A sizable portion of the thermometers that make up the Global Historical Climatologically Network (GHCN) are located at airports. While much digital ink has been spent condemning the surface record for this fact, there have yet to be many substantive looks at what, if any, aggregate effect airport locations have on the surface record.
It seems like common sense that being located at an airport would lead to higher temperature readings than, say, a pristine wilderness. After all, airports tend to feature lots of pavement, which can radiate quite a bit of heat. However, measurement instruments are almost always located in grassy areas adjacent to runways, and its not clear how large the bias will be. Additionally, changing the location of a station to an airport (or building an airport around an existing station) will introduce a step-change in temperature readings, but apart from that step change there is little reason to expect that the trend in anomalies for the station will be different from non-airport stations.
Before we dive into examining the data, lets look at how many GHCN stations are located at airports, and how both the absolute number and percent has changed over time. Its important to note that the airport classification is based on a current snapshot in the metadata; there is no guarantee that stations currently classified as airports were located at airports in the past, though its probably safe to assume (with a few exceptions) that stations currently not located at airports were also not located at airports in the past. This assumption is important, because it allows us to compare non-airport stations to airport stations to tease out the relative effect of airport location.
The figure below shows the number of stations in each year that are airport rural, airport urban, non-airport rural, and non-airport urban. The urbanity designation is obtained by geolocating GHCN stations based on their lat/lon coordinates and using the GRUMP database to determine if the area in question is currently rural or urban (note that GRUMP uses a combination of high-resolution population density mapping and satellite nightlight brightness to designate areas as either urban or rural) via work by Ron Broberg.
Figure 1: GHCN station count by GRUMP urbanity and airport/non-airport metadata designation.
Here we see that the majority of GHCN stations are not located at airports (though a decent amount are), and that the percent of urban stations located at airports is considerably higher than that of rural station. The number of stations available has changed over time as well, with the large drop post-1992 due to the fact that GHCN has only been updated with CLIMAT-submitting GSN stations in the interim (note that this will change when GHCNv3 is released in the next year or so).
Figure 2: GHCN station percent by GRUMP urbanity and airport/non-airport metadata designation.
Figure 2 shows the percent of GHCN stations each year for each type. The percent of total stations located at airports has slowing increased from around 25 percent in the early part of the century to close to 40 percent in recent years.
To attempt to tease out the effect of airport location on the temperature trend, we can undertake a pair-wise comparison of airport and non-airport stations in the same 5×5 lat/lon grid cells. This prevents spatial coverage biases from sneaking in, as in many cases stations in more remote areas of the world will be located at airports. Unfortunately, it also shrinks the number of stations we have to work with down a bit, since not all grid cells have both airport and non-airport stations.
The temperature trend for each series is generated by calculating the anomalies for each station for the common reference period of 1961-1990. Stations are assigned to a 5×5 lat/lon grid based on their coordinates, and grid-cell anomalies are calculated by averaging individual station anomalies. Grid-cells are spatially-weighted based on there area to generate an estimated global land temperature. For more on the model used, see this post.
Figure 3, below, shows global land temperature reconstructions from airport and non-airport stations from grid cells containing both classes.
Figure 3: Temperature record from airport and non-airport stations in grid-cells containing both. Dotted lines show the number of stations available in each set.
Here we see only a small difference between the two records, with non-airport stations trending about 8 percent lower than airport stations for the 1900-2009 period and about 3 percent lower for the 1960-2009 period.
However, looking at all airports might not be appropriate, as some are located in dense urban areas and others in rural areas. We might expect urban airports to actually have a slightly lower trend than urban non-airport stations, simply because airport stations tend to be placed in open grassy areas. On the other hand, assuming rural stations tend to be better sited than urban stations, we would expect the airports to lead to a larger warm bias.
Figure 4: Temperature record from airport and non-airport urban stations in grid-cells containing both. Dotted lines show the number of stations available in each set.
For urban stations, we indeed find that airport stations warm slightly slower than non-airport stations, at least over the full period. The 1900-2009 trend for non-airport stations is about 5 percent higher than for airport stations. For 1960-2009 trends, however, non-airport stations are about 4 percent lower than airport stations. In both cases, the differences are fairly negligible.
Figure 5: Temperature record from airport and non-airport rural stations in grid-cells containing both. Dotted lines show the number of stations available in each set.
Rural stations show a much larger difference over the full period. Rural non-airports trends are a whopping 27 percent lower than airport stations, though it is worth noting that the sample size of rural airports sharing grid cells with rural non-airports is rather small, especially prior to 1945 and after 1992. Interestingly enough, rural non-airports actually trend 3 percent higher than rural airports in the 1960-2009 period.
Figure 6: Comparison of 1900-2009 trends for rural and urban airport and non-airport stations.
If we look at 1900-2009 trends and confidence intervals for both groups, the high trend in airports for rural stations is clearly visible.
Figure 7: Comparison of 1960-2009 trends for rural and urban airport and non-airport stations.
However, for the “modern warming period†post-1960, there appears to be little difference between the trends of airport and non-airport stations in either rural or urban areas. This is somewhat puzzling, since I’d imagine many of the airports were constructed post-1960. It might be possible that only pre-1960 airports experienced a case where an existing station was transplanted at an airport (or an airport grew around an existing station), where in most post-1960 cases new stations were added when airports were constructed, but unfortunately the GHCN station history data is not complete enough to make this anything but speculation.
Addendum
Note that David Jones over at Clear Climate Code undertook a similar analysis (that I came across after finishing my own) with somewhat different results, showing non-airport stations trending higher than airport stations over both the full century and the past few decades. This is most likely due to bias introduced by differing spatial coverage, as he did not restrict himself to the same gridcells for airport and non-airport stations. Indeed, if I run my model without the same-cell pair-wise comparisons, I get results similar to his, with non-airport stations warming 22% faster than airport stations over the 1900-2010 period and 9% faster over the 1960-2010 period.
Figure 8: Temperature record from all airport and all non-airport stations. Dotted lines show the number of stations available in each set.
Looks to me that nothing matters, unless your subset has so few stations that you lose signal/noise.
You said ‘pairwise’, but I assume that if a certain grid box had 2 stations from subset A and 4 stations from subset B, you use all of them for the comparison, and not just pick out one pair? Meaning, each grid box has at minimum a pair, but could have more?
If you’re looking for a growing divergence between two subsets, it might be more intuitive to show the residuals between the two, and then maybe fit a line through that.
Carrot,
Yep, by pair-wise I mean comparing grid-box averages for each criteria using the same group of grid-boxes. E.g. I might be comparing the anomalies of one airport to the averaged anomalies of 5 non-airports in the same grid cell, or vice-versa.
Also, divergence doesn’t appear to be growing; in most cases its shrinking. But good point about the residuals, I’ll include them in future posts.
Isn’t the main key to the airport criticism the step change (from non-airport to airport) you mentioned? The absolute view doesn’t really address this; more work to do right?
Nice work Zeke.
A couple of things:
Since the biggest differences are in the rural airport to non airport
How about this. ( just for clarity )
1. A chart ploting Non_Airport- Airport over time. I think this chart
The differences in approaches ( paired analysis versus not paired) is interesting. That makes me want to see the station list.
2. Ron had listing for other kinds of airports.. thots?
I think the other thing for people to note is that as folks like Carrick have suggested the horizontal fetch at an airport can make up for a lot of micro climate sins. Translated.. one of the causes of UHI is building height. Buildings do all sorts of nasty things to the boundary layer. At an airport, ask yourself are there buildings that disrupt the flow of air around the runways?
How are runways oriented when they are built relative to the dominant winds in the area? If the local winds can blow freely over the surface without being disrupted by tall local buildings then you have an area where excess heat is rapidly transported away. As Zeke notes you may very well get a step function up when the airport is first built ( changes in materials ) but having a long fetch is WAYYYY more important than a little concrete on the surface.
Anyways, Zeke It might be interesting to list up the stations used in your paired approach. The clear climate code guys have a peice of code that does Google earth files. If you post up the list of Rural aiports I’ll build a google earth file.
Finally. I’m still unclear by what you mean by pairwise. and how you figure the pairing.
For each 5 x 5 grid you check if there is
an airport site and non airport site? more than one?
Or do you find an airport site, then search for the closest Non airport site.. then check if they are in the same grid box?
This paring has never been clear to me. If you could be explicit about that it would help. There are a couple ways to do it.
1. Airport versus non airport: no considerations for spatial issues.
2. Airport versus non airport, by grid cell. Only use those grid cells that have both airport and non airport stations. ( I think you do this) there is a spatial problem here as well.
3. Station pairs. For each airport station, find the closest non airport station. Thats your pair. the weather doesnt care about 5×5 grids. think about an airport/non airport station that are 10 miles apart but on other sides of a grid line. .
And as far as station histories go.. I have zero interest in doing it, but for the US at least, you could probably go through station histories and see when each airport station was currently moved to the airport. Or at the very least, look up when any airport was built.
This maybe gets to my biggest complaint about the GHCN – if a station move is known, then it should be marked somehow in v2.mean. I’m not asking for tons of worldwide historical metadata, but at least, station moves can be logged so far as they’re known. If I was running my own database, I wouldn’t necessarily put some of these things into the same record with the same ID. But since the GHCN already did, you tend to just go with it.
Hmmm,
If the step change had a large effect on the trend, it would show up in the analysis such that airport stations had a higher trend than non-airport stations. If the step change isn’t significant enough to show up in the trends, its too small to worry much about on aggregate.
For rural airports pre-1960, we see something that might be significant. Post-1960 not so much. My guess is that there are few cases where airports grew around pre-existing stations; rather, stations were often constructed concomitant with airports. I have no way of proving this hunch, however.
Zeke, as you mention, the step change can go the other way, too. If you move a station from a warm spot in a city to a grassy spot at an airport 30 miles outside the city, maybe it goes the other way. You don’t know a priori
Nice work Zeke.
A couple of things:
Since the biggest differences are in the rural airport to non airport
How about this. ( just for clarity )
1. A chart ploting Non_Airport- Airport over time. I think this chart
The differences in approaches ( paired analysis versus not paired) is interesting. That makes me want to see the station list.
2. Ron had listing for other kinds of airports.. thots?
I think the other thing for people to note is that as folks like Carrick have suggested the horizontal fetch at an airport can make up for a lot of micro climate sins. Translated.. one of the causes of UHI is building height. Buildings do all sorts of nasty things to the boundary layer. At an airport, ask yourself are there buildings that disrupt the flow of air around the runways?
How are runways oriented when they are built relative to the dominant winds in the area? If the local winds can blow freely over the surface without being disrupted by tall local buildings then you have an area where excess heat is rapidly transported away. As Zeke notes you may very well get a step function up when the airport is first built ( changes in materials ) but having a long fetch is WAYYYY more important than a little concrete on the surface.
Anyways, Zeke It might be interesting to list up the stations used in your paired approach. The clear climate code guys have a peice of code that does Google earth files. If you post up the list of Rural aiports I’ll build a google earth file.
Finally. I’m still unclear by what you mean by pairwise. and how you figure the pairing.
For each 5 x 5 grid you check if there is
an airport site and non airport site? more than one?
Or do you find an airport site, then search for the closest Non airport site.. then check if they are in the same grid box?
This paring has never been clear to me. If you could be explicit about that it would help. There are a couple ways to do it.
1. Airport versus non airport: no considerations for spatial issues.
2. Airport versus non airport, by grid cell. Only use those grid cells that have both airport and non airport stations. ( I think you do this) there is a spatial problem here as well.
3. Station pairs. For each airport station, find the closest non airport station. Thats your pair. the weather doesnt care about 5×5 grids. think about an airport/non airport station that are 10 miles apart but on other sides of a grid line.
Gimme a little time and I can generate a list of rural ( by nightlights) airports and the closest rural non airport. .
Mosh,
I also see airport locations as probably being open to convection and radiation.
But look at that picture! Just look at the yellow arrows for the jet blast. Case closed, we don’t need to do any math.
Mosh,
By pair-wise comparison I mean doing an analysis of airport and non-airport stations separately, but limiting the analysis to only grid-cells that have both types.
E.g. for the rural airport case, I throw out all stations that are not in a 5×5 grid cell containing both a rural airport and rural non-airport station. I take the remaining stations and run only airport stations through the model, only non-airport stations, and compare the results.
The data used in the graphs is available here: http://drop.io/0yhqyon/asset/airports-blackboard-xls
I’ll work on a station list and a diff graph for rural airports.
Regarding David’s version at CCC, note that it was done by simply removing stations right at the beginning of the analysis (after step 0). So, for instance, the UHI adjustment is done on urban airports using nearby rural airports only. In other words, it was quick-and-dirty. One could do a more sophisticated analysis, but the quick-and-dirty approach strongly suggests that there is no point.
“Interestingly enough, rural non-airports actually trend 3 percent higher than rural airports in the 1960-2009 period.”
A rural airport may be a grass field sitting in the middle of a farm field that is tilled in the spring and harvested in the fall.
In the case of Antarctica the ‘rural airport’ is actually an ice sheet.
In the case of Barrow Alaska the ‘rural airport’ is 300,000+ sq ft of black pavement sitting in the middle of white snow and ice.
Mosh,
Rural airports – rural non-airports (in grid-boxes containing both):
http://i81.photobucket.com/albums/j237/hausfath/Picture420.png
Spreadsheet with specific station IDs: http://drop.io/0yhqyon/asset/mosh-airports-csv
So, no trend in the residuals back to WW2. Is that a step change before that, or just noise?
I wonder what the v2 mean adj set does here.
Zeke,
Congratulations on a very clear presentation.
I also did an airport analysis. For both a recent period (1979-2009) and the century period, I found airports had a slightly lower trend than land stations in general.
