have been emphasizing for years: that poor thermometer siting has likely led to spurious warming trends, which has then inflated the official IPCC estimates of warming. These results are roughly consistent with the McKitrick and Michaels (2007) study which suggested as much as 50% of the reported surface warming since 1980 could be spurious.
I would love to write this work up and submit it for publication, but I am growing weary of the IPCC gatekeepers killing my papers; the more damaging any conclusions are to the IPCC narrative, the less likely they are to be published. That’s the world we live in.
UPDATE:
I’ve computed results for just the United States, and these are a little more specific. The ISH stations were once again stratified by local population density. Temperature trends were computed for each station individually, and the upper and lower 5% trend ‘outliers’ in each of the 3 population classes were excluded from the analysis. For each population class, I also computed the ‘official’ CRUTem3 trends, and averaged those just like I averaged the ISH station data.
The results in the following plot show that for the 87 stations in the lowest population class, the average CRUTem3 temperature trend was 57% warmer than the trend computed from the ISH station data.
These are apples-to-apples comparisons…for each station trend included in the averaging for each population class, a corresponding, nearest-neighbor CRUTem3 trend was also included in the averaging for that population class.
How can one explain such results, other than to conclude that there is spurious warming in the CRUTem3 dataset? I already see in the comments, below, that there are a few attempts to divert attention from this central issue. I would like to hear an alternative explanation for such results.
added 15:07 CDT: BTW, the lowest population class results come from approx. 4 million temperature measurements.
What would such a revised surface temperature trend mean for the studies that have examined the difference between the ground trend and the lower troposphere trend as measured by satellites? Lindzen, for example, has emphasized that this difference in trend ought to be present as a matter of basic physics, hinting that the difficulty of finding the predicted difference may be due to errors in the measured ground temperatures. If the revised ground trends resulted in a comparison with the satellite data that is more consistent with the physical theory, might that further bolster the case you are making?
On another issue, Jonathan Lowe (http://gustofhotair.blogspot.com/) has examined Australian data on temperatures taken at fixed hours of each day. He has shown there has not been a trend increase in temperatures at 9pm, midnight, or 3am even though the minimum has trended up over time. He argues that the minimum occurs right at sunrise, so it is not a good measure of what is happening at night. More generally, he argues that the trends in fixed time of day temperatures are consistent with the sun playing a dominant role in recent temperature changes. I wonder if one could do a similar analysis with this fixed time of observation ISH data.
Most of the “missing hotspot” is in the tropics, which is mostly ocean, which has a different temperature measurement source. So at this point I don’t believe these results say anything about the hotspot controversy.
Don’t give up on publishing, Roy. Here are a few suggestions for getting this in shape for publication.
1. Acknowledge the fact that the people who do surface temperature aggregations correct local temperature trends for the urban heat island effect.
2. Do the same thing yourself, and see whether your bias still exists.
3. Try to estimate HOW MUCH any remaining bias would affect the trends. You say, “These results are roughly consistent with the McKitrick and Michaels (2007) study which suggested as much as 50% of the reported surface warming since 1980 could be spurious.” But what do you mean by “roughly”? 50% is a lot.
1. I already implied they had been somewhat adjusted…what I’m interested in is how much has remained unadjusted.
2,3. Steve Mosher claims it has already been well established that the CRUTem3 dataset has spurious warming, so maybe Phil Jones will make the necessary corrections for the next IPCC report. 😉
Hi Roy,
You say you “already implied they had been somewhat adjusted”. So, is that all the adjustment they do before computing global averages and such? I don’t know much about CRUTem.
You could also address the criticisms of McKitrick and Michaels, such as those noted here:
http://www.realclimate.org/index.php/archives/2007/12/are-temperature-trends-affected-by-economic-activity-ii/
Barry, McKitrick already addressed qualitative, hand-waving criticisms of his work in 2010 with a rather thorough statistical analysis. My results stand on their own.
Thanks. Can you give a link?
Roy,
If you didn’t see our poster at the AGU this year, you might be interested. We found a small but (apparently) significant UHI effect using GHCN-Daily data across a slew of different urbanity proxies (Modis, Nightlights, ISA, Population Density, and Historical Pop Growth). The AGU ePoster site seems to be down at the moment, but you can find our poster here: http://www.2shared.com/document/ciJ9qH1t/Hausfather_et_al_2011_AGU_Post.html
I should note that I’m not very comfortable with your choice to fit a linear projection of UHI to the log of population, especially when the three points you are fitting are not particularly linear. Its also worth pointing out that ISH data has very little quality control or correction for inhomogenities, and it would be interesting replicating your approach with GHCN-Daily or the Berkeley monthly data.
The problem with looking at the NH only especially when using population density as a proxy for UHI is that we know that the UHI effect has a statistically significant latitudinal bias. As Oke and others showed population density ( in conjunction with regional wind speed ) is indicitive of MAX UHI and further that the relationship is different for the NH than it is for the SH. basically, it’s lower in the SH.
That said, as Zeke mentioned, we did find a small bias in the record (~.04C) per decade. For that we used a more robust proxy for urbanity than population density.
As we add more SH stations the Bias diminishes. This is consistent with other literature indicates that UHI is modulated by latitude. The number of grids you have by the coast will also impact your results. In regressions of stations with long and complete data coverage I found the key independent variables were
1. latitude ( NH warms faster)
2. distance from coast ( inland warms faster)
3. amount of built urban area ( more land transformation = more UHI)
What that means is to estimate the bias in the TOTAL record your analysis has to include representative samples of both coastal locations and SH & NH samples.
Finally, the 20 day test you use for inclusion of data will cause you problems. In tests of all stations ( daily data ) I’ve found that as you relax the constraint from 100% of data coverage to less than 100% you introduce a cooling bias in trends. This is likely due to the fact that missing data is not uniform with respect to seasonality. have uniform data coverage across seasons is also critical for UHI studies since the effect has a strong seasonal component. That has been shown in the literature repeatedly
Yes, I know that the warming increases with latitude, our satellite dataset has shown that for close to 20 years now. I could slice and dice the data in all different ways, but I wanted an overall feeling for whether there is a residual warming bias in CRUTem3.
But, if it has already been well established (as you seem to suggest) that the CRUTem3 dataset has a significant residual warming bias, can we assume that the bias will be corrected for the next IPCC report? If so, I won’t waste any more time on this.
Also, if the population data I used is so bad, why does does the rate of warming change with population? Are you saying a better population dataset should give even stronger results?
Perhaps there is an underlying trend in population density with latitude(?), especially in the NH.