I looked in detail once at what airports in New Zealand were actually like – where situated, how busy etc. It’s pretty varied. Some are close to town, some not. The rural/urban distinction is of variable help – some urban airports are well out of town, and big cities often have several airports, some busy, some less so.
My favourite there was Campbell Island – uninhabited, way down in the Southern Ocean – only ever visited by people maintaining the weather station. But it’s listed as an airport.
Nick,
Neat. I must have missed the airport stuff when I read that post awhile back. I’m pretty sure the lower airport trend observed when looking at all airports vs. all non-airports is somewhat biased by spatial coverage. E.g. there may be a lower airport density in higher latitude grids, making the airport-only series under-sample faster-warming areas. That’s just a hunch though, and mapping the 5×5 grid cells covered by each might show coverage differences more clearly.
Thanks zeke, I’m prepapring for weekend away, so cant spend much time on this.
my thought was this. As carrot slyly notes and as Carrick and I have argued there is a strong theoretical case to make that would say that airports– over time– should be good locations ( in fact parkers UHI study uses mostly airport– and he looks at temps as a function of local winds ) anyways, if a bias grows more strongly in non airport than airport ( The trend of non-airport – airport being positive) THEN there is argument that says the adjustments that should be done are NOT UHI adjustments according to population, but rather according to ‘airport’ That is, nick could
take his emulation of Giss Urban adjust, and see what happens when you adjust the higher trending Rural Non airports to the lower trending rural airports.
In a rush, sorry if it not more clear.
A look at your rural non airport – airport suggests a change point, as CE has intimated this may have something to due with increased air travel.. etc.
Anyway. airports are probably not as bad as some have thought. Pictures of tarmacs dont capture the meso scale issues which are decidely IN FAVOR of using airports.
Just conceptually, I’m trying to get to the bottom of petersons mystery. why dont urban warm more than rural.
he argued: 1 urban was well sited. I think its more complicated than that. primarily I think that folks have concentrated too much on the population aspect of UHI and not enough on the physical geometry of the problem.. Hmm, when I get back I’ll have thought some more.. might be good to read parker again.
2836,1065
2869,3100
3026,3952
3965,3957
8314,3970
8397,3974
11782,3976
14307,3980
16045,8306
16088,8495
I have to assume these are not ghcnids or wmo numbers
Mosh,
Having a bit of an argument with Anthony over the “rawness” of GHCN v2.mean. Specifically, he thinks that it is adjusted for TOBS and FILNET (might be confusing it with USHCN v2.mean F52 series?).
Wonder if you could lend a hand in pointing out that v2.mean is indeed the rawest temp data available (e.g. comes directly from CLIMAT reports with minimal QA, at least for the last few decades).
Mosh,
2836,1065 is two columns in a csv file:
wmo_id 02836 for the rural airport set, and wmo_id 01065 for the rural non-airport set.
Excel truncates leading zeros, unfortunately.
Hi Zeke – another great analysis, thanks!
Can you explain what “airport” meant in 1900? I mean, 25% is a good fraction of the stations, but the Wright Brothers didn’t fly until 1903!
Arthur,
Airport is a metadata designation associated with each station, reflecting a snapshot in time (which, I believe, was compiled in the late 90s). So in this particular case, a 1900 record from an “airport” station is the temperature reading from a station that was presumably not at an airport in 1900 but was subsequently either moved to an airport or had an airport constructed around it.
.
The idea behind this comparison is that any bias due to the measuring of temperatures at airports (mostly due to the step-change caused by moving a station to an airport, or having an airport built around an existing station) will be reflected in the trend of the airports series.
My concern about the use of airport data is that in my experience the airports at Albany, Syracuse and Rochester were orginally rural but in the past 40 or so years the area around them as been built up as suburbs developed around them.
For example, the Syracuse airport was orginally built in WW2 and was pretty much wide open. By the time I moved to Syracuse in 1981 there were some local developments but and the land on the airport proper west of the temperature instrument was still not developed much. Since then a rental car parking lot was built, the airport parking area was expanded, and buildings were added with parking lots all on the airport grounds.
I strongly suspect that development would add a urban heating bias to any trend for this airport. How much of a bias would be a difficult thing to tease out. Would doing this kind of pair-wise analysis for specific airports and nearby rural sites on a smaller scale than by grid cell be useful?
Carrot/Mosh/Nick,
Any of you guys know where to obtain a copy of the USHCN v1 raw.avg file? All they have on the ftp site now is v2…
RogerCNY,
It probably does have some effect, which is why I control for urbanity in Figures 4 and 5. Though I suspect UHI is less significant at urban airports than urban non-airports.
Funny you ask that Zeke, I was looking for the exact same thing before I came over here… I suspect you may have to email your friends at NOAA for old USHCN v1.0 files, or ask around if anybody has something saved on their computer.
Carrot,
Yeh, google searches aren’t turning up anything.
I strongly suspect that for USHCN, v1 raw.avg and v2 raw.avg contain the exact same data, and that Anthony’s assumption that USHCN v2 adjusted the “raw” data is based on a comparison of v1 F52.avg to v2 F52.avg (since the major change between USHCN v1 and v2 was the adjustment procedure). Actually need to dig up the data to check though.
Agreed on all counts.
If you can track it down, there should be four files.
Though I should add – correcting a random typo/data entry error here and there doesn’t count as an adjustment. It’s just getting the raw data in right. That’ll happen here and there.
Arthur is correct, there is a timing problem here. In the US (at least), before 1948, most of the stations at airports now were in town at places like the window of the local custom house etc… NCDC’s favorite record to display on the v2 page is Reno, NV where mean temps dropped 2 deg when moved to the airport from town, then the temps slowly regained that 2 deg as the airport and surrounding area was developed
http://www.ncdc.noaa.gov/oa/climate/research/ushcn/
graphs down twards the bottom of the page
Zeke (#44227) – thanks, of course that explains it. Is there any data on what fraction were moved vs just had an airport grow around them?
If I had to guess, I’d say very few stations randomly had an airport sprout up next to them.
but a bunch were actively moved from city centers to airports at some point.
But for the US, this can be sorted from online station histories. if somebody has a ton of motivation for grunt work..
Zeke, you probably know this already, but your lead-off photo is a famous picture of how to “lie” with photographs.
This site explains how the temperature measurement station is located in a grassy strip adjacent to an airplane “parking lot” used rarely (only when air traffic is very high, and they need to keep some planes out of the take-off queue).
Maybe we should correct some terminology here… airfield rather than airport… there are many stations at small airfields… although the airports tend to have ASOS
And Here is the link that shows just how carefully this photo was cropped.
Paulk2, Intresting how Mr John Graham doesn’t know what he is talking about but provides the evidence which he is disputing.
One of the fundamental building blocks (which I recently learned that even some of those with the most potential do not understand) of temperature monitoring as used by all of the AGW studies, is that the high and low temperatures of the day are what is averaged then reported.
John uses the first photo to show what has been said… then he uses the second photo to say that the area is a parking lot, implying that the jet engines do not have an impact on the thermometers (of course not, they’re parked…AH HAAAA!). But what he did not say, even after being told by a commenor on his blog, is that the photo’s show two different planes. The conclusion is that the planes there in fact do move, probably on their own power… and if during that few minutes when the jet engines are on, blowing hot air in the direction of the station, then their jet wash will bias the high temperature of the day.
Zeke,
go back to that WUWT topic. You’ll find that EM Smith has popped up, if you have an appetite for that. Something about duplicate #3 seems to be his new thing.
Okay, I screwed that first sentence up… it was meant to say that he provides the evidence which disputes his point
mike, what you meant to say was pretty easily discerned.
see how smart carrot is? The darned veggie consumer comprendes my screw ups!
But it would be unfair to conclude that the problem with airports and airfields is that the stations get bombarded with jet wash, the problem is with the concrete and such which absorb heat and slowly release it.
But I will disagree with mosher re: the idea that the buildings block the wind and therefore affect the temps… I have yet to see a modern airport/field where distance of the structure or tree to the station is less than 4X the height of the structure or tree (that is a wind issue).
It’s all the vitamin A.
Zeke, was it you or Dr. Spencer that was suggesting earlier that rural areas seemed to warm up “quicker” with advancing population? Sorts like CO2, it seems to take a doubling to create more heat. 🙂
Could this in any way have an effect on the earlier “divirgence”?
Having some trouble grasping the argument. Is it:
— There is no difference between absolute recorded temperatures at airport and non-airport sites? Ie airports are not hotter than rurals?
— if this is true, is the argument then that if we take a given population, and swap out rural and airport sites in it, it will make no difference? Since they are both recording the same temperatures, it doesn’t matter how many sites are airport and how many rurals, and it does not matter if the proportions stay the same over a given period?
— Or is it that there is no difference in anomalies over a given period if we compare a population of airport sites with a population which are rural sites? That is, airport sites and rural sites may differ in absolute temperatures, but if we compare a given rural and a given airport series over time, the rise or fall will be identical. Airports will warm or cool at the same rate as rurals?
— if this is the argument, does it leave open the possibility that if there is an absolute difference in airport temps compared to rural temps, if one varies the proportions, it might change absolute temperatures, and this would in turn change anomalies?
OMG.
MikeC… don’t what you are talking about. Here is another picture of the site, showing how the planes are parked in front of a hangar, with what appears to be refueling connections. You can even see the stairs pulled up to one plane. The planes are parked; not running. Parked.
Hangar at Rome
Finally the Stevenson screens are a long way from the planes; even with a slight breeze, the hot exhaust won’t reach the boxes. If this was the end of a runway, and the engines were running full blast (as this photo was “sold” by the propagandist who cropped the photo), then maybe you might have a point. But this information was so twisted out of context and doctored up by the opportunistic liars who used this photo to push their POV, it is really clear how much we can’t trust these sources.
Key Measurement: Look at how much you twist even one photo to mean something not consistent with reality, as an indicator to show how much you are in denial of the facts.
Bugs,
The closet’s quiet, dark, and deep,
but I have promises to keep,
and miles to go before I sleep,
and miles to go before I sleep.
A question that comes to mind is this:
If the temperature data is as stable in the face of large influences as seems to be implied in this study, is this method of data gathering actually showing anything of significance?
Off topic, but if anyone is interested in the real time feed from BP at the site of the blowout, here is a good link:
http://www.bp.com/liveassets/bp_internet/globalbp/globalbp_uk_english/homepage/STAGING/local_assets/bp_homepage/html/rov_stream.html#
MikeC: In fact, if you rotate the photo, you can see the jet engines on one of the planes are covered by what appears to be brownish-red covers. Hmmm, how much hot air is coming out of a jet engine that has the intake canvassed over?
Answer: Less than the hot air that came out of the big mouths who pushed this nonsense onto their rather ignorant and (too) trusting audience.
My problem is that the heat off of a tarmac is intense, and with the simple survey that Warwick Hughes performed with a temperature data logger with a tiny town (pop. 227): http://www.warwickhughes.com/blog/?p=575
how does one find the signal among this type noise/bias?
intrepid_wanders: What happens at night? How about at night, with a cloudless sky? What happens when a breeze is blowing? What happens in the winter?
The fact is that no one has done an intensive study to show the impact of tarmac near a temperature monitor over the course of a year, compared to a similar monitor mounted nearby in a field. And it isn’t clear at all, what the net effect will be over the course of a year. But the point is, that year after year, these impacts will likely even out, so that a trend anomaly will likely be accurate.
I checked out the information Warwick Hughes collected. He DROVE through town taking temperature readings at around 9:30 AM in late May, with no wind…
How do you know what the nighttime temperature is? How about the exhaust from vehicles on the road? How many measurements were made when the car was in motion versus stationary? How about shady and sunny spots on the street? I would expect at least a one degree C difference just by turning off a sunny street onto shaded street only a couple of hours after sunrise.
But temperature measurement stations aren’t driving around. And the temperature records show the highest temperature anomalies in winter at night! And the highest anomalies are in places like northern Canada, Greenland, and Siberia. Where are the teeming cities in those locations? This whole UHI causes rising global temperature anomalies myth is shot full of so many logical holes, it is shocking any intelligent person still believes it.
Check out Peter Sinclair’s short video on the UHI effect on global temperature anomaly myth.
Here is the archive of all of Sinclair’s videos… These are well worth watching and learning from, especially if you have sceptical views.
re: bugs (Comment#44251)
Not just Zeke. He back tracked some college IP and scolded the poster. And I could have sworn I saw a Timpaganos post there earlier. So may be a post delete there too. Watt’s is feeling a little defensive today.
basically a single look at one f the satellite data maps should end every discussion about UHI effect and airports at once.
.
http://www.remss.com/msu/msu_data_monthly.html
.
satellite data is not affected by UHI, but gives extremely similar results.
.
—————–
.
the “plane engine” claim is extremely similar to the “barbecue under stevenson screen” one: even a few minutes of thinking about it, should tell you that the effect will be tiny.
.
IF serious heat would reach the thermometer, it would show absolutely abnormal spikes. those spikes would show up in min/max readings of the day, all the time. and those spikes would be the same in winter as in summer.
.
the temperature reading is important for pilots. the effect of such spikes on them, would be much bigger than on the climate record!
Well bugs, as usual, you added value to a great discussion.
here is the story that anthony wrote on the subject:.
.
http://wattsupwiththat.com/2010/02/14/christy-and-mckittrick-in-the-uk-times-doubts-on-station-data/
.
and here the “jet blast” video that i have seen posted, in this context, by denialists multiple times:
.
http://www.youtube.com/watch?v=DFP4xl0V0mk
.
Mosh,
Having a bit of an argument with Anthony over the “rawness†of GHCN v2.mean. Specifically, he thinks that it is adjusted for TOBS and FILNET (might be confusing it with USHCN v2.mean F52 series?).
Wonder if you could lend a hand in pointing out that v2.mean is indeed the rawest temp data available (e.g. comes directly from CLIMAT reports with minimal QA, at least for the last few decades).
Sorry been out all day. point me too a link where I can clarify.
Out of town and off line for a few days so tommorrow AM
Mosher, it’s in the topic about a CEI lawsuit
http://wattsupwiththat.com/2010/05/27/cei-files-suit-on-giss-regarding-foia-delays/
discussion prompted by Chris Horner saying things that don’t make any sense.
Carrot/Mosh/Nick,
Any of you guys know where to obtain a copy of the USHCN v1 raw.avg file? All they have on the ftp site now is v2…
I thought i saw it the other day
USHCN v1 is what I was talking about at Comment#44231
I haven’t got anything saved, nor can I find it on the website.
I cant find it. It was in the intermediatries file.
BUT, please pass this onto zeke and others, especially Ron
http://www.ncdc.noaa.gov/oa/climate/…/article3abstract.pdf
peterson and owen 2005. Some very kool metadata. I gotta sleep man.. early trip tommorrow
US 1km population back to 1930
shit busted link
http://journals.ametsoc.org/doi/abs/10.1175/JCLI3431.1
Anybody who wants to look at jetwash.. just get the hourly data. The probability of jet wash corrupting Tmin or Tmin is vanishingly small.
WRT the changes in material properties ( concrete) those effects are Tmin issues. They are mitiagted or modulated by the absence of high structures in the area that would disrupt the boundary layer.
I always get a bit peeved at this because when all this stuff came up nobody wanted to go back and actually read Oke and others.
lets see. If you think ( like peterson, parker, jones) that the UHI signal is small or non existent, then that’s a mystery of sorts as Peterson notes. Now, peterson thought he solved the mystery. I think he didnt. How to put this.
Peterson noted no difference between urban and rural. he concluded it was due to good siting.
Watts and others say the siting was bad, and conclude there will be a UHI/Microsite signal in the mix.
Folks miss the 3rd possiblity: peterson was right, there is no signal ( or its small) but the reason ISNT good siting in urban locations.
CE does that position make sense to you? i mean just logically, not whether u buy the 3rd way
Just a notion on moving aircraft on an airfield re the photo on top:
Commercial aircraft usually taxi under their own power only for takeoff and after landing to terminal. Otherwise, tow tractors are used as much as possible. Saves fuel. Just to spool up a big jet engine takes more fuel than a tractor uses for the whole chore.
This is cool Zeke. thanks! (been out and away from computer)
I have sort of the same questions as
michel (Comment#44250) May 27th, 2010 at 9:07 pm
and hunter (Comment#44254) May 27th, 2010 at 10:04 pm
I keep thinking of my airport (in the middle of urban sprawl) plus the micro climate (temperature gradients because of the coast) (sometimes 8-10 degree higher temperature differences between my house and there). Micro climates are every where and they aren’t necessarily “unnatural” either.
“The temperature trend for each series is generated by calculating the anomalies for each station for the common reference period of 1961-1990… ”
Something about this paragraph bugs me.
Who or what decides this reference period? (That reference period for here where I live would see a heck of a lot of changes to the landscape, and even more changes after 1990)
(I will read the link to the model used too)
MikeC (Comment#44247) May 27th, 2010 at 8:44 pm
We have a giant outdoor shopping mall with movie theater here. It’s designed to look like a giant mediterranean villa. It’s plopped in the middle of a vast parking lot that also has restaurants and banks etc situated around the parameter of it too. I swear all that heat coming from the asphalt and the way those buildings are situated, I think this creates wind. At least for me; I always remember to bring a sweater when I go there. It could be calm here at my house but when I go there I remember to bring a sweater. (the temperature difference is chillier) It’s about a mile from my house. I’ll have to pay attention to what it is like this summer.
Who or what decides this reference period? (That reference period for here where I live would see a heck of a lot of changes to the landscape, and even more changes after 1990)
It’s arbitrary. You just select a period and go. I vaguely recall that one of the hacked emails has Jones talking about moving his reference period forward, but was afraid it would confuse people and open CRU up to more claims of data manipulation.
Over at WUWT, EM Smith is decrying the GISS reference period. Apparently he doesn’t realize that, since he has the code and runs it, he could just go in and change the reference to whatever he wanted. Or he does realize it but hopes no one else is smart enough to figure it out. But he’s wrong – I’m the slow kid in the class and I figured it out. 😉
On the NASA GISTEMP website, their web
page lets you choose whatever baseline you want to display.
http://data.giss.nasa.gov/gistemp/station_data/
Baselines are ‘arbitrary.’ You just pick a period and go. You want it long enough to provide a good ‘average.’ You want it recent enough to have lots of stations in it.
I vaguely recall that one of the purloined emails had Jones discussing moving CRU’s baseline forward but that he was afraid that moving it would confuse people and provoke calls of data manipulation???
NASA GISS lets you change the displayed baseline on its web page presentations:
http://data.giss.nasa.gov/gistemp/station_data/
This is the third time I’ve tried posting this message. Something wrong with the website …?
.
Baselines are arbitrary. You just pick a period and go. You want it long enough to provide a good average. You want it recent enough to include lots of stations.
.
EM Smith is bemoaning GISS’s selection of a baseline over at the WUWT thread. Even though he has the GISTEMP code, apparently he doesn’t realize he could just change the baseline himself and present a new version. I realize it and I’m though slow kid in the class. Maybe he does realize it and just doesn’t want to quantify the change it will make to the trend (hint: the trend will not change at all). It’s all about ‘talking points’ and nothing about ‘data analysis’ over there.
.
You can experiment with changing the baseline yourself on the GISS GISTEMP web site. Note that this does not display the trend – it displays the difference from whatever baseline you choose.
http://data.giss.nasa.gov/gistemp/maps/
This is the fifth time I’ve tried posting this message. Something wrong with the website …?
.
Baselines are arbitrary. You just pick a period and go. You want it long enough to provide a good average. You want it recent enough to include lots of stations.
.
EM Smith is bemoaning GISS’s selection of a baseline over at the WUWT thread. Even though he has the GISTEMP code, apparently he doesn’t realize he could just change the baseline himself and present a new version. I realize it and I’m the slow kid in the class. Maybe he does realize it and just doesn’t want to quantify the change it will make to the trend (hint: the trend will not change at all). It’s all about ‘talking points’ and nothing about ‘data analysis’ over there.
.
You can experiment with changing the baseline yourself on the GISS GISTEMP web site. Note that this does not display the trend – it displays the difference from whatever baseline you choose.
http://data.giss.nasa.gov/gistemp/maps/
Liza the reference period is arbitrary. If you were to give me an anomalized time series centered on 1960-1991, I could convert it to one which is centered on your birthdate to birthdate + 10 for example, just by subtracting the average value of that anomaly over that period from the original series.
Regarding microsite effects, we could represent it this way:
T_meas = (1 + bias) * T_true + T_offset + T_noise
when you convert the temperature into an anomaly you get multiple sources of error: One is from the bias of the microsite, another is from changes in T_offset.
T_noise is noise associated with the site selection (weather plus microsite). Averages over time plus over the surface of the Earth substantially reduce the magnitude of this.
This is something you and others need to come to grips with: If I start with ±2°C errors in daily measurement for an individual site, what is the effect of a 30-day average on this? What is the affect of averaging over sites that are widely spaced (so the noise isn’t correlated?).
Change in T_offset affect the anomaly value but the magnitude and frequency of this T_offset change is bounded. This T_offset can be positive (UHI) but very low frequency (cities grow slowly) or negative and very high frequency (station move=step function). Moreover, this episodic noise is in general out of sync with that at other stations.
Average over e.g. 30 days and over all of the stations on the Earth and you can set a bounds on the error from offset shifts (it turns out to be really small, contributing perhaps a ±3% error to the monthly-averaged global temperature anomaly).
The one that matters the most is the “bias” error.. Good metadata helps us test for the importance of this effect, and on global mean anomaly the net affect seems to be small for everything except maybe latitude effects, and marine versus land in general.
Ron EM knows that. His issue is rather arcane. kinda like the selection of colors on charts
EM does raise an issue that I haven’t really had a chance to look into much; namely, the duplicate flag in GHCN data.
Right now I treat them the same way I treat imods; they all get combined into a single wmo_id record prior to processing.
Anyone know what exactly the duplicate/imod distinction is, and why all duplicates aren’t, well, exact duplicates of each other?
Is it a given data red flag that as 5000 or so stations are removed, starting around 1980, that the trend line for temperature seems to mirror image the decline in station numbers?
I didn’t see a reference to this above, so I’m assuming it has been discussed and resolved in the past.
(I got an error message trying to submit this comment earlier…. please ignore this one if it’s a duplicate)
Is it a given data red flag that as 5000 or so stations are removed, starting in 1980, the temperature trend follows in the reverse, the decline in station numbers?
I didn’t see this mentioned in the above discussion, so I’m assuming that it has been resolved. Can someone point me to a good explanation? Thx,
Dan
DanC: see http://rankexploits.com/musings/2010/timeline-of-the-march-of-the-thermometers-meme/
Zeke,
Different imods should not be combined. When the imod is 000, it means that the data is for the station that the WMO number is assigned to. When the imod is not 000, it means that the data is for a different station, and the WMO number used is the closest WMO station. You wouldn’t combine data for 42572290000 SAN DIEGO and 42572295000 LOS ANGELES because it is obvious they are different stations, but you should also not combine 42572469000 DENVER and 42572469004 BERTHOUD PASS, even though they use the same WMO number – the difference is in the imod.
The duplicate number indicates different timeseries of data for the same station. Duplicate 0 is the longest timeseries, duplicate 1 is the next longest timeseries, and so on. The difference in the data values is probably usually due to the different sources, but in some cases such as 42572469000 DENVER, the difference comes from a station move. If all of the different duplicate numbers are due to different sources, the values are probably all similar enough that averaging is okay. But in a case like Denver, averaging will give you numbers that really are not good.
Carrick,
you said: “(cities grow slowly)”.
Not around here!
Especially in that time frame 60’s-90s.
No trying to be contrary – I just notice the assumptions.
“Liza the reference period is arbitrary.” I see.
These are tiny timescales in regards to time on this planet and we can’t do this for older time periods on earth-pick arbitrary time lots that small and “look at them” and say what is “unusual”or not because there isn’t that kind of data resolution. I think that is the part I can’t “get over”. So sue me. 😉
This may be a dumb question. Is data sorted / adjusted for altitude “all over the Earth” ?
torn8o,
Fair enough, I agree in retrospect that treating imods as separate station records might be a better approach than averaging the data pre-anomaly-generation.
There are a few different methods worth testing:
1) The old method of averaging absolute temps from all duplicates and imods into a single wmo_id record.
2) Averaging absolute temps of duplicates into imods, calculating anomalies, and averaging imod anomalies into a single wmo_id record.
3) Averaging absolute temps of duplicates into imods, calculating anomalies, but treating each imod like a separate station.
The results for all three are shown in the chart below:
http://i81.photobucket.com/albums/j237/hausfath/Picture425.png
I suspect 2) is ideal, but it doesn’t really matter much either way.
Zeke
there is a difference between imods and duplicates.
I’ve posted on this in the past. sorry no links. and I’m out the door.
There is a manual on it: linked on CA in a thread on hansen step 1.
see the threads on airvent or ask nick.
in short. GHCN gets a bunch of records with the same WMO.
12345. up to 10 records for one site. They then try to determine if these records are ACTUALLY duplicates. if they think they are they delete them. If not they preserve them
In the end you have a record that looks like this
42512345000X: X is the duplicate indicator.
so
425123450000: 23 24 25 26 27 27 28 28 NA NA 12 23
425123450001: 23 NA 25 28 27 23 28 30 NA 16 12 21
425123450003: 23 20 25 28 27 23 28 30 NA 16 12 21
Are 3 records from the same WMO that GHCN preserves.
Duplicate 2 has been dropped because its a duplicate of some
other record. there are many reasons for these “duplicate” records. like two diffrent instruments at the same location.
two scribal records of the SAME instrument with transcription errors.
How to handle them:
Average: dup 0,1,3: this is what most do. it overweights TRUE
duplicates.
Median:
Midpoint of the range: ( max+ min)/2
Reference method: thi is what hansen does in step one.
The differences in these ways of handling them contributes a very minor uncertainty to the final answer. When I get back from
vacation I will post you a file that calculates the various ways of combining duplicates. All except Hansens. the differences are for the most part in the 1/100.
Please see the threads on CA that describe the “scribal methods”
Also, Roman has decided to have a look at it after I raised some issues about it on the Airvent. Ask him
start here:
http://climateaudit.org/2007/09/28/russian-bias/
http://climateaudit.org/2007/09/10/the-bias-methods-perfect-siberian-storm/
http://climateaudit.org/2007/09/07/hansen-and-the-ghcn-manual/
GHCN manual
http://www1.ncdc.noaa.gov/pub/data/documentlibrary/tddoc/td9100.pdf
Zeke,
The main difference I see between 2) and 3) is that you might be moving some anomalies out of one grid cell and into another in 2) that would not happen in 3). Otherwise you are right that there is not much difference. However, my opinion is that it is best to use the cleanest method, rather than the argument that it doesn’t matter much. That increases confidence in the results and eliminates an arguing point. You can see that the long-term trend was reduced between method 1) and methods 2) and 3). Whether it is a lot or a little is not the point.
torn8o,
Sorry, I meant that the distinction between 2) and 3) doesn’t matter much. Both are significantly different than 1), however, decreasing the trend a bit over the century and increasing it a tiny bit over the last few decades.