The reason UHI is modulated by latitude is because (as I explained elsewhere in this thread) the modus operandi is less evaporation and convection in urban areas. As you move towards higher latitudes there is less evaporation occuring and hence less opportunity to reduce it through impervious ground cover and subsequent high volume low surface area removal of rainwater.
See how easy it is to produce elegant explanations of the observations once you have the correct understanding of the physical processes underlying them? Funny how it works that way. 😉
That said, there is an offsetting principle. Due to non-linear nature of temperature increase with increasing power (temperatures rises as the fourth root of power) it takes less energy to raise a low temperature by one degree than a high temperature by one degree. Another commenter in this thread noted that population centers produce about 5W/m2 of waste heat that must be added in addition to solar power. As solar power diminishes with increasing latitude this 5W/m2 becomes a higher percentage of total power as well as having a lower base temperature to act upon. One might also presume that more power is consumed by higher latitude urban areas than lower latitude locations because of winter heating requirements. These would tend to offset the difference caused by impervious ground cover replacing deep rooted plants.
The classification of station sites by population around year 2000 is not, in my view, an adequate tool to measure the UHI regarding warming (as opposed to warmth). More populated sites ordinarily mean higher temperatures, granted, but that does not mean higher temperature TRENDS, unless the density of population (and related structures and equipment such as buildings, factories or cars) is INCREASING. Of course, most cities grow in population (through extensions of the urbanized area and/or increases in density per unit of area), but possibly some sites increase or decrease their density at different rates, and that CHANGE in density would be the driver of a change in TRENDS, not the absolute level of the density.
Population density, on the other hand, is not a surefire indicator of local heating: population may diminish while heating-causing structures and equipment increase, e.g. when people move to suburbs while downtown or industrial areas become mostly occupied by offices or factories. Probably fewer people per sq km live now in many downtown or industrial areas than 40 years ago, but that does not mean that vehicle circulation, density of buildings, % area under concrete, or density of engines or power (HP or MW/h per sq km) has diminished in those areas.
By the same token, the survey of US met stations led by Anthony Watts a couple of years ago revealed that many “rural” stations (probably on green surroundings some years before) were now on top of a tin or concrete roof, or in a paved parking lot, or near an airstrip or highway.
In other words, population density is not all that counts for UHI.
However, this study is a valuable contribution to the issue, It would be very interesting to deepen our understanding of the issue by including population census data from, say, 1970 to 1990 and also 2010, to assess whether changes in population or urbanization have anything to do with temperature trends. And including indicators of the presence or absence structures (such as a vegetation index or a geo-referenced map of buildings and paved roads in the vicinity of stations) may be another useful development of the study.
Zeke,
on the log() of population density. Others have used popdensity^1/4. That functional form has the benefit of actually being defined at 0 population. One of the issues with population density is that the same density can give rise to vastly different anthropogenic heat and different transformations of the land based on regional building practices. In most studies ( since Oke) folks have looked at actual physical causes such as changes to the surface.
Simply. people change the surface. more people more change.
Its the change to the surface that drives the majority of UHI.
So you get a clearer picture if you look at the proximate cause: changes to the land. If people made the same changes ( if building practice was the same in the SH as in the NH) then population density would be less confounded. Other problems are industrial areas and airports. In global population data products the population density in industrial locations ( mining camps, refineries ) is very low. sometimes zero. The same with airports. That’s because the population products measure residential population and not ambient population.
“Changing the surface” is one way of putting it. Another way of putting is replacing deep rooted plants that keep the air cool and humid with hot dry air. People effectively create urban deserts with impermeable surfaces and rainfall drained away via sewer systems and other high volume low surface area conduits to even higher volume lower surface area repositories like rivers, lakes, and oceans.
This notion that water vapor is the primary greenhouse gas greenhouse is highly misleading. Where there is water vapor there is also evaporation, convection, and rain. The net result is a negative feedback and the empirical evidence to support is that the highest mean annual temperatures at any given latitude happen where there is the least rainfall.
The so-called greenhouse effect is the result of a liquid ocean thousands of meters deep on average that covers 71% of the globe. The albedo of the ocean is practically zero and solar energy penetrates at the speed of light to a depth of about a hundred meters. Removing the energy imparted by that light at depth requires mechanically moving the deep water to the skin layer (the topmost 1 millimeter) where it cools primarily by evaporation and secondarily by radiation.
The result of this is that SST has negligible diurnal variation. This lack of dirurnal variation allows the surface temperature to approach the black body temperature which is otherwise greatly skewed downward due to far higher rate of heat loss with higher temperatures. The highest mean surface temperature, in other words, is attained with the least variation when everything else is equal. This is why the moon has a mean surface temperature far, far below the temperature of the mean power input of 342W/m2. The reason why deserts on earth attain mean temperatures higher than ocean temperatures despite deserts having higher diurnal variation is that evaporation and convection are because evaporaiont and convection remove heat from the surface without raising sensible temperature, causing a smaller lapse rate, whereas deserts have a higher sensible component and higher lapse rate to go along with it.
The other problematic issue with 1km population data is that it is heavily modeled. Some of the key data it relies on ( Roy cites GRUMP here ) is administrative boundary data and urban boundary polygons. The net result is that lower population densities, population densities not associated with any administrative boundary or urban polygon are not very accurate. Also, the distinction between urban population and rural population is lost.
Read the description here
http://sedac.ciesin.columbia.edu/gpw/documents/Balk_etal_GeostatPaper_2010pdf-1.pdf
One thing you’ll note that is missing from this is any kind of metrics for producer accuracy. Spend a few minutes with GRUMP data. compare the population density it gives with reality.
You can do that pretty easily. Answers are interesting to say the least.
Dear Roy,
I too was convinced that there must be an urban warming effect on global temperature, based on simple physics. Total Average World Energy consumption Rate = 15 TW (IEA 2005) and increasing by 2%/year. My guess is that 80% of this is converted to heat (2nd law thermodynamics)
The land Surface Area of the Earth = 150 x10**12 m2 of which urban areas are roughly 1.5% So if we assume that energy consumption is concentrated in these urban areas then the human heating effect there works out at 5.5 watts/m2 in cities. Direct heating by Man in urban areas comes out at approximately 2% of direct solar radiant heating.
What temperature increase would this lead to ? Using Stefan Boltzman’s law.