Zeke
there is a difference between imods and duplicates.
I’ve posted on this in the past. sorry no links. and I’m out the door.
There is a manual on it: linked on CA in a thread on hansen step 1.
see the threads on airvent or ask nick.
in short. GHCN gets a bunch of records with the same WMO.
12345. up to 10 records for one site. They then try to determine if these records are ACTUALLY duplicates. if they think they are they delete them. If not they preserve them
In the end you have a record that looks like this
42512345000X: X is the duplicate indicator.
so
425123450000: 23 24 25 26 27 27 28 28 NA NA 12 23
425123450001: 23 NA 25 28 27 23 28 30 NA 16 12 21
425123450003: 23 20 25 28 27 23 28 30 NA 16 12 21
Are 3 records from the same WMO that GHCN preserves.
Duplicate 2 has been dropped because its a duplicate of some
other record. there are many reasons for these “duplicate” records. like two diffrent instruments at the same location.
two scribal records of the SAME instrument with transcription errors.
How to handle them:
Average: dup 0,1,3: this is what most do. it overweights TRUE
duplicates.
Median:
Midpoint of the range: ( max+ min)/2
Reference method: thi is what hansen does in step one.
The differences in these ways of handling them contributes a very minor uncertainty to the final answer. When I get back from
vacation I will post you a file that calculates the various ways of combining duplicates. All except Hansens. the differences are for the most part in the 1/100.
Please see the threads on CA that describe the “scribal methods”
Also, Roman has decided to have a look at it after I raised some issues about it on the Airvent. Ask him
start here:
http://climateaudit.org/2007/09/28/russian-bias/
http://climateaudit.org/2007/09/10/the-bias-methods-perfect-siberian-storm/
http://climateaudit.org/2007/09/07/hansen-and-the-ghcn-manual/
GHCN manual
http://www1.ncdc.noaa.gov/pub/data/documentlibrary/tddoc/td9100.pdf
The point about handling duplicates in my mind is that it is a decision with attendant uncertainty. Small uncertainty.
For example: if you have 3 duplicates all reporting for July 1943:
22C 22C and 25C. What do you do? average them? this is a tactic assumption that they are different records and that the two 22C are not records from the same instrument with different scribal versions. My first thought was to take the mid point of 22 and 25
Roman suggest the median. Nick uses the average. Hansen uses the reference method. Neat small problem that has nothing to do with global warming.
yes Zeke it doesnt matter much. Its just a few hundreths added to the uncertainty in monthly temps.
The average station has 1.89 duplicates. the vast majority only have two duplicates. some have up to 10. in these cases
Carrick (Comment#44280) May 28th, 2010 at 7:31 am
I’ve tried to submit a reply a couple of times too. I will make this one short for now.
“Cities grow slowly”. What is the basis of that?? Not trying to be contrary, I just tend to notice assumptions.
Paul K2,
My point was 2 fold. First, those parked planes were not in park with their engines shut off when rolling to where they are. I do not see any sort of cart or dolly attached to the front of the planes so I assume they got there on their own power. Okay? They were definitely parked when the picture was taken, but they got there somehow… probably by the thrust of those jet engines.
Second, Daily mean temps are a mean of the high and low temperature of the day. So when those jets were turned on and parking, then the jet wash (which I understand is pretty hot) can reach the station and be the source of the heat which was recorded as the daily high. The daily high temperature only takes a moment to record. Put a flame under it for a second and that will raise up the temperature.
Some other observations… there is one CRS (Cotton Regional Shelter or Stevenson Screen) at that location. A few feet to the left of the CRS is an MMTS, not an ASOS as John claims, and there are not 2 Stevenson Screens as John claims, the other square object measures rainfall. To the left of the 2 story building which is to the left of the station is what looks like a HGO (the thermometers used in ASOS stations.)
Next is your claim that the jet does not have enough thrust during taxi to reach those instruments. But then to be fair, I’m not an aerospace engineer so I wouldn’t know.
However, the point made in the Sunday Times article seems to suggest that John did not understand what was written.
“Watts has also found examples overseas, such as the weather station at Rome airport, which catches the hot exhaust fumes emitted by taxiing jets.”
It says taxiing jets, not taking off jets. You seem to be the only one saying taking off jets.
However, if one of those jets did taxi to that parking area at about an hour before sunrise (generally the coolest part of the day), then the heat from that jetwash would combine with the heat from the runway, two story building, parking lot and etc to create warmer temperatures than if none of those objects were not there (as was certainly the case 100 years ago)
Doesn’t matter much, but I also think different imods and duplicates shouldn’t be treated the same way. use the inv file to see how different the locations of the imods can be. but duplicates ought to be nominally the same location, though there can be a station move difference in there. Mainly you get duplicates because NOAA got data from 30 different sources, and they don’t always agree 100% for the same station.
It’s said this is related to different ways of calculating Tmean; see Peterson 1997. I don’t know if that’s the best explanation, but it’s an explanation.
This is my pet peeve with GHCN – the way they set it up, it funnels everybody into treating all these different records as essentially being duplicates. Some are, some aren’t.
As for EM Smith, I’m not at all convinced that he understands the point about baselines. He’s written all sorts of odd things on this topic, and sat there and played with the baseline on the GISS webpage mapper, to find out.. who knows what.
And ron, you don’t need anybody’s code to change the display baseline in GISS. Just a numerical listing of the output and a spreadsheet would do. So it’s that much easier… Now, in the case of somebody using the CAM, like Zeke, then you have to go back into the original calculation to see any effects of a baseline. But the RSM has no true baseline, in that respect.
GISS really should feature the trend maps more heavily, so people know they’re there, because a lot of the time, that’s what people are actually looking for, whether they realise it or not.
Paul K2,
“This whole UHI causes rising global temperature anomalies myth is shot full of so many logical holes, it is shocking any intelligent person still believes it.”
Given that UHI is discussed in every meteorology textbook, has been shown to exist in hundreds of scientific studies, is the cause of adjustments in pretty much all of the data sets including Hansen and Jones… I’d say the logical holes are the ones created by those who do not know what they are talking about.
Strawman, mike. UHI (or at least, some combination of complicated meso and micro scale effects) is real. But just because something exists doesn’t mean it has that much weight on the final results.
carrot, since the final results are .6 degrees that is believed to be the background GHG signal, then very small numbers become more than just “not that much.”
Liza:
On the scale of weather noise it is slow.
Always good to explain the reference frame.
Liza:
On the scale of weather noise it is slow.
It’s always good to explain the reference frame.
Zeke – sounds good, that was my misunderstanding.
carrot – I agree about the GHCN duplicate issue. I wish they would have preserved the source as a code or something instead. I don’t think GCHN v2 is junk by any means – any dataset is going to have some warts. It’s more about how easy it is to figure out what is a wart so that it can be dealt with. I hope GHCN v3 does it better.
MikeC:
There is a right answer here: Either UHI is important or it’s not. The right answer doesn’t change by you or anybody else debating the validity of the studies.
This has been tested empirically a number of different ways, and what has been found is, wrt to global mean temperature trends the effect of UHI is small. My own estimate is, over a 50-year period, the effect of UHI bias is less than 3% of the total trend on global mean temperature (or roughly 10% of global land temperature).
If you think there is a flaw in the studies please have at it to point at the holes in them. But seriously, this vague hand waving is just a waste of everybody’s time. Nobody is going to bank their work on anybody’s intuition of orders of magnitude.
That’s what we have math and quantitative analysis for, and every quantitate study done says you’re wrong.
Zeke,
Underlying all this discussion is an assumption (possibly by others) that the thermal characteristics of airports are sufficiently different from other places that data series compiled from airport measurements will be biased. And that this bias will be high or low and consistent from airport to airport (at least by sign) so as to have an effect on derived “global temperatures.”
Airports have a lot of open space, are usually located in areas that are relatively flat, have low building footprint (square feet per acre), and that usually of relatively light construction (hangars and service buildings), and lots of paving.
The paving at major airports can be concrete to thicknesses of 48 inches. This sounds like a big heat sink which will suck-up heat during the day and then radiate it away during the night.
But if the night was windless, the radiation, which likely wouldn’t last through the night, might create convection which draws in air from the surrounding neighborhood which might be cooler than that at the airport. And this is what would be measured, not the air above the runway.
What if the thermometer location were most often upwind from the runways and ramps?
I would think that before trying to divine (tease out) the distinctive characteristics of airport temperature measurements from the un-adjusted GHCNs, it might be more revealing to see if the low temperature trends are different at airports than rural stations. These would show the effects of radiating heat sinks if there were any.
I think I understand the usefulness of the anomaly method of temperature change analysis, but if there is a pattern in the raw temperatures, such as i suggest above, that would not show up in the anomalies, are the anomalies still reliable indicators?
I am very impressed by the methodical analyses you and your colleagues have been doing to look at each aspect of GISTEMP (If I have the name right) and that aspect’s possible contribution of bias.
You, Chad, JeffID, and RomanC and our hostess can certainly not be adequately thanked for showing that what can be plumbed from GHCN and the other series comes out similarly even if you run the pipes in different patterns.
Now on to similarly qualifying the GHCN data sets.
BTW – the dupe -v- imod distinction is new to me as well.
I appreciate the clarification, torn8o, steven.
carrick, that sure is a lot of finger pointing, scolding, and yapping for someone who hasn’t read the material and as of two days ago, didn’t even understand how or what temperatures are reported.
“This has been tested empirically a number of different ways, and what has been found is, wrt to global mean temperature trends the effect of UHI is small.”
Carrick, MikeC’s question is like mine about: “Cities grow slowlyâ€.
I asked: What is the basis of that??
You answer: On the scale of weather noise it is slow.
It’s always good to explain the reference frame.”
Thanks! I still have questions however. Sorry! How is it “tested empirically a number of different ways” ?
And/or how do we know what is “slow”??
What exactly does “slow” mean anyway?
(Seems to me that the effect of a city, or growing city is immediate and constant on the temperatures)
thanks!
Re: steven mosher (May 28 10:13),
I have playing with using difference between the mean and the median in combine duplicates as a method of locating errors in the supposedly QCed GHCN data set – and there appear to be quite a few. If I get organized (it IS golf season here, however), I will post something up on Statpad.
j ferguson,
The lack of wind is what causes the problem. There is no mixing and the warmer air does not rise in a manner which you suggest. It sort of hangs over the area (for the lack of a better term) and the heat slowly coming from the concrete prevents the local air temperature from getting down as far as it normally would.
j ferguson,
The lack of wind is what causes the problem. There is no mixing and the warmer air does not rise in a manner which you suggest. It sort of hangs over the area (for the lack of a better term) and the heat slowly coming from the concrete prevents the local air temperature from getting down as far as it normally would.
.
Mike, you are in denial of basic physics. try again.
MikeC
I used to get significant bumps (lift) flying sailplanes over areas we thought were generating convective movement – thermals. I had supposed the same thing would happen over a warm runway at night, but maybe there isn’t enough energy to make the air rise much-but rise it must.
I do remember that thermal activity in the upper midwest tended to die in the late afternoon.
sod,
“Mike, you are in denial of basic physics. try again.”
Yes sod, we’ve all heard of the term “heat rises”
Now, sod, take your thumb outta your mouth and put your rattle down and read about boundary layers and inversions. Here, I’ll even give you the links:
http://en.wikipedia.org/wiki/Planetary_boundary_layer
http://en.wikipedia.org/wiki/Inversion_(meteorology)
j ferguson,
j, if I recall correctly, when the air is calmest (an hour or so before sunrise), it is because the convection has stopped. What causes wind?
Sail planes… reminds me of ….quiet. I love sailing(boating) and I went up in a glider once… one thing that always stands out… the quiet.
Re: MikeC (May 28 17:40),
Does Gavin Schmidt Understand Boundary Layer Physics?
and
Comment On The Inaccurate Response By Gavin Schmidt Of Real Climate On The Role of Land Use Change On Temperature Trends
Nothing in any of these threads attempting to validate the surface station network has included evaluating the actual quality of the individual station data nor considered climatic effects as referenced above. It is akin to concluding an investigation of a plane crash by simply observing there was a crash and skipping over the flight, if the plane even got off the ground. Replicating error is still error.
Also, as Zeke linked to Watts’ paper, Case #13 by Edward R. Long summarizes:
Have his results been replicated?
MikeC:
You haven’t read the papers I’ve read apparently. Nor done any calculations on your own.
Reading papers doesn’t teach you everything. Most of them are junk in any case.
At some point you have to do hands on calculations. Rhetoric is pointless, so if that’s all you can do, your options are pretty limited.
Re: Carrick (May 28 13:11),
well I found the data from peterson 2005. us population at 1sq km from 1930 on.
Should be able to test some stuff with that. Not that it will convince some in the room. Now back to vacation.
Hey! Don’t tease!
Links please!
🙂
What a bunch of BS.
ron its on the noaa ftp.
ftp://ftp.ncdc.noaa.gov/pub/data/uhi/
Have a crack at it, I’m on vacation.. so wont be able to have a wack at it.
I think with Zeke’s and Nicks programs and you metadata we could do some interesting stuff. Its all ascii,
DG,
“Does Gavin Schmidt Understand Boundary Layer Physics?”
Of course he does. He also knows how to dance real well. ie: skirt the issue.
DG,
On the second paper”Comment On The Inaccurate Response By Gavin Schmidt Of Real Climate On The Role of Land Use Change On Temperature Trends”
Here’s another skirt job by Gavin. The subject is the effect of land use changes on temperature trends. Gavin changes the context to the effect of land use change as it applies to models, not measured surface air temps.
carrick 44319,
carrick, you can’t fake your way through this… well, maybe with some people you can.
So far, every comment or suggestion you’ve made on this issue has made no sense… it’s like having a conversation about baking chocolate chip cookies and you don’t know about the vanilla, flour or sugar.
Just a few days ago I offered you the bio on USHCN which has all of the relevant papers on temperature monitoring and you made it clear you did not need to read the literature which is the foundation of temperature monitoring… you also made clear your frustration with not being given the exact cut and paste and link. I’m not your butler. If you want to have a discussion about temperature monitoring or UHI then read the papers… especially the ones which serve as the issue’s foundation and building blocks.
If you’ve missed something then please take the time to read the material, it will only increase your understanding of the issues and improve the overall knowledge of people who read what you have to say. It would be much more productive than polishing your crystal ball and guestimating how much calculating I have done. Okay, since you need to know, I’m not done calculating yet. The issues involved in temperature monitoring are many and complex. But I continue to move forward…
DG,
“Nothing in any of these threads attempting to validate the surface station network has included evaluating the actual quality of the individual station data nor considered climatic effects as referenced above. It is akin to concluding an investigation of a plane crash by simply observing there was a crash and skipping over the flight, if the plane even got off the ground. Replicating error is still error.”
I’ll disagree with you on that one. The analysis done by zeke n co has been more of a macro sense… they are looking at improving their tools and knowledge, and they’re making an honest try at it… and they do have some strengths with the computer.
MikeC:
Well I’m faking nothing but if you can point to something that I’m “faking” beyond some vacuous comments on which peer reviewed papers one has read, have at it.
If you find a substantive issue that is supported by peer reviewed literature, bring up the point combined with the literature. The idea that I can’t sit down and put a bounds in UHI bias effect without reading every single paper that is marginally related and is good, bad and worse is just a friggin’ hoot.
If you know papers that contain apropos facts by all means share them. If you expect me to read your mind and figure out what is important about them, well I’ve better things to do with my time. I’m referring to a concise link to relevant papers with a synopsis of what it is that is relevant to the point you appear to be making wrt UHI impact on global mean temperature trend. I’m not your student and I don’t have time to do your work for you.
And If you’ve done calculations that’s great. But if you haven’t discussed your calculations, from a scientific viewpoint that’s essentially the same thing as having never done them. If you have calculations that you wish to share with the world, have at it. We’d love to see them.
Finally I didn’t “make it clear” that I didn’t need to “read the literature”. I made it clear one doesn’t need to read every paper on any topic to understand the issues on a given topic. I appreciate the importance of coming up to speed on measurement and analysis methods, but it’s not like I’m a noob with no experimental measurement or analysis experience of my own.
In any case, I don’t see much point in slighting others for having not exhaustively read ever possible paper on a topic before they are allowed to open their mouth in a public forum. What nonsense.
carrick, try the papers on the list that is the foundation of the temperature record that I offered. You don’t need to write paragraph after paragraph of excuses or try to change the subject.
MikeC, you brought this up. You keep bringing it up. Why not tell us us which papers in the webpage you linked are relevant?
You know this is the standard, not just complaining about somebody else not reading what you think are the relevant papers without explaining what it is that you think is relevant about those papers? Seems like a better idea than demanding that other people invest in crystal balls, to me.
MikeC,
I think I can appreciate your frustration that concepts contemplated here may already have been dealt with in detail in the literature and that you may think that wheels are being spun.
I don’t doubt that smart unbiased people have already been down some of these paths and that their observations have influenced the collection and use of the “raw” temperatures. And that they have published and the publications are worth reading.
But, isn’t it worthwhile to try to put the various conjectures to the test going back to the data sets themselves? And to do it here?
Since it doesn’t seem, to me at least, to be impossible to assemble the components and effects driving the temperature series and to look at each of the contributions they make for reliability and maybe authority, Carrick’s approach makes good sense.
And I suspect, he feels that wading through the literature would spin wheels.
But maybe it’s a blogging misdemeanor to speculate about something like this.
Can you point to a single document that would enlighten one of his statements?
One way to handle this is for MikeC to make a statement about how much he thinks UHI infects the record. Its time for MikeC to make a testable claim and then offer up his suggestion about how we would go about testing that claim.
Let’s take airports and UHI. MikeC assumes that airports will show a larger warming trend than non airports. How much mike? what kind of effect size are we looking for? How would you propose looking for that effect? That’s a simple request.
[q] Let’s take airports and UHI. MikeC assumes that airports will show a larger warming trend than non airports. How much mike? what kind of effect size are we looking for? How would you propose looking for that effect? That’s a simple request.[/q]
http://books.google.com/books?id=wokqNDknbLIC&pg=PA7&lpg=PA7&dq=Montávez+et+al.,+%5B2000a&source=bl&ots=cgioPnllWH&sig=G7CxkAjCjNBMUIPYB7XkFW4vONs&hl=en&ei=1-oDTMqLHYy0NoOunDs&sa=X&oi=book_result&ct=result&resnum=4&ved=0CCQQ6AEwAw#v=onepage&q&f=false
Here I googled. That’s a really long link so I hope it works! It is a link for a book preview.
Lot’s there in the preview. Just look at page 8 and read on. For these questions above from Mosher look in figures 1.12 and 1.13 on page 10
“Heat islands: understanding and mitigating heat in urban areas By Lisa Gartland”
carrick, I linked you to the USHCN page, told you that to understand temperature monitoring (which includes UHI) you need to review the studies on that list.
j.ferguson, I’ve already told carrick I’m still looking at the issue. I am not done calculating, I’ve also told carrick that the UHI issue is going to depend on the amount of objects such as concrete and buildings etc which is in the area. Now, carrick and steven the merlin mosher can play crystal ball, guess and give conjecture all they want, but until they read the studies, they have no clue… like I said, it’s like discussing baking chocolate chip cookies and not knowing anything about sugar, vanilla and flour.
wow, steven the merlin mosher is claiming that i’m assuming stuff? hmmmm, very intresting… merlin, your wizzard skills suck
my balls are hge
Hellooooo! Did you see my link?
Wake up and smell the UHI! 😉
j ferguson, I believe it is MikeC who is spinning wheels, covering ground that has been covered before. I have already read what I believe to be the relevant literature (Jones, Peterson, Menne, Easterling, Stone, McKitrick, Mills, Ungar, Parker, Hansen are papers authors that come to mind) but the idea that I have to read every paper that’s ever been written on measurement collection and analysis, which is what MikeC appears to be suggesting, in order to comment on this is itself risible.
Rather than continuing with this incessant bickering, I figured it would be more productive to include my own line of reasoning.
The correction I’ve seen from UHI on global temperature trend is typically less than 0.1°/century in the land record compared to perhaps 0.7°/century for land only (CRUTEM3), and 0.03°/century compared to 0.65±0.1°/century for global land+ocean (HADCRUT, I’m using 1900-2000 numbers).
The fact is of course that UHI only affects the land record, land is only 30% of the total surface area, so UHI has a steep hill to climb to explain a dominate portion of the global land+ocean temperature trend.
Further if we look at SST by itself (I’m using hadsst2), I get 0.63±0.1°/century. It’s hard to explain how UHI could be a dominant player for land record and still obtain such a similar number for SST trend.
Finally, if we look at the latitudinal contributions to land temperature trend (I believe I used 1960-2010 for this case), what we see is a pattern that is not consistent with the hypothesis that the dominant contribution to land temperature trend is UHI (peak urbanization does not occur at 60°N of course).
figure.
That’s about all I have to say on this, unless somebody wants to respond with substantive comments on the matter.
yes carrick, UHI has a huge hill to climb to explain all of the global warming. It cant get there from here. the basic math of 70/30 is huge. MikeC doesnt get that. PS I like the latitude stuff.
Sorry MikeC I’ve read everything you linked to. care to tell us about OMR? you claim to be calculating. Really? who was it again that helped you with Menne?
Heh, “nothing to see here” move along! and look at the sea!
Let’s hear it for peer review! LOL what a joke.
BTW o’worthinesses, the UHI doesn’t have to explain ALL of the warming when graphs are plotting a LESS THEN ONE DEGREE rise “in global average temperature”. The urban temperature readings DO MATTER they are on Zeke’s chart and that’s where some of the data comes from and they are part of the globe!!!!
From that link I provided, all that work done for that book; the preview shows in just those view pages that MIN temps may certainly be affected by UHI and airports could be warmer! What is so very lame is that you guys refuse to look.
Liza. read that. doesnt answer my question.
Yes it does steve mosher. Sheesh.
Liza:
It hardly explains any of the land temperature change, and none of the SST change. We can ignore it if we choose. That’s the point.
Liza:
It hardly explains any of the land temperature change, and none of the SST change. We can ignore it if we choose. That’s the point!
Liza:
It hardly explains any of the land temperature change, and none of the SST change. We can ignore it if we choose. That’s the point…
The corrupting influence of the UHI effect caused by regular snow removal from runway tarmac — making landbased readings higher during the winter at airports than for the rest of the French countryside — was the conclusion of a recent study done there on this subject. Adjusting for the erroneous thermometer placement, there has been no actual ‘global warming’ in France for decades. In other words, even the French understand that you don’t have to dangle a thermometer in the exhaust of a commercial jet deceive the ignorant and superstitious herd.
evilincandecentbulb, is it then your argument that globally there has been no warming?
Some coherency in thought around here would be appreciated.
Lisa, the point is this.
POINT meaurements of UHI indicate that cities are warmer than rural. We know this. Jones knows this. peterson. parker. everybody. We also know that the number can be very large.
In 2003 peterson tried to isolate this effect. he took rural stations and compared them to urban stations. There were of course issues with his site selection BUT the proceedure was clear. Take stations that are supposed to be rural compare them with urban. YOU WOULD EXPECT a difference. peterson did. he didnt find one. He concluded that the stations in urban areas must be in “cool parks” Cool parks are a DOCUMENTED phenomena in urban settings. They are pockets of cool. See the primary literature:Oke. Parker also looked at the problem. he used different stations and a different methodology. Temperatures on windy versus calm nights at urban/rural sites. Again, not just a look at a single station but a brad look at stations all over the world. he found no difference. Also a mystery. he too explained it by “cool parks.” Jones also did a study. same approach. same result. When we select a single city and look at rural stations close by we see a lower temp for the rural stations. When we look at the TREND over time, we see no difference. The desert is hotter than the arctic. But if you look at the trend, the change over time, you will see that the hot gets hotter and the cold gets hotter. So while we do see an difference in the absolute temps, when we look at anomalies, the change over time, the effect of UHI becomes very small. ITS TREND we care about. if your average temp goes from 0C to 1C over 100 years that is no different than going from 15C to 16C. its still 1 C per century.
In any case, as carrick notes the SST has also been going up. No UHI there. Further we measure the temps at the troposphere, the air above the cities and the countries. The coverage is global. the resolution is fine. Whatever UHI effect there may be at ground level, vanishes at the troposphere. Mixing. What trends do we see in that data? trends that are close to the land trends.
Whatever, UHI there is in the land record, in the TRENDS of the land record is limited. We know this by looking at the trop data.
“Further we measure the temps at the troposphere, the air above the cities and the countries. The coverage is global. the resolution is fine. Whatever UHI effect there may be at ground level, vanishes at the troposphere.”
You didn’t look at that book preview.
Sorry, I find no proceedure there for testing the difference in trend at airport sites versus non airports sites. None. rather than cite something that doesnt answer the question, put your words in a post.
“From that link I provided, all that work done for that book; the preview shows in just those view pages that MIN temps may certainly be affected by UHI and airports could be warmer! What is so very lame is that you guys refuse to look.”
yes. Airports could be warmer because of the heat retaining properties of the materials. This would impact Tmin. This is WELL KNOWN. So construct a hypothesis and a way of testing that.
The trend in Tmin over time ( WARMING) at airport locations will be greater than the trend in Tmin non airport locations.
Since Tmean = (Tmin+Tmax)/2 it follows that the trend in Tmean at airports will ALSO be greater than the trend at Non aiprort locations.
Rememeber we are talking about TRENDS.. the rate of increase.. getting warmer over time. TREND.
How to test:
1. select all airport locations create the average trend
2. Select all NON airport locations: create the average trend.
Compare.
then scratch your head when you find out that airports warm about the same or less rapidily..
then remember what the primary research says about boundary layers. then listen to carrick.
Liza, rather than present the material in context that you think is relevant, you link to an entire book. Then everybody else is supposed to then magically divine what it is you thought was germane to the argument.
At the least you should point to the specific text that undermines the statement:
Carrick, if you search for the word troposphere in the book preview, you get 2 examples (well I did anyway): pages 32 and 33. Page 33 says:
“Figure 3.5 shows typical daytime temperature profiles in the boundary layers over rural and urban areas. Warmer urban surfaces create a thicker boundary layer, above which the temperature reverts to that of the rest of the troposphere.”
Now I’m no scientist, but I don’t think that undermines Mosher’s claim…
Carrick it’s a book preview. I can’t highlight the text. Besides I gave pages for airport data and you ignored it.
here’s an illustration.
http://www.actionbioscience.org/environment/figures/voogt1.jpg
How can you possibly know what the UHI does to clouds? Yeah, that question is just an example because the UHI effects the weather and that book preview also gets into MODELS just for the UHI.
Seems to me that a lot of this is understood but way more complicated then the AGW crowd will admit.
I personally don’t believe some magic formula you add to data representing “all the world” fixes these things down to fractions of a degree. I spent a lot of time here just using where I live as an example! I don’t think many of you can grasp how VAST this land is and how much data you do not have, and that you have no way to claim these “fixes” are so perfect to make claims about “the whole world’s” temperature.
And just like the MWP; the UHI “you’ve got to get rid of it” .