(T+DT)**4 = 1.02T**4
(1 +DT/T)**4 = 1.02
DT = (0.02/4)T ( T = 285)
DT = 1.4 degrees ! (i.e. the averaged temperature increase in urban areas)
This also corresponds to our everyday experience where cities are generally a couple of degrees warmer than the surrounding countryside. So there is an urban heating effect without doubt. The question is how does this effect anomalies (changes in temperatures with time) ? I read the BEST (no relation!) paper http://berkeleyearth.org/pdf/berkeley-earth-uhi.pdf which after excluding urban areas using MODIS land cover, deduced that there was no urban heating effect at all ! What I think this means is that there is a subtle difference between temperatures ( which I believe are increased by urbanization) and anomalies. Yearly Anomalies effectively measure the rate of change DT/dt at one single place. So for stations based within cities the anomalies would only increase if local heating density increased. Only stations originally in rural areas which are swallowed up by development would indeed show anomaly warming.
Meanwhile I downloaded all the station data for HADCRUT3 and identified all those stations which have warmed more than 1 degree C. from the periods 1891-1920 to 1991-2010. I found that SAO PAOLA had warmed by 2.4 degrees, and among the top 50 stations were Beijing, Tokyo, San Diego, New York and San Francisco. However I then repeated the BEST procedure and simply excluded the top 70 most warming cities. Then I recalculated the anomalies and compared them with the original full CRUTEM3 set. The difference is almost negligible. – since 2000 the global average falls a maximum of just 0.02 degrees. C. ( see http://clivebest.com/blog/?p=3434 )
Therefore I conclude that the BEST result is valid and that urban heating effect has little direct effect on the way global anomalies are derived. However it is still possible that this simply means that the energy density for those stations already urbanized by 1900 remains constant, while those rural stations which have consequently been effected by urban sprawl warm .
Since then we now have CRUTEM4 which has apparently managed to boost recent temperature anomalies. This is basically a systematic effect due to their addition of over 600 arctic stations which bias geographic coverage ! – see http://clivebest.com/blog/?p=3528 . Thus we have the press release from the MET office that 2010 is 0.01 (+-
keep up the good work !
Very interesting result, indeed.
I don’t know if it’s possible, but it might be also interesting – instead of comparing land station data with another land station data – to compare station data with satellite data for the same location, yet again stratified by the population. I mean, I believe satellite data are not as much affected by UHI as land station data as they measure temperature of the mass of the atmosphere rather than the temperature right above the ground.
I agree with Hector M, higher trends should be caused by a changing level of urbanization, not a higher but constant level.
There is reason to look very closely at the northern hemisphere trends: http://www.woodfortrees.org/plot/crutem3vnh/mean:36/plot/crutem3vsh/mean:36/plot/hadsst2sh/mean:36/plot/hadsst2nh/mean:36
Which is not to say there are for sure significant errors, but the recent divergence is at least surprising.
When I graphed the HAD and UAH numbers I get same trend down to .001 degrees per year. Surely this isn’t statistically significant? If they are talking about the GISS records, then they might have a point. GISS seems to be marching to a different drummer of late. GISS seems to show a much smaller drop than the other indexes after 98.
http://www.anupchurchchrestomathy.com/2012/02/comparing-giss-hadcru-rss-and-uah.html
Click on the graph to expand it.
I may have to change my mind when I see the new improved HADCRU data, but luckily I had the old data archived. A comparison might make a interesting entry in my blog.
Thanks Dr. Spencer. Again UHI is shown to be a major “forcing” on thermometer’s data.
And even then, they showed a cooling trend after 1998.
Dr. Spencer, with respect;
You are finally arriving at the conclusion that the flow of heat through the climate system is limited by the thermal diffusivity, from a system level perspective this is essentially the ”speed of heat”.
As I’m sure you realize, Urban Heat Islands are largely composed of manmade materials like concrete and asphalt.
Here is a reference that lists the thermal diffusivity of several different materials;
http://www.hukseflux.com/thermalScience/thermalConductivity.html
Please Note that the values for Saturated Sand (102) and Quartz (141) are BOTH higher than Concrete (66). This supports the conclusion that Urban Heat Islands SLOW the flow of heat through the system, which results (temporarily) in a higher temperature.
Also, please note that the value for “air” (comprising N, O, CO2, etc.) is significantly higher (i.e. 1938, more than an order of magnitude larger, it should be noted). This indicates that heat flows much faster (i.e. at a higher velocity, propagation rate, distance travelled per unit time, etc.) through “air” than through solids like sand, rocks, and water.
If you think this whole “thermal diffusivity” thing is more ”pseudo-science” from a “denier” you might want to research this documented method used to measure the “thermal diffusivity” of materials;
http://www.stormcable.com/uploads/Thermal_expansion_6702_Ch1_web.pdf
The ASTM group has detailed and documented almost all of the methods used to determine how well different materials perform under all kinds of conditions.
And in the unlikely event that you want to measure the thermal diffusivity of some stuff lying around in your backyard here is somebody that will sell you a thermal diffusivity meter. I personally have a pretty cool tool collection, but I must admit that I do not have a thermal diffusivity meter.
http://www.ulvac.com/thermal/laserPIT.asp
Cheers, Kevin.
See the graphs below which were made in such a way that differences are very clearly seen. Is it possible that UHI is at least partly responsible for the difference between the satellite slopes and the slope of the others?
http://www.woodfortrees.org/plot/uah/from:1978.9/trend/offset:0.31/plot/rss/from:1978.9/trend/offset:0.22/plot/gistemp/from:1978.9/trend/plot/hadcrut3gl/from:1978.9/trend/offset:0.08
This was noted at WUWT by DR
http://www.real-science.com/important-point-from-roy-spencer
“if the satellite warming trends since 1979 are correct, then surface warming during the same time should be significantly less, because moist convection amplifies the warming with height.”
(this maybe a double post, if so I apologize in advance)
Dr. Spencer, with respect;
You are finally arriving at the conclusion that the flow of heat through the climate system is limited by the thermal diffusivity, from a system level perspective this is essentially the ”speed of heat”.
As I’m sure you realize, Urban Heat Islands are largely composed of manmade materials like concrete and asphalt.
Here is a reference that lists the thermal diffusivity of several different materials;
http://www.hukseflux.com/thermalScience/thermalConductivity.html
Please Note that the values for Saturated Sand (102) and Quartz (141) are BOTH higher than Concrete (66). This supports the conclusion that Urban Heat Islands SLOW the flow of heat through the system, which results (temporarily) in a higher temperature.
Also, please note that the value for “air” (comprising N, O, CO2, etc.) is significantly higher (i.e. 1938, more than an order of magnitude larger, it should be noted). This indicates that heat flows much faster (i.e. at a higher velocity, propagation rate, distance travelled per unit time, etc.) through “air” than through solids like sand, rocks, and water.