“From that link I provided, all that work done for that book; the preview shows in just those view pages that MIN temps may certainly be affected by UHI and airports could be warmer! What is so very lame is that you guys refuse to look.”
yes. Airports could be warmer because of the heat retaining properties of the materials. This would impact Tmin. This is WELL KNOWN. So construct a hypothesis and a way of testing that.
The trend in Tmin over time ( WARMING) at airport locations will be greater than the trend in Tmin non airport locations.
Since Tmean = (Tmin+Tmax)/2 it follows that the trend in Tmean at airports will ALSO be greater than the trend at Non aiprort locations.
Rememeber we are talking about TRENDS.. the rate of increase.. getting warmer over time. TREND.
How to test:
1. select all airport locations create the average trend
2. Select all NON airport locations: create the average trend.
Compare.
then scratch your head when you find out that airports warm about the same or less rapidily..
then remember what the primary research says about boundary layers. then listen to carrick.
And Mark, yes. What Liza forgets is that Cities represent A TINY portion of the earth. Like a kid peeing in a pool it may be locally warm, but that warmth diffuses. It has too, 2nd law and all. so when we look at the upper trop or even the lower trop we get a good idea how warm the planet is. Funny how liza missed that portion of the book. maybe she just found it on google.
Lisa, you seem so positive about a MWP.. but those “measurements” are a lot less scarce than todays measurements. You believe that a diary written in china or grapes in england represent the whole planet. funny. or pathetic? or both
Steve Mosher, You mean “today’s measurements” as in 32 yrs of being able to measure the temps or ice from space?
Better question, do you think 32 yrs or 300 of years of the historical record gives you a real picture of what the climate does on earth? And by the way those people back then who wrote things down had the same size brain as you do.
(I also gave you links to biology papers. There was one about salamander fossils found in Yellow Stone and the MWP. Did you that forget again? Fossils aren’t good data? hmmmm.)
mark, I think Lisa is still choking on that bone about the troposphere. probably searching furiously
How the trop modulates UHI
http://journals.ametsoc.org/doi/full/10.1175/1520-0450%282002%29041%3C0519%3AAONNTT%3E2.0.CO%3B2
Here is one that uses reanalysis data to estimate UHI.
http://www.ncbi.nlm.nih.gov/pubmed/12774119
That is, we look at the trends in the surface data and then we compare that to data from the atmosphere where the effects AT THE SURFACE are removed.
This figure is .27C century. Jones figure is .05C.
I guess for those who have followed my argument over the years, I would say that .27C is closer to the truth than .05C
carrick, your thoughts. What upper bound do you put on UHI trend bias?
Steve Mosher I am not choking on anything. You guys are full of it. I think you need AGW to feel some what important.
“Funny how liza missed that portion of the book. maybe she just found it on google.”
I didn’t miss it. My preview will not show those pages. It does show the next two pages which talk about night time temperatures into the higher air.
Funny on those graphs of the Arizona airport, Mesa, and Tempe etc there’s no jump in MAX temperatures over the years; at the height of “global warming” alarm
here lisa:
urban cooling:
http://www.springerlink.com/content/f4l07h118281881l/
Mesoscale and macroscale phenomena were examined during the different UHI classes through a weather type scheme. It was emerged that high UHI classes are associated with anticyclonic conditions or advection in the lower troposphere, while low UHI classes are associated with strong northeasterly winds. Anticyclonic conditions which frequently occur in spring and early summer, reduce or reverse the UHI to Urban Cooling Island.
Or here Lisa, look at literature about verticle profiles of UHI
http://journals.ametsoc.org/doi/full/10.1175/JAM-2176.1
or this
pielkeclimatesci.files.wordpress.com/2009/09/ppt-54.pdf
in a nutshell:
Apart from its horizontal range, the UHI also has some vertical structure. It usually reaches up to between 200 and 300 m into the air, about 3 to 5 times the height of the buildings. In a cloudless sky it may reach up to 500 m into the atmosphere. Two distinct layer are seen:
1. the urban canopy layer occurs nearest the ground and results from heat emitted by low level emitters such as house chimneys, from the buildings themselves (as they absorb lots of solar radiation and emit it back as heat) and also from vehicles.
2. the chimney layer occurs above the urban canopy layer. Here the heat is emitted into the air from the high level emittors, e.g. the chimneys of power plants.
nice try lisa.
Like I said. Whatever LOCAL effects there are at the surface dissipate by the time you measure at the trop. We know this because folks have been measuring and modelling the UHI profile.
So when the Land record increases at .17C per decade for the past 30 years and the trop increases at .15C per decade. GUESS WHAT?
you have a pretty good estimate of the size of the UHI right there.
Say .02C per decade or .2C per century.
When you go hunting for a signal like UHI you better damn well have a good idea of the size of the signal you are looking for.
Now, JONES and others think the figure is more like .05C per century. I’m more toward the .2C camp.
Either way, GHGs still cause warming.. .8C or .6C, shrugs
Liza:
Urban environments account for about 3% of the total land surface area, or about 1% of the total area of the planet.
You seem to be arguing on one hand that humans have a fine-tuned control where 1% of the surface area can cause a 0.7°C rise (via some mystery mechanism) but a global change attributed to human activity (increase in atmospheric CO2 concentration) has no effect. That’s a bit cracked up logic.
In any case, it can be quantitatively addressed: large-scale eddy simulation codes over an urban canopy. I’ve got some references in my list to read. My interests go beyond just the effect of UHI, since I collect data in urban environments for my own work already.
have plenty of personal experience with measurement and analysis in an experimental setting (meaning repetition is possible), and I do know what Jones, Hansen etc claims is not only possible but highly plausible.
Whether you personally “believe” it, so what? That’s like taking a poll on which Ben and Jerry’s ice cream one prefers. If you can’t quantify why you don’t believe it, it’s just an unsubstantiated and unquantifiable belief. And that has no place in scientific discourse.
Liza:
Urban environments account for about 3% of the total land surface area, or about 1% of the total area of the planet.
You seem to be arguing on one hand that humans have a fine-tuned control where 1% of the surface area can cause a 0.7°C rise (via some mystery mechanism) but a global change attributed to human activity (increase in atmospheric CO2 concentration) has no effect. That’s a bit cracked up logic.
In any case, it can be quantitatively addressed: large-scale eddy simulation codes over an urban canopy. I’ve got some references in my list to read. My interests go beyond just the effect of UHI on global temperature trends, since I collect data in urban environments for my own work already.
I have plenty of personal experience with measurement and analysis in an experimental setting (meaning repetition is possible), and I do know what Jones, Hansen etc claims is not only possible but highly plausible.
Whether you personally “believe” it, so what? That’s like taking a poll on which Ben and Jerry’s ice cream one prefers. If you can’t quantify why you don’t believe it, it’s just an unsubstantiated and unquantifiable belief. And that has no place in scientific discourse…
Stephen Mosher:
In terms of it’s effect on biasing the temperature trend (in the sense of a measurement error) no more than 0.1°C/century land, maybe 0.03°C/century land+ocean. My best guess is closer to 0.05°C/century land, 0.015°C/century land+sea.
“Urban environments account for about 3% of the total land surface area, or about 1% of the total area of the planet.
You seem to be arguing on one hand that humans have a fine-tuned control where 1% of the surface area can cause a 0.7°C rise (via some mystery mechanism) but a global change attributed to human activity (increase in atmospheric CO2 concentration) has no effect. That’s a bit cracked up logic.”
Um no I’m not arguing that at all!
Geez you guys are useless. I showed you a video last week of how many stations are not rural which includes HUGE areas of land like Canada, Russia, Asia. And you jumped all over me then as well. You’ve got an answer for everything and it all adds up to a FRACTION of a degree. This is religion not science. Amazing. ( Of course you didn’t answer my questions about how many years makes climate on Earth a trend!!)
Mosher:
But land and trop are not expected to warm at the same pace. The upper troposphere is supposed to warm faster than the surface – about 1.2 times as fast as I recall. The satelites show slower warming. If you take that into account and you think we get a pretty good estimate of the size of the UHI-bias from the difference between land and satellite measurements then we should estimate the effect as larger than your .2C?
So, warming not attributable to UHI is GHG’s? I don’t think you meant to say that. Or are you talking about sensitivity?
Liza:
Then do you even have a point? 70% of the Earth’s surface is ocean. Obviously there are no UHI microsite issues there. The only way UHI could influence SST temperature is via indirect effects like clouds. Without that, you have no argument.
And as I pointed out, for land, the dominant contribution to land temperature trend is high latitude, not consistent with UHI have a substantive effect. As as I pointed out, SST temperature trend is nearly (but slightly less) than land temperature trend. This is consistent with the differences in land versus marine atmospheric boundary layer physics.
Also,”FRACTION of a degree”. We’re talking trends here not absolute temperature. So the units are DEGREES PER CENTURY (for example). What is important is changes in temperature WITH time, and when you average over 1000s of stations and thousands of measurements at each site, a lot of these issues you worry about just get washed out.
I’ve “measured” (and published) quantities to a part in 10 to the eighth. More than once. You expect me to feel like measurements to a part in 100 is undoable?
I’m amused that you accuse us of “religion.” You’re the one who thinks rhetoric is a substitute for quantitative analysis. It’s not.
Carrick:
“Urban environments account for about 3% of the total land surface area, or about 1% of the total area of the planet.”
I’m not sure that’s the whole point of the worry about the impact of Airport UHI on the aggregate temperature. My take is that the worry is that many more than we think of the “rural” thermometers are located at airports and their (supposedly elevated) readings exported into the adjoining grids. If this were happening, AND the airport temperatures were different from actual rural, then of course they would have a greater effect on the global number,
I know there has been some sampling of the data to see if there is anything to this worry, and at least to the NOAA folks, this worry seems groundless.
Carrick what is the average temperature of the Pacific Ocean then? What is the average rain fall on Earth supposed to be exactly? See how silly that sounds?
“What is important is changes in temperature WITH time,”
Hello. I asked how much time on Earth 32 yrs or 300 yrs of various written down history? And I got NO ANSWER.
I think you don’t have enough data. (I said that clearly above) I think that the more you play with these measurements, do your statistical tricks on them, the more you get away from this real VAST world. BTW I am not the only one asking questions!
I am having the hardest time getting a comment in…PLUS I am painting my bathroom at the same time! LOL
Liza,
Don’t stand on the potty. they are not always as well secured to the floor as you’d expect.
Ah yes, as Philip Stott recently observed, a glimpse of the the spirit, the essence — indeed, the very ethos — of the UN-IPCC and all of the freakish modern day AGW academics, is truly captured, albeit many years ago:
“The principle of private or natural conscience is extinguished in each individual … and nothing is considered but how the united effort of the whole (released from idle scruples) may be best directed to the obtaining of political advantages and privileges to be shared as common spoil.†(Wm Hazlitt, ~1802)
j fergunson, thanks. I know it is temping. It is that lid on the potty that might move on you too. My arms hurt-had to do the ceiling too.
evilincandecentbulb- interesting quote you got there. 🙂
“As pointed out by Oke (1979), the internal pattern of the heat island of each city is dominated by microscale features related to land use and building density. In large-scale studies of climate change using hundreds or thousands of stations, the microclimate of each station cannot, at present, be properly parameterized. This presents a major obstacle to the quantification of the heat islands produced by small cities and necessitates the use of population as a measure of urbanization. The noise produced by the microclimate of each city, however, may disguise the effect of urbanization.”
Karl, T.R., H.F. Diaz, and G. Kukla, 1988: Urbanization: its detection and effect in the United States climate record, J. Climate, 1, 1099-1123.
pdf link: http://www.met.sjsu.edu/~wittaya/journals/Urbanizations.pdf
now that it seems that folks are reading the literature, we can have a chat…
Many clues in that statement… carrick, if you haven’t read that one, it’s some fairly decent work… I’m going to be back and fourth between this and a work crisis so it’ll be spotty, but I’ll show some points that will hopefully help understand my point…
But lets get to the first point… population is a method which can be used to study urbanization because there will be more urban objects such as streets, structures and etc.with larger population.. however, a point made by merlin the mosher is cool parks (simply stated, cooler areas of the same heat island) often at airports where ASOS is properly sited… also, the photodocumentation provided by Watts which shows small scale urbanization much closer to stations regardless of the population… and we’re talking, what, 89% (although there is uncertainty because he did not separate a small number of cool biases, such as shading, from the warm biases) … both of these instances (cool parks and microclimate) create uncertainty when using population to estimate UHI contamination of the temperature record.
… but cool parks and bbq’s aren’t the only problems for the population scheme, then there are them darn trees… next point which is important why peterson came to the conclusion he did (aside from the fact that he is a stern advocate)
be back in a bit
Okay, them wonderful trees… what happens to a temperature record when a tree grows near the station depends on where the tree is in relation to the station. Trees growing to the west of a station will cause a shading issue if the shading occurs during the warmest part of the day (Tmax). If trees grow close to the station (other than west) then there will be warming in both day and night temperatures, this is a wind issue. Every one here blows on their food when it is too hot in order to cool it down… that is how it works during Tmin, if the wind is blocked then things just dont cool down like they should. During the day, if the wind is blocked, the station sits there in the sun (where it is designed to sit) and warms, a problem during Tmax.
As I’m sure everyone is aware, trees grow slowly. So a bias will be introduced as the tree grows.
The NWS has a siting guideline that requires stations to be a minimum distance of 4x the height of the tree (or structure). And the observers and NWS station managers I have spoken to indicate that there is a lot of concern about shade trees, but not wind block trees.
NCDC used to have meta data available which gave the height, width and direction of any trees or etc which was near a station. That metadata has since disappeared from the MMS site (probably because they use the automated version 2 system for adjustments now).
I’m sure of this because I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems… with the overwhelming majority being wind block problems, so if anyone wants to know why Peterson could not find UHI when comparing clusters of rural and urban stations, there’s your answer… oh, and he did not use stations from the network he was studying…
so the moral of the story is, trees too close to the station in rural areas will warm a station just like urbanization will, making it harder to detect urbanization when comparing rural and urban stations
Okay, I gotta run for a little bit, next bit will be on wind and Parker
UHI is most profound at night, just a touch before sunrise when the wind is calmest… the heat from the concrete and such lingers… and keeps that Tmin thermometer from getting as cool as it would have without the concrete…
Wind and Parker… I’ve read a few different reasons why Parker did not find UHI in his wind study. Basicly Parker claimed he compared station records on calm and windy nights. but… what Parker failed to mention was that he compared the third most windy to the third least windy. So, despite what was advertised to the IPCC and friends, Parker compared windy vs less windy.
Next, in order to get the hourly wind data he needed, Parker would have (at least in the US) to get data from ASOS stations which are usually well sited in grass fields at airports. These airports will have varying amounts of UHI depending on how much urbanization there is around the airport itself. So an airport with lots of hotels and warehouses will have more UHI. But, in general, ASOS stations will experience less UHI… so, his sample was biased twards cooler urban stations.
Next there is an argument out there about the thickness of the boundary layer… something to do with large cities having thick boundary layers, so when the wind picks up and there is supposed to be mixing, there is not really much along the lines of mixing.
next bit of chat will be GISS night lights adjustment…
Well, the GISS nightlight adjustment is pretty simple. Satellite pics taken at night show lights which are pretty much in urbanized areas. So a station in a more lit up area is going to be more urban than the station in a dark area. The nightlight adjustment suffers the same as the population adjustment in that the satellights will not show the biases created by the bbq’s, air conditioners, sidewalks, waste treatment plants and etc which Watts has documented. The same goes for trees… more trees do not give off more light. So rural stations which are biased by microclimate or wind blocks will bias the GISS adjustment. Otherwise, the GISS adjustment would get most of it…. which brings us to our next point… and that is the point which merlin the mosher likes to highlight… how much is it…really… in the end…
but that has to wait til tomorrow…
I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…
Gosh I picked some peterson site randomly
https://mi3.ncdc.noaa.gov/mi3qry/locationGrid.cfm?fid=1422&stnId=1422&PleaseWait=OK
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=899&PleaseWait=OK
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=921&PleaseWait=OK
Google earth shots. hmm I see catcus. I see low shrubs. I see fields.
Not so many tree problems.
For peterson sites you can start here
http://climateaudit.org/2007/08/03/petersons-urban-sites/
Then look them up. Then google earth. Then over to MikeC
how come you said all but 1 peterson rural had a tree problem and the first three I pick dont appear to.. one might some low tree to the east.
Mike, if you want just send the lat lon for peterson sites and I can make google earth tours. Takes a few mintues to confirm what you said.. or disconfirm
“I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…”
Thats according to Mike C.
So lets check…
I picked some peterson site randomly
https://mi3.ncdc.noaa.gov/mi3qry/locationGrid.cfm?fid=1422&stnId=1422&PleaseWait=OK
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=899&PleaseWait=OK
https://mi3.ncdc.noaa.gov/mi3qry/identityGrid.cfm?setCookie=1&fid=921&PleaseWait=OK
Google earth shots. hmm I see catcus. I see low shrubs. I see fields.
Not so many tree problems.
For peterson sites you can start here
http://climateaudit.org/2007/08/03/petersons-urban-sites/
Then look them up. Then google earth. Then over to MikeC
how come you said all but 1 peterson rural had a tree problem and the first three I pick dont appear to.. one might some low tree to the east.
Mike, if you want just send the lat lon for peterson sites and I can make google earth tours. Takes a few mintues to confirm what you said.. or disconfirm
MikeC you need to read the parker thread at CA where we discussed all of those issues and even sent questions to parker.
He used NCEP reanlysis winds BTW. whenyou have something that we have not already discussed, let me know
Just thought that I would throw this into the mix –
{Cities can be up to 10C warmer at night than surrounding rural areas, partly because they absorb more heat from the sun but also because they generate more heat from vehicles, lighting, machines and air conditioning units. Even the metabolism of millions of city dwellers adds to the temperature.
Mark McCarthy, a climate scientist for the Met Office who led the research, said: “The impact of this waste heat on a global scale is very small, but it is hugely significant at the city scale, where it can have a big influence on urban climates.†}
http://www.timesonline.co.uk/tol/news/environment/article7141432.ece
Liza:
Averages are well defined concepts. I think it’s really obvious you don’t grok them, nor the statistical theory surrounding it.
At all. Nada.
Also, “exact values” has no role to play in statistics.
J Ferguson:
I don’t think grids has much to do with it, but I do think that airports are anomalous in the sense of having very large fetch, low surface roughness and associated fully developed turbulence and larger scale/lower frequency motion than is the case for teh ABL over more typical terrain.
In a sense they may be the most ideal sites for characterizing regional scale temperature, especially compared to urban or forested canopies or especially over rough terrain. “Actual rural” may have more microsite surface information encoded in it… some interesting issues here…
This is certainly a question that can be studied and answered using state of the art micrometeorological tools (e.g., LES codes that can handle large surface roughness and/or sloped terrain). Unfortunately I’ve got some other real world concerns I have to deal with now. When I get done with them, I might see what I can do with some LES codes…
MikeC, I think you are over-analyzing this a bit from the perspective of global mean temperature. I feel the arguments I’ve given above: i) precent ocean to land (70/30), ii) nearly identical trends for ocean vs land, and iii) high latitude dominance of contribution to land temperature trend together sets a pretty strong upper bounds the maximum contribution of UHI to global temperature trend. It certainly can’t be the dominant contribution to land record (since urbanization certainly peaks well below 60°N), land temperature trend > ocean temperature trend in every model out there, it is unlikely that you can account for more than about 0.1°C/century of the land temperature trend from UHI alone… and when you factor in i) that pretty much bottoms out the global effect of UHI at around 0.03°C/century.
Now I happen to be interested in the “urban forest” for my own purposes. If I do end up running LES codes it will be mainly for that, but I’ll keep an idea on an overlap with UHI effect on global climate monitoring.
Karl is a bit dated, but I’ll give him a look anyway. Probably this weekend. I have a full plate till then.
hi very good.thank you very much.
In the first half of the 20th century the land and ocean surfaces warmed at about the same rate. Subsequently, the warming rate of the ocean surface moderated slightly. The land surface, on the other hand, first cooled until the mid-70s then warmed considerably faster than the ocean surface. This pattern is inconsistent with the AGW hypothesis: how would increased atmospheric absorption of outgoing radiation decelerate the warming rate on the ocean surface and accelerate it on the land surface? Rather, the pattern is suggestive of being influenced more by surface conditions than atmospheric ones. The most obvious influence related to surface type is human population and associated economic activity including burning fossil fuels and urbanisation: there is much more of it on land than on the oceans. The urban heat island effect is usefully defined as thermometers recording additional heat generated at the surface locally by structures, rather than generally in the atmosphere by CO2 throttling outgoing radiation.
Adding weight to the reality of the UHI effect: first, the northern hemisphere land masses, where most of the world’s population live, warmed faster than those in the south; second, the differential was greater in winter than in summer; and third, the winter differential widened in the second half of the temperature record corresponding with increasing populations, economic activity and urbanisation.
The UHI effect implies, profoundly, that temperature could be driving emissions: as the weather turns cold, the world’s expanding stock of increasingly urbanised households turn up the heat, which is then detected by the climate system’s monitoring thermometers. Inverse correlation between emissions and land surface temperature supports this view. The inter-decadal proportionate change in each variable cycles with a period about 60 years, the cycles being out of phase by 180 degrees – as one variable goes up the other goes down. Alternatively, if emissions drive temperature, the inverse correlation suggests they are cooling the climate.
The UHI effect implies, profoundly, that temperature could be driving emissions: as the weather turns cold, the world’s expanding stock of increasingly urbanised households turn up the heat, which is then detected by the climate system’s monitoring thermometers.
.
you do not have the slightest clue about UHI. “turning up the heat” has basically no influence on UHI.
.
urban areas cause local warming, because concrete stores heat better than other substances do. because houses break strong winds. “heating” is a minor issue.
The UHI effect implies, profoundly, that temperature could be driving emissions: as the weather turns cold, the world’s expanding stock of increasingly urbanised households turn up the heat, which is then detected by the climate system’s monitoring thermometers.
.
you don t understand the UHI effect at all. please educate yourself, before you try to lecture people. (short version, longer one got eaten by server bug..)
The UHI effect implies, profoundly, that temperature could be driving emissions: as the weather turns cold, the world’s expanding stock of increasingly urbanised households turn up the heat, which is then detected by the climate system’s monitoring thermometers.
.
you don t understand the UHI effect at all. please educate yourself, before you try to lecture people. (short version, longer one got eaten by server bug..)
lucia, there are serious errors with the server…
Millett
You’ve never noticed that water has a particularly high heat capacity?
Took me a second to dig up the paper..
This could be of interest in regards to land/sea differences, and what the hypothesis actually is
http://www.agu.org/pubs/crossref/2007/2006GL028164.shtml
Carrick (Comment#44559) June 1st, 2010 at 11:52 pm
“Averages are well defined concepts. I think it’s really obvious you don’t grok them, nor the statistical theory surrounding it.
At all. Nada.
Also, “exact values†has no role to play in statistics.”
Oh the horror!
Um..Wrong! I’ve even read Stranger in a Strange Land.
To grok (all these statistical maneuverings) as if you have all the data you need for a vast complicated planet; down to a fraction of one degree, is silly too!
sod (Comment#44573) June 2nd, 2010 at 5:20 am
“you don t understand the UHI effect at all. please educate yourself, before you try to lecture people”
Do as I say not as I do?
http://books.google.com/books?id=wokqNDknbLIC&pg=PA7&lpg=PA7&dq=Montávez+et+al.,+%5B2000a&source=bl&ots=cgioPnllWH&sig=G7CxkAjCjNBMUIPYB7XkFW4vONs&hl=en&ei=1-oDTMqLHYy0NoOunDs&sa=X&oi=book_result&ct=result&resnum=4&ved=0CCQQ6AEwAw#v=onepage&q&f=false
Page 22 and 23 sod. Energy consumption is part of the UHI so he isn’t as far off as you are.
Re: carrot eater (Jun 2 06:34),
carrot eater:
I could not discern the theory about the discrepancy between land and sea temps from the abstract and I am too cheap and lazy to obtain the full doc. Can you summarize the theory?
Also, the paper predicts greater warming in the tropics but it appears that almost all of the current warming is up in the 50-60 latitudes (BTW, why is that?).
Zeke:
Nice work as always. However, I am not sure the quality of the metadata supports the sophistication of your analysis.
In figures 4 and 5 it appears that rural stations recorded higher temperatures in the 1920s and 30s than did the urban stations. I don’t get why warming would be more anomalous in the rural regions during previous warm decades than during the present warming. Perhaps broad regional changes ala Pielke Sr like the Dust Bowl?
Also, the rural and urban records converge as they approach the present which is either just a tad suspicious or a glowing tribute to the various adjustment and homogenization processes.
Tobin,
Fig 5 is rural airports vs rural non-airports. Probably many of the stations labeled as rural airports were not at their present locations in the 1930s. They were moved at some point to airport locations, as airports/airfields were built.
If I had to guess then, what you’re seeing in Fig 5, with the two series in agreement recently but a bit off before, is the result of station moves. Step changes due to station moves will show up in these raw unadjusted data.
Again and again, there is no homogenisation done in the data that Zeke is using.
Interesting short piece at Pielke, Sr.’s site on the effect of an airport on local weather.
I’m wondering if the effect of an airport is more related to the fact that the surface is dry and therefore gets a lot hotter during the day than a grass covered field where evapotranspiration acts to keep the surface cool. Does concrete have higher heat capacity and thermal conductivity than vegetation covered soil? The table in Wikipedia says the thermal conductivity of concrete (stone) and dry soil are very close at 1.7 and 1.5 W/mK.
Nice work Zeke. Looking at the urban and rural anomaly charts, I don’t see much difference between urban airport and rural airport trends. I would have expected urban airports to be trending much higher due to the surrounding industrialization.
A few months back I compared the trends between Toronto’s Pearson International and Muskoka Airport (rural). The trend in Tmean was much higher in Toronto over the period 1950-2004 (~1.5C/century vs 0.5) and the difference was even more pronounced in the Tmin trends.
Once again… nice work!
AJ
Do as I say not as I do?
http://books.google.com/books?…..mp;f=false
Page 22 and 23 sod. Energy consumption is part of the UHI so he isn’t as far off as you are.
.
this is garbage lucia, and you know it.
.
John’s claim was a strong one:
.
the world’s expanding stock of increasingly urbanised households turn up the heat, which is then detected by the climate system’s monitoring thermometers.
.
he made the claim, that people turning up their heating would influence temperatures. that is garbage.
.
page 16 of your link shows 8 factors. “increased energy use is only one of those. John was wrong. he read too much about barbecues.
George Tobin:
You can’t safely compare the temp record between the charts because they have somewhat different spatial coverage (e.g. grid cells wih both urban airports and urban non airports are often different from grid cells with the rural counterpart. I did some more traditional urban/rural station comparisons in my old In Search of the UHI signal post a month or so back (on my mobile at the moment so no link handy).
sod:
Part of UHI effect is waste heat and it is not a negligible effect.
You obviously don’t know what you’re talking about.
DeWitt Payne (Comment#44607)
” Does concrete have higher heat capacity and thermal conductivity than vegetation covered soil? The table in Wikipedia says the thermal conductivity of concrete (stone) and dry soil are very close at 1.7 and 1.5 W/mK.”