If you think this whole “thermal diffusivity” thing is more ”pseudo-science” from a “denier” you might want to research this documented method used to measure the “thermal diffusivity” of materials;
http://www.stormcable.com/uploads/Thermal_expansion_6702_Ch1_web.pdf
The ASTM group has detailed and documented almost all of the methods used to determine how well different materials perform under all kinds of conditions.
And in the unlikely event that you want to measure the thermal diffusivity of some stuff lying around in your backyard here is somebody that will sell you a thermal diffusivity meter. I personally have a pretty cool tool collection, but I must admit that I do not have a thermal diffusivity meter, YET.
http://www.ulvac.com/thermal/laserPIT.asp
Cheers, Kevin.
Thanks again Dr. Spencer,
I have published quotes from this excellent article in the section on Global Temperatures of my page “Climate Change (“Global Warming”?)
– The cyclic nature of Earth’s climate”, at http://www.oarval.org/ClimateChangeBW.htm
(Spanish version at http://www.oarval.org/CambioClimaBW.htm)
FYI.
Using your population classes and Berkeley Earth station data
I find
1 2 3
15348 12797 6471
that’s the count of stations of each population class.
Roy
Surface temperatures need to be put in their place. They sould be weighted ..
Land Surface : Sea surface = 6 : 89
or thereabouts, based on the relative theraml energy contained in land and oceans. This is the only weighting that makes any sense. We know that ocean temperatures have a strong influence on islands and coastal land, and even some effect further inland.
It seems you would be in a good position to recommend such.
And when is NASA going to do something about the failed sea surface measurements, as these are by far the most important.
So warming is greatest where human population density is greatest. That’s not news. News would be that this has significant influence on the global temperature trend. High population centers have been knocked out of the dataset before and there is no material difference in the global trend. There are just too few high population centers to make much of a difference.
Aside from that, playing devil’s advocate here, high population centers are high anthropogenic CO2 producers. The atmosphere appears to be well mixed over time but that mixing takes time and in the meantime population centers are going to have a far higher level of the gas simply because it takes time to get from production centers to sinks.
This is hardly worthy of note.
A mighty valuable piece of work, most admirable. Dr Spencer is of course right to point out that “Analysis of the raw station data is not for the faint of heart.” He is far from faint of heart. The more strength to his arm.
As Steven Mosher points out, the log/lin straight line fit seems particularly inappropriate.
In as much as any relationship can be defined by three dots, this does not appear to be well characterised by log(pop) and you are extrapolating well out side your very limited data with unknown uncertainty.
Since this is at the very heart of your headline result which is stated again without any uncertainty estimations, I think you are going out on a limb with comments like:
“I’ve appended the results for the U.S. only, which shows evidence that CRUTem3 has overstated U.S. warming trends during 1973-2011 by at least 50%.”
While the gatekeeper conspirators have been exposed by their very own communications and have shown no indication of contrition since, not submitting this for review does not seem to be a good reason not to apply the usual rigour required for science.
Please don’t be tempted to adopt the flawed science of the AGW proponents in order to refute them. That road will only lead to pointless “did-didn’t…” shouting games.
I have a lot of respect for you published work, please stick to rigorous science, the world needs more of that not less.
BTW, it seems most of the divergence in CRUTem3 is post 1995. Knowing how this crowd work, I’d be looking for some other “bias correction” they are doing to warm up the later period that is not fitting their preconceived ideas.
Their “partners in crime” at Hadley have been inserting ever increasing “corrections” into the SST data over the same period.
best regards.
Roy: Your analysis seems promising. Since you asked how can one explain such results, I’ll note that results don’t have error bars or evidence that they are statistically significant. SInce no theory exists explaining why UHI should vary with the log of population, the bar graph of the US data is more appropriate than the scatter plot of global data with log population on the x-axis.
You’ve developed important information about how recent temperature TRENDS vary with population density. Biases in trends are far more important than UHI (a term you don’t even need to mention) or the general temperature record. Averaging all of the temperature trends inside every grid cell no longer makes sense – the average temperature station has higher population density than average surface location. The resulting temperature trends MUST be biased and this bias applies to every method using grid cells, not just CruTemp. Can you calculate the size of the bias?
How much of the surface of the earth has high, medium and low population density? Let’s suppose that the numbers are 5%, 15%, and 80%. One might try to remove the bias by calculating the overall global land trend as: 80%*0.235 degC/decade + 15%*0.252 degC/decade + 5%*0.285 degC/decade. (I estimated these trends from your graph.) However, one probably needs to do this type of analysis for individual grid cells.
Clive Best links his examination of CRUTemp4 here: http://clivebest.com/blog/?p=3528
Though there seems to be a large QC problem for 2009 and 2010 which present a huge spike, it seems that much better coverage of high northern latitudes has considerably reduced late 20th c. warming in NH , hence globally.
This could be a challenge to Hansen’s ridiculous arctic extrapolations.
To answer your question, Roy, I think we can assume from CRUTem4 that not only will the warming biases in CRUTem3 not be corrected in the next IPCC report, they will even be amplified.
Great post just now by Warwick Hughes on what CRUTem4 has done to the rural gridcell in southeast Australia known as the Murray-Darling basin: http://www.warwickhughes.com/blog/?p=1460. How to smear urban warming over a huge inland area which in previous versions showed no significant warming, and where there aren’t even any big city stations!
The valleys in Switzerland are densly populated, and the homogenized data shows warming (some only after homogenization, as in Sion).
But the mountain peak of Säntis shows no significant warming for the last quarter century.
I strongly suspect that the (real) warming is due to a growing population, getting richer. So it is AGW, since immigration, legal and illegal, is man made.
(The homogenized data for Säntis can be found at Meteoswiss, or, adjusted a second time, at GISS.)
The man made global warming theory is dead. Temperatures have failed to go up for the past 15 years or so, and all their stupid reasons for the temp. rise have failed to materialize. Some of those being the hot spot in the upper troposhere near the equator(due to positive feedbacks between an increase in co2 and water vapor), a more positive AO, stratospheric cooling (especially near the poles), Arctic Ice Cap meting away, less extremes in weather ,due to their more zonal outlook(+ao) wrong again,more El Ninos, wrong again. Rising sea levels wrong again. Less snow especially in the N.H., wrong again.
Their total disregard for solar influences on the climate ,along with the PDO/AMO ,volcanic activity, etc. etc .etc.
Their asinine excuses for why the global warming has stalled, due to heat being stored in the deep oceans ,to aerosol increases, etc etc. A joke.
They are a bunch of frauds.
The theory will be dead before this decade ends officially, although it is currently dead as of now ,as far as I am concerned.
Be prepared for cooler temp. as this decade proceeds ,especially if the solar flux falls to sub 90 readings.
Please entertain a naïve question from a non-climate scientist.
Is the extent of CO2-caused warming likely to be affected by altitude? Large population densities tend to arise along major rivers, so it would not be surprising for there to be a confounding correlation between population density and elevation. Of course, higher elevations are cooler, but the critical question is whether temperature trends attributable to increasing greenhouse gases would be ameliorated.
I don’t doubt that the UHI effect is there but it is not my main concern. What concerns me far more is the outright fraud of raising the temperature in the eighties and nineties to create a so-called “late twentieth century warming.” I came to this conclusion when comparing HadCRUT3 to satellite temperature curves. That means both UAH and RSS. They are so close in that time period that they practically fall on top of each other. That in itself should raise confidence that their data are real. And they do not show a steady warming in the eighties and nineties, prior to the appearance of the 1998 super El Nino. What you see in satellite data for this period is nothing but a series of ENSO oscillations about a mean temperature that remains constant for almost twenty years. There are five El Nino peaks in this period with La Nina valleys in between. And by the way, that Pinatubo cooling you have marked out on your web site is imaginary and does not exist. It is just the 1992/93 La Nina misidentified as volcanic cooling thanks to Best who did not understand in 1996 what the satellite record really contains. To bring out the temperature trend from the satellite record you must first use a magic marker by hand, wide enough to cover most of the fuzz that that is part of the record. Do not use computer smoothing. That fuzz is not random noise but randomly distributed cloudiness effect. Once you have done that you will get a series of peaks and valleys. Next, put a dot at the midpoint of each line connecting an El Nino peak and its adjacent La Nina valley. Now connect these points. You should get a linear horizontal line with slight scatter from 1979 to 1987. Fit the best straight line through these points by hand. What you now have is the best possible estimate of global mean temperature for this period that you can get. Now get hold of a graph of HadCRUT3 on the same scale and do the same with it. You will notice more scatter and outliers but all the El Nino-s and La Nina valleys are there. You will notice immediately that when you draw the straight line that defines the mean temperature it is no longer horizontal but slopes up at 0.1 degrees per decade. Closer inspection reveals that the first four El Nino peaks of both curves coincide. The fifth and last El Nino peak is lower in satellite data but has been raised in the ground-based data to line it up with the first four. But the major difference is in the La Nina valleys in between the peaks. They have all been raised so that the depth of a valley in the ground-based record is only half the depth of the same valley in the satellite record. Check out figure 24 in my book “What Warming?” I know of no natural process that can zero in on low temperatures this way and raise them without raising the adjacent high temperature points as well. The time interval between the peaks is somewhat variable but averages five years. It requires a precise and long term process, active for twenty years, to raise all the valleys involved as consistently as they are. I can only think of one thing that can do it, and that is anthropogenic warming. It started approximately at the same time as the satellite record does, about 1979. It is not limited to HadCRUT3 alone and is still continuing to the present as comparison with the rest of the satellite record reveals. To get such long term and widespread and coordinated action one needs to have some central control. I looked at the records available on the web and came across a document posted by GISS. It is called “GISS (Goddard Institute of Space Studies) Surface Temperature Analysis.” in the “History” section it says: “The basic GISS temperature analysis was defined in the late 1970s by James Hansen when a method of estimating global temperature change was needed.” Here I want to point out that the period from 1950 to the late seventies did not experience any temperature rise. There was a method of recording temperature already in use and it is somewhat puzzling why a method for determining temperature change was even needed at that point. In his book about his grandchildren Hansen relates that he left the Pioneer Venus Project in 1978 to join GISS because “The composition of the atmosphere of our home planet was changing before our eyes.” Didn’t even finish his experiment that was on the way to Venus. Is it just a coincidence that ground based temperatures started to rise as soon as his method was invented? We know what happened: from the late seventies on global temperature began to rise and by 1988 Hansen could declare to the US Senate that anthropogenic warming had arrived.
Dr. Spencer,
Isn’t the UHI effect corrected in the BEST project of Muller and Curry? And isn’t UHI also a cause AGW since man-made structures absorb more solar radiation? Why are you excluding this effect if it’s causing artificial warming?
Man-made structures don’t absorb more solar energy on average. Not everything manmade is fresh blacktop. Impervious ground cover and channeling away of rainwater in high volume low surface area conduits greatly reduces evaporative cooling.
It’s all about water in all three phases. Non-condensing greenhouse gases can be ignored on water worlds like ours. Water is not just necessary to life at the level of biochemistry it’s also necessary to a stable climate that stays within reasonable surface temperature bounds.
Dr Spencer, you write:
“These results are roughly consistent with the McKitrick and Michaels (2007) study which suggested as much as 50% of the reported surface warming since 1980 could be spurious.”
If that is the case, can you tell me why the surface temperature trends are not twice as large as the UAH or RSS trends, which are unlikely to be subject to the UHI effect?
woodfortrees.org gives the following trends (1980-present):
CRUtemp 3 (global land only) 0.0207769 per year
RSS (global land only) 0.0190954 per year
UAH (global land only) 0.0173541 per year
so the differences seem fairly slender. WFT doesn’t have GISTEMP for land only, but the global land sea trends are
GISSTemp slope = 0.0155201 per year
RSS slope = slope = 0.013664 per year
UAH slope = 0.0136882 per year
So again the surface measurements are a bit higher, but no way near twice as high as if 50% of the trend was spurious. Obviously the surface temperature measurements are not measuring exactly the same thing as the satelite measurements, but the headline figures do not seem to support the contention that 50% of the warming is spurious.
Dr. Roy’s results are indeed consistent with McKitrick and Michaels (M&M), and the fact that Dr. Roy and M&M use rather different techniques is evidence for the robustness of the result. M&M did not disaggregate stations by population density. Rather they applied an econometric model that looked for correlations between the gridded warming trend and indices linked to human activity. Their spatial resolution is therefore fairly gross and they remark on this in their paper: “Since we detect significant effects on temperature trends even with low spatial resolution we conjecture that if future studies are able to examine the issues at the subnational level, even more significant and detailed results will emerge.” Dr. Roy’s analysis is a step in this direction.
M&M also remark that the removal of evident bias brought the histogram of trend observations in close accord with the satellite measurements, mainly by eliminating the large tail in the uncorrected data.
What about Dikran’s post Dr. Spencer? The 34+year trendlines of the UAH and CRU data do not lend much support or significance to your argument. Could it be because your argument doesn’t mean much to the overall global temperature data trends, even if you conservatively account for the likely 0.06C UHI global contribution to the ground-level CRU data?
See the attached graph of actual data and trends for your review:
http://www.woodfortrees.org/graph/hadcrut3gl/from:1978/offset:0.39/mean:12/plot/gistemp/from:1978/offset:0.3/mean:12/plot/uah/from:1978/offset:0.54/mean:12/plot/rss/from:1978/offset:0.54/mean:12/plot/hadcrut3gl/from:1973/offset:0.39/trend/plot/uah-land/from:1978/offset:0.54/trend
UAH land based temps vs CRUTEM3 (land only)
http://www.woodfortrees.org/plot/uah-land/mean:13/plot/crutem3vgl/from:1979/mean:13
Can anybody spot a difference other than the baseline climatolgy for caculation of anomalies.
Makes a mockery of your assumptions Roy.
“sillyfilly says:
April 2, 2012 at 4:28 PM
Makes a mockery of your assumptions Roy.”
I am not so sure. Please see how I enhanced your graphs:
http://www.woodfortrees.org/plot/uah-land/mean:13/offset:0.34/plot/crutem3vgl/from:1979/mean:13/plot/uah-land/mean:13/trend/offset:0.34/plot/crutem3vgl/from:1979/mean:13/trend
re: sillyfilly
Not sure what your point is sillyfilly. The CRUTEM ols trend in your graph is 0.21C per decade. The UAH ols trend is 0.17C per decade. If you detrend Crutem by 30% to 50% wouldn’t that put the tropospheric trend squarely into the theoretical surface temp amplification range?
“Makes a mockery of your assumptions Roy.”
If you’re asking me to do an eyeball test, then the diverge cira 1980-1990 is obviously significantly less than the divergence 2000-2010, so not sure what you’re smoking but maybe you should stop, because it seems to be scrambling your brain ‘SillyFilly’.
Could there be a built-in bias for larger trends at larger population densities (Figure 1), given polar amplification combined with an approximate pattern of larger cities in the mid-to-higher latitudes of North America, perhaps Europe, perhaps eastern Asia, etc?
(Viz. is there any rough trend in population density as a function of latitude?)
To the LL and WB;
So lets add RSS land only to the graphics. Thus we see RSS trends lying between UAH land only and CRUTEM.
http://www.woodfortrees.org/plot/uah-land/mean:13/offset:0.33/plot/crutem3vgl/from:1978.92/mean:13/plot/uah-land/trend/offset:0.33/plot/crutem3vgl/from:1978.92/trend/plot/rss-land/from:1978.92/to:2012.08/offset:0.23/mean:13/plot/rss-land/from:1978.92/to:2012.08/offset:0.23/trend
Underlying trends:
CRUTEM: 2.07 / century
RSS: 1.91 / century
UAH: 1.71 / century
Still seems like 50% is a gross overstatement and still unsubstantiated.
Dr Spencer,
I wrote a post about this and would like to hear your comments: The spurious signal is a higher proportion of the the long term trend.
http://naturalclimate.wordpress.com/2012/03/31/northern-hemisphere-uhi-crutem3-18/
sillyfilly, you should have a look too. It might be 100% or more…
It is very simple: if the half of the warming in the surface station data were a spurious artifact due to UHI effect, then their true value would be half what it actually is.
Now if we assume that the warming in the lower trophospheric temperatures over land should be about the same as that in the true surface temperatures, then the biased surface station data ought to show a trend about twice that of UAH or RSS.
The problem is that they don’t. The warming in the surface data is only 10% more than RSS and about 20% more than UAH. This suggests that any bias in the surface temperature warming is much less than 50%.
This is such a straightforward sanity check on the basic claim, I am surprised it wasn’t mentioned by M&M nor Dr Spencer.
Dr Spencer, I have some questions;
1) Does your US assessment include Alaska and Hawaii?
2) How have you area weighted your results?
3) What happens to the trend and intercept in your first graph if you plot individual station data, rather than binning them?
4) What happens if you average box-by-box? E.g. say you have 3 high density and 3 low density stations in a box, average each of them and compare, so that you compare box by box?
5) Are there any biases in latitude, altitude or microclimate effects that correlate with the population density?
6) How does your data fit with Menne, the Watts paper, BEST and the satellite data?
I’m very skeptical of your 50% claim, I think it’s a very big push based on a linear extrapolation from 3 data points.
The problem many people have when it comes to climate change is they try to pin it on one particular item. When the reality is , it is a combination of a zillion items, and how they phase or not phase in relation to one another ,along with the duration of time.
My phase in theory addresses this , which I have posted many times on these message boards.
I also address abrupt climate change.
I say ,my theory is as good or better then anything out their, and I will have a good chance to find out just how good it is, before this decade ends, if the sun can stay quiet, which is looking more and more likely , as sunspot cycle 24 fails to show any siginifcant signs of strength.
All indications are in general with some lulls from time to time, for colder temp., more geological activity, and more extremes in climate as the decade continues.
As ,I might add ,has been the case thus far for this decade.
The GISS surface station trend is about 0.2 C per decade, but you need to use excell or some other statistical tool and do the analysis yourself.
The CRUTem3 data set is not the official IPCC data set, maybe you should read the latest IPCC report to confirm whether or not they advocate one data set over any of the others.
CRUTem3 doesn’t even provide complete land station coverage and can by no means be called a global record.
All the data sets have their biases and those should be taken into account when comparing trends.
“sillyfilly says:
April 2, 2012 at 10:45 PM
Still seems like 50% is a gross overstatement and still unsubstantiated.”
Granted, the numbers shown do NOT show a 50% increase. Now I could be wrong here, but I see two possible reasons for this:
1. Dr. Spencer just used US land and not the whole earth’s land, and/or
2. Just cities were compared so many land sites in rural areas of the US partially neutralized the spike due to the UHI of the cities.
@Werner I evaluated the land only trends for UAH, RSS and CRUTemp. The CRUtemp trends are no where near double the satellite trends as you would expect if 50% of the surface trend was spurious.
We all know that the UHI effect affects measurement made by surface stations, which is why the gridded products such as CRUTemp adjust to compensate for such non-climatic biases.
If Dr Spencer’s analysis used the station data AFTER it had been adjusted, rather than before, THEN he might have some evidence that the surface gridded products such as CRUtemp had a spurious bias, but even in that case it can be no where near 50% of the trend being spurious, and the comparison with satellite products (which are not susceptible to UHI effect) makes that very clear.
Dikran Marsupial,
“We all know that the UHI effect affects measurement made by surface stations, which is why the gridded products such as CRUTemp adjust to compensate for such non-climatic biases.”
Do you have a reference to this?
The only adjustments that I know of CruTem (Alpine cells) correspond to an artificial warming of the raw series of 0.5 ° C over the twentieth century.
@phi, I don’t know the details of hand, but I remember reading a paper on the homogenisation etc. that is used by pretty much all of the datsets. The code for GISSTemp is available in the public domain, which along with the GHCN adjustments should be representative of what is done.
Note Dr Spencer writes “Here I will show based upon unadjusted International Surface Hourly (ISH) data archived at NCDC …” Of course the unadjusted data shows UHI bias; the question is whether the adjusted data show a bias and whether the gridded products show UHI bias. A comparison with the trends from UAH and RSS suggest the answer is probably “no, not very much”.
GRUMP population data.
I think I pointed this out many moons ago over at Lucia’s.
For much of the world population records are kept at ‘administrative’ level.
As an example…if I lived slightly outside the city limits of Enumclaw Washington in an 8 house/acre patio home subdivision I would be a resident of Unincorporated King County Washington which includes Mt Rainier National Park and a fair portion of the uninhabited Cascade Mountain Range. There is actually a thermometer there and Grumps would tell you that it is a rural simply because the population density for ‘Unincorporated King County’ is extremely low ‘on average’.
If you go to the Middle East Riyadh is an administrative district that covers maybe 20% of the entire land mass of Saudi Arabia and a very large and densely population city…almost all the the Riyadh Administrative district is uninhabited. If I look at Grumps it’s listed as ‘rural’. The thermometer is at the airport. The same problem exists for Basra and Mosul. The ‘city boundary’ is the whole ‘state/county’ or whatever they call it.
If I read the gist of this properly, Roy has looked at very raw station data and observed a bias associated with higher population densities. Put simply, the Urban Heat Island effect.
This isn’t particularly ground breaking news. This has been recognised for quite some time and all the major temperature products include a compensation to remove the effect of UHI.
So to say anything meaningful Roy would need to show that these compensations are inadequate. This post doesn’t seem to do that.
Next Roy calls HADCruT3 the ‘gold standard’! Say what! There are 5 major surface temp products – Had/CRU, NCDC, GISS, BEST, JMA as well as a number of re-analysis products. All producing comparable reults. No particular one is ‘a gold standard’.
And why hase Roy used HADCruT3 when HADCruT4 has been released?
The whole surface stations thing has been done to death and only the die-hards still keep pushing that broken down wheelbarrow. If anyone is interested in an overview of much of this, I authored this 4 part series about it some time ago, before the BEST results came out http://www.skepticalscience.com/OfAveragesAndAnomalies_pt_1A.html In particular, parts 3 & 4 look at station bias issues.
“And why hase Roy used HADCruT3 when HADCruT4 has been released?”
In fairness, HADCruT4 hasn’t been available all that long and presumably Dr. Spencer started working on this analysis a while ago. I can’t see any reason why he shouldn’t publish what he has done so the rest of the world can consider it rather than have to wait until he manages to redo the whole thing using HADCruT4.
The end of the Global Warming scare hinges on understanding that there has been a false assumption made by the early physicists. Let me explain why ….
Not all photons are equal. You cannot assume (as the IPCC energy diagrams imply) that the electromagnetic (EM) energy being radiated will always be converted to thermal energy in the Earth’s land, ocean or ice surfaces.
The early physicists did in fact assume this when considering (and trying to explain) the observed effect of radiation from a cooler body slowing the rate of heat transfer from a warmer body. This does happen, and is well measured and documented. The “net” heat transfer (as they called it) can be calculated from the difference in the two amounts of radiative flux.
So they assumed that the energy in the radiation was in fact converted to thermal energy in each body. The mathematical subtraction then of course only showed a “net” effect of heat transfer from hot to cold. But the underlying assumption was still there that there were in fact two separate heat transfer processes. Yet one of these would violate the Second Law of Thermodynamics (SLoT).
So, can we find an example of EM radiation not being converted to thermal energy when we might expect it to be?
A microwave oven can warm items with water molecules in them, including liquid water. This does not violate the SLoT simply because energy is added using electricity. But it can only melt ice by conduction from adjacent water molecules that it has already warmed, not by direct action on the ice.
However, the process is nothing remotely like the normal natural absorption of sunlight which also warms water and melts ice.
Not all photons striking water or ice molecules automatically convert EM energy to thermal energy as happens with solar radiation. If they did (as some people imply they do because they assume there is two-way heat flow which results in an apparent net one way flow) then why does far less energy flow into ice in a microwave oven than into water?
There is another mechanism which explains all observed facts without violation of the Second Law, and this is discussed in my paper linked from my site http://climate-change-theory.com
There’s more to it, so please consider many other points raised in the paper.
If your model is fudged then you have to fudge your results to save face.
The IPCC implies with their energy diagrams that the electromagnetic energy in radiation from a cooler atmosphere is all converted to thermal energy in a warmer surface, which is contrary to the Second Law of Thermodynamics, because it would be an independent process. There is certainly no contrary statement by the IPCC, or any footnote for the energy diagrams.
The IPCC does not specifically state that only the solar radiation transfers thermal energy to the surface, whilst the backradiation does not do so, but only slows the radiative component of cooling for that component of the radiation from the surface which does not get through the atmospheric window. Yet this in fact is all that can happen.
The above is the kind of explanation that should have been what was in their most comprehensive explanation of their GH conjecture, for the benefit of scientists.
Even then, they should further have discussed what compensating effects there might be in the rates of evaporation and sensible heat transfer, both of which would obviously increase if radiation did less for its share of the cooling. It is well known that these processes accelerate if the temperature gap widens, as it would if radiative cooling slowed. None of this is in any IPCC explanation, now is it?
Anybody interested in how these sorts of issues are dealt with in gridded analyses suchas CRUtemp should indeed check out Glen’s articles that he mentions in his post up-thread.
Doug Cotton: radiative transfer calculations are either done on a ‘line by line’ basis, i.e. across all wavelengths, some of which are absorbed and some of which aren’t, or by including a spectrum weighted emissivity.
This accounts for everything you talked about in your first post.
In your second post you discuss changes in sensible and latent heat flux, this is included in the ‘lapse rate feedback’.
You’ve thought a lot about this, but I think you missed a few crucial papers in your reading. A more mathematical treatment is available in:
Atmospheric Radiation Theoretical Basis 2nd Edition – R M Goody & Y L Yung.
p33 onwards demonstrates how the basic equations of radiative transfer are defined on a wavelength by wavelength basis, and absorption is calculated based on the quantum structure of the material, again on a wavelength by wavelength basis. Include these effects and you get the IPCC conclusions.
MieScatter
I am quite aware of how radiative transfer equations are applied “wavelength by wavelength” but this is not a contra argument to my posts. What I am talking about is the physical mechanism involved.
I can explain and quantify everything and get the same results, but the mechanism I describe in my peer reviewed Radiated Energy and the Second Law of Thermodynamics is totally different, and, whereas the concepts in Goody and Young violate SLoT, mine don’t.
There’s no need for you to explain why a microwave oven can boil water but not melt an equivalent amount of ice, for I am aware of what happens. But my point in discussing this is to demonstrate that equal radiation striking the water and the ice leads to totally different amounts of energy being converted to thermal energy even in the same H2O molecules receiving the same radiation, but some molecules in solid form and the others liquid.
Try your radiative transfer theory on that simple case and explain what happened to the radiation which struck the ice. Not much was reflected or transmitted, and nor was it absorbed. There is another process I call “resonant scattering” which is happening and which explains and is consistent with all observed empirical data, without violating the Second Law of Thermodynamics (SLoT).
By the time you get to Section 5 of my paper you will be starting to understand, if your read with an open mind.
But since you mention the need to take into account all wavelengths, I would comment that this is recognising that, as you know, S-B is based on integration of the Planck function, using effectively all wavelengths for a composite solid surface, similar in some ways to a blackbody.
But with gases there are of course just the spectral lines. However, those lines cannot exceed the intensity for which the Planck function sets a maximum for any particular wavelength and temperature. So radiation per molecule from CO2 is less effective on average than that from water vapour which has more spectral lines – a fact somewhat at odds with IPCC claims, and putting CO2 in its place with an overall effect on the rate of radiative cooling of the surface of probably less than 1% that of all the water vapour.
There’s more discussion on the above here
Here’s another more detailed post.
MieScatter, Glenn Tamblyn,
I do not think there is any consideration of UHI by CruTem. Some series are adjusted, in particular a number of European cells are aligned on ALPCLIM.
For these adjustments, see here: http://onlinelibrary.wiley.com/doi/10.1002/joc.689/pdf
CRU uses now (CruTem3) these adjusted data.
Roy…I second Barry Bickmore’s sentiment, don’t give up on publishing. There are other journals that are not controlled by the peer review police.
Put it out there and let the legions of skeptics get it around the Net by word of mouth. I have been hammering anyone who will listen with your UAH satellite graph, which single-handedly destroys the AGW theory, although you still support that pseudo science. We in the Clausius fan club have not given up on you.
We all know the surface record is corrupt. NCDC has been chopping stations right and left, and although the stations are still there, they are selecting which ones to use. I think spurious results are the least of our problems.
http://chiefio.wordpress.com/gistemp/
Doug Cotton,
if you want to “correct” the current understanding of the physics please go and convince mainstream physicists and then come back and tell use how it changes analysis of climate. I don’t thing Roy’s site needs filling with your pet theory. At least respect Dr. Spencer’s site enough to discuss what he chooses to post on not just use it as a spring board for your own subjects.
P.Solar and others
As a member of Principia Scientific International I am in constant contact with main stream scientists, including professors and PhD’s in various disciplines such as physics, applied mathematics, chemistry, climatology and astro physics. The numbers are approaching 40, including well known new members just announced.
These are not just my theories. We are all in agreement that standard physics and empirical results back us up.
What empirical proof do you have of IPCC theories, or Roy Spencer’s middle-of-the-road beliefs? Lay your cards on the table. Explain, for example, what you think happens to the low frequency radiation in a microwave oven. Just like that from the atmosphere it does not undergo atomic absorption but resonant (some call it pseudo) scattering. But you try proving that it is absorbed and I will show you why it’s not, neither is much reflected. Discuss its action on water, ice, black metal, food or anything else.
What is your explanation of just exactly what valid physical mechanism leads to a greenhouse effect, for which there are several such explanations, all invalid? But, never-the-less, spell out your specific belief as to what you think the mechanism is and I will guarantee to be able to show you with standard physics where you are mistaken. Or you could just read my peer-reviewed paper first.
Springer,
It’s not related to evaporative cooling. Man-made structures like concrete, asphalt, tiles, G.I. sheet have higher absorptivity than grass, trees, soil, vegetation. That leads to higher temperature given the same incident solar radiation.
Wonderful post! McKitrick & Michaels Were Right: More Evidence of Spurious Warming in the IPCC Surface Temperature Dataset Roy Spencer, Ph. D. really makes my morning somewhat nicer 😀 Continue on alongside the fascinating articles! Regards, We have
Dr. Spencer, with respect, a follow up to my previous post re: Thermal Diffusivity,
Please note the following text from the website offering a Thermal Diffusivity Meter for sale commercially;
“AC TEMPERATURE RESPONSE“, their words, not mine!
As I have posited before, the climate science community has performed a rudimentary “DC” analysis of a system that is inherently an “AC” system. As a result, the Climate Science Community has produced an answer that is incorrect, i.e. “the GHE necessarily results in a higher equilibrium temperature”.
Also, from my link to measurements of Thermal Diffusivity please note the text;
“This immediately explains why it is so very difficult to measure thermal conductivity. Ideally this would require a steady state situation. This is far from easy because it usually requires a carefully planned laboratory experiment and a lot of time to get to an equilibrium.” Again, their words, not mine!
Unfortunately the whole concept of “Equilibrium Thermodynamics” only exists in textbooks and climate science models. It does not exist in reality, even in very carefully prepared laboratory experiments.
Cheers, Kevin.
Doug darlin, ice absorbs quite strongly in the IR, but has a minimum absorption where microwave ovens emit, 122 mm. This has everything to do with the atomic structure of the ice and nothing to do with your handwaving.
In simple words, just as CO2 molecules only absorb at some frequencies because of their structure, so do solids, although in solids one observes absorption bands rather than lines (as in the band structure of solids). Band structure, of course, is what gives us semiconductor technology.
You can put the dust back on that Nobel Prize
Hi, Neat post. There is a problem with your website in web explorer.