Well, most drunks would be able to testify that concrete is defiantly warmer than dirt as far as sleeping on…. In saying that, there is no such thing as “dry” dirt where im from. And the water content of the soil directly effects its thermal conductivity… and rain has the quickest immediate effect on soil temperatures for obvious reasons. (it certainly takes a long time for atmospheric air temp to effect soil temperatures, i keep an eye on these things… as grass and clover dont like below 12C, and im normally counting down the winter/spring days til temperatures rise above this) Concrete on the other hand, is pretty immediate in response to solar and atmospheric warming(closer to black sand than soil i would have thought)…. I’ve spent a few years sleeping in jungles/subtropical forests…. dirt sucks the warmth from yah… even in warmer parts o the world.)
So im surprised by those figures…
By now, hopefully, everyone has taken a look at steven moshers links to 3 of Peterson’s rural stations in comment #44556.
mosher’s assertion that these three stations are not affected by trees or other windblocks is entirely false. He also failed to mention that each station is located in the immediate vicinity of structures and other objects which will cause small scale UHI. I also question his assertion that these stations were randomly selected as they are all located in Arizona, where you would EXPECT to NOT see trees.
The first station is Tuscon 17, coop # 028795, located at the Sagura National Park (east) visitors center. The MMTS and rain gauge are located at 32 15 15.77N 111 11 49.11W. I couldn’t find the station on Google Earth so I called the visitors center and one of the observers walked me to the exact location which is between the building and a stairway which goes down to the wash (15 – 20 feet or so separate these two objects). The vegetation in the immediate area includes bushes and cactus which were planted when the visitors center was constructed and is the vegetation which was listed by NCDC when that information was available on MMS. The majority of the vegetation is up slope and above the height of the sensor. Not only does this station have vegetation and a large structure as windblocks, but there are the structures and objects which will cause artificial warming (building, stairway, parking lots and etc). In this instance, the vegetation will have a smaller effect of blocking wind than the visitors center building.
The second station is located at Santa Rita Experimental Range building complex which is run by the University of Arizona. The instruments are located at 31 45 42.91N 110 50 45.32W at the Range Headquarters building. The instruments can be seen in one of the pictures on a slide show here:
http://ag.arizona.edu/aes/cac/srerslides/slides.htm
As is also clear by the photos in the slide show, the trees in the area are full and close to twice the height of the buildings. In yet another case, the station will be artificially warmed by windblocks (structure and trees) and small scale UHI from objects such as the stone walls, gravel driveway and the HQ building.
The third station is located in the building complex at Anvil Ranch (31 58 46 76N 111 22 59 79), once again in the immediate vicinity of structure and trees which, from the satellite images, are tall enough to cast shadows on the rooftops of the buildings. So here again, there is going to be artificial warming from the windblocks (structures and trees) and urbanization (structures and gravel roadways).
This is a far cry from moshers claims about randomly selected sites with some cactus, shrubs or one tree. These stations, however, are a good example of why Peterson could not find UHI. His rural stations are artificially warmed by windblocks (and to some extent small scale urbanization) next to the stations. This is the same problem with most of the other UHI studies where population or night lights were used as proxies to estimate urbanization.
“I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…â€
“mosher’s assertion that these three stations are not affected by trees or other windblocks is entirely false. He also failed to mention that each station is located in the immediate vicinity of structures and other objects which will cause small scale UHI. ”
You asserted TREES. You asserted ONE and ONLY ONE didnt not have TREE PROBLEMS.
1 pulled three sites. 2 of the three did not have tree problems.
As I Said
“how come you said all but 1 peterson rural had a tree problem and the first three I pick dont appear to.. one might some low tree to the east.”
You said all but ONE had TREE problems. I pull three randomly.
one might. You claimed I ” [asserted] that these three stations are not affected by trees or other windblocks is entirely false.”
I said no such thing. I’m sure that everyone can read what you said. only ONE was free of tree problems. They can see what I said. I looked at three and one of the three might have tree prblems. hard to tell without knowing the kind of tree, the height and distance from the sensor. They can read you when you write that “mosher makes assertions about other wind blocks” they can read my words and see that I make no such assertion.
Like I said, post up the lat/lons and I’ll give them a google earth tour. confirm or disconfirm.
Oh what the hell,
From NOAA lat lons
HINT: in this I hint at what the real issue with many “rural” sites MAY BE
HINT: I started down this path when I looked at orland, CA, which warmed more rapidly than other rural sites in the early 20th century. Something changed there WRT land use.
Anyways for now its an untested hypothesis. The problem mikeC is you trusted NOAA metadata: MikeC
“I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…â€
Roll tape!
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.6025++-110.73444&sll=36.43821,-115.363124&sspn=0.002227,0.004206&ie=UTF8&ll=32.6082,-110.732478&spn=0.004663,0.008411&t=h&z=17
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=35.3694+++-111.5436&sll=32.6082,-110.732478&sspn=0.004663,0.008411&ie=UTF8&ll=35.3683,-111.54346&spn=0.018058,0.033646&t=h&z=15
trees? shrubs? windblocks? sunblocks?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.7433+-111.4139&sll=35.3683,-111.54346&sspn=0.018058,0.033646&ie=UTF8&ll=34.744434,-111.414692&spn=0.009098,0.016823&t=h&z=16
dirt trees?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=+35.1766+-105.9438+&sll=34.744434,-111.414692&sspn=0.009098,0.016823&ie=UTF8&ll=35.176641,-105.942761&spn=0.002263,0.004206&t=h&z=18
Big dirt circles pretending to be trees
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.82417+++-106.03444&sll=36.437099,-115.359109&sspn=0.008908,0.016823&g=36.4377++-115.3597&ie=UTF8&ll=34.826874,-106.035147&spn=0.018178,0.033646&t=h&z=15
I FOUND TREES!. hmm more than 100 feet away,
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.26306++++-96.63694+&sll=34.826874,-106.035147&sspn=0.018178,0.033646&ie=UTF8&ll=32.262965,-96.636592&spn=0.00117,0.002103&t=h&z=19
hmm might be a tree here somewhere or NOT.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.4047++++-111.5286+&sll=32.262965,-96.636592&sspn=0.00117,0.002103&ie=UTF8&ll=40.405956,-111.527828&spn=0.004216,0.008411&t=h&z=17
Gosh UHI here must be worse than NYC
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=39.9578+++++-111.7794&sll=40.405956,-111.527828&sspn=0.004216,0.008411&ie=UTF8&ll=39.958093,-111.779033&spn=0.004243,0.008411&t=h&z=17
huge windbreaks here
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.7908+++++-111.4077+&sll=39.958093,-111.779033&sspn=0.004243,0.008411&ie=UTF8&ll=40.790927,-111.407767&spn=0.002096,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=46.96+++++-116.855+&sll=40.790927,-111.407767&sspn=0.002096,0.004206&ie=UTF8&ll=46.962213,-116.85992&spn=0.007557,0.016823&t=h&z=16
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.3667+++++-106.105+&sll=46.962213,-116.85992&sspn=0.007557,0.016823&ie=UTF8&ll=42.368703,-106.106091&spn=0.002045,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.6338++++++-106.3775&sll=42.368703,-106.106091&sspn=0.002045,0.004206&ie=UTF8&ll=42.634764,-106.37774&spn=0.008146,0.016823&t=h&z=16
Finally a tree issue. or is it.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=28.9275++++++-98.7494+&sll=42.634764,-106.37774&sspn=0.008146,0.016823&ie=UTF8&ll=28.927383,-98.749489&spn=0.002423,0.004206&t=h&z=18
TREESSSSSSS
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.66667++++++-85.91667&sll=28.927383,-98.749489&sspn=0.002423,0.004206&ie=UTF8&ll=32.679524,-85.911938&spn=0.00466,0.008411&t=h&z=17
Oh what the hell,
From NOAA lat lons
HINT: in this I hint at what the real issue with many “rural” sites MAY BE
HINT: I started down this path when I looked at orland, CA, which warmed more rapidly than other rural sites in the early 20th century. Something changed there WRT land use.
Anyways for now its an untested hypothesis. The problem mikeC is you trusted NOAA metadata: MikeC
“I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…â€
Roll tape!
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.6025++-110.73444&sll=36.43821,-115.363124&sspn=0.002227,0.004206&ie=UTF8&ll=32.6082,-110.732478&spn=0.004663,0.008411&t=h&z=17
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=35.3694+++-111.5436&sll=32.6082,-110.732478&sspn=0.004663,0.008411&ie=UTF8&ll=35.3683,-111.54346&spn=0.018058,0.033646&t=h&z=15
trees? shrubs? windblocks? sunblocks?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.7433+-111.4139&sll=35.3683,-111.54346&sspn=0.018058,0.033646&ie=UTF8&ll=34.744434,-111.414692&spn=0.009098,0.016823&t=h&z=16
dirt trees?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=+35.1766+-105.9438+&sll=34.744434,-111.414692&sspn=0.009098,0.016823&ie=UTF8&ll=35.176641,-105.942761&spn=0.002263,0.004206&t=h&z=18
Big dirt circles pretending to be trees
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.82417+++-106.03444&sll=36.437099,-115.359109&sspn=0.008908,0.016823&g=36.4377++-115.3597&ie=UTF8&ll=34.826874,-106.035147&spn=0.018178,0.033646&t=h&z=15
I FOUND TREES!. hmm more than 100 feet away,
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.26306++++-96.63694+&sll=34.826874,-106.035147&sspn=0.018178,0.033646&ie=UTF8&ll=32.262965,-96.636592&spn=0.00117,0.002103&t=h&z=19
hmm might be a tree here somewhere or NOT.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.4047++++-111.5286+&sll=32.262965,-96.636592&sspn=0.00117,0.002103&ie=UTF8&ll=40.405956,-111.527828&spn=0.004216,0.008411&t=h&z=17
Gosh UHI here must be worse than NYC
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=39.9578+++++-111.7794&sll=40.405956,-111.527828&sspn=0.004216,0.008411&ie=UTF8&ll=39.958093,-111.779033&spn=0.004243,0.008411&t=h&z=17
huge windbreaks here
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.7908+++++-111.4077+&sll=39.958093,-111.779033&sspn=0.004243,0.008411&ie=UTF8&ll=40.790927,-111.407767&spn=0.002096,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=46.96+++++-116.855+&sll=40.790927,-111.407767&sspn=0.002096,0.004206&ie=UTF8&ll=46.962213,-116.85992&spn=0.007557,0.016823&t=h&z=16
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.3667+++++-106.105+&sll=46.962213,-116.85992&sspn=0.007557,0.016823&ie=UTF8&ll=42.368703,-106.106091&spn=0.002045,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.6338++++++-106.3775&sll=42.368703,-106.106091&sspn=0.002045,0.004206&ie=UTF8&ll=42.634764,-106.37774&spn=0.008146,0.016823&t=h&z=16
Finally a tree issue. or is it.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=28.9275++++++-98.7494+&sll=42.634764,-106.37774&sspn=0.008146,0.016823&ie=UTF8&ll=28.927383,-98.749489&spn=0.002423,0.004206&t=h&z=18
TREESSSSSSS
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.66667++++++-85.91667&sll=28.927383,-98.749489&sspn=0.002423,0.004206&ie=UTF8&ll=32.679524,-85.911938&spn=0.00466,0.008411&t=h&z=17
dirt trees?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=+35.1766+-105.9438+&sll=34.744434,-111.414692&sspn=0.009098,0.016823&ie=UTF8&ll=35.176641,-105.942761&spn=0.002263,0.004206&t=h&z=18
Big dirt circles pretending to be trees
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.82417+++-106.03444&sll=36.437099,-115.359109&sspn=0.008908,0.016823&g=36.4377++-115.3597&ie=UTF8&ll=34.826874,-106.035147&spn=0.018178,0.033646&t=h&z=15
I FOUND TREES!. hmm more than 100 feet away,
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.26306++++-96.63694+&sll=34.826874,-106.035147&sspn=0.018178,0.033646&ie=UTF8&ll=32.262965,-96.636592&spn=0.00117,0.002103&t=h&z=19
hmm might be a tree here somewhere or NOT.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.4047++++-111.5286+&sll=32.262965,-96.636592&sspn=0.00117,0.002103&ie=UTF8&ll=40.405956,-111.527828&spn=0.004216,0.008411&t=h&z=17
Gosh UHI here must be worse than NYC
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=39.9578+++++-111.7794&sll=40.405956,-111.527828&sspn=0.004216,0.008411&ie=UTF8&ll=39.958093,-111.779033&spn=0.004243,0.008411&t=h&z=17
huge windbreaks here
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.7908+++++-111.4077+&sll=39.958093,-111.779033&sspn=0.004243,0.008411&ie=UTF8&ll=40.790927,-111.407767&spn=0.002096,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=46.96+++++-116.855+&sll=40.790927,-111.407767&sspn=0.002096,0.004206&ie=UTF8&ll=46.962213,-116.85992&spn=0.007557,0.016823&t=h&z=16
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.3667+++++-106.105+&sll=46.962213,-116.85992&sspn=0.007557,0.016823&ie=UTF8&ll=42.368703,-106.106091&spn=0.002045,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.6338++++++-106.3775&sll=42.368703,-106.106091&sspn=0.002045,0.004206&ie=UTF8&ll=42.634764,-106.37774&spn=0.008146,0.016823&t=h&z=16
Finally a tree issue. or is it.
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=28.9275++++++-98.7494+&sll=42.634764,-106.37774&sspn=0.008146,0.016823&ie=UTF8&ll=28.927383,-98.749489&spn=0.002423,0.004206&t=h&z=18
TREESSSSSSS
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.66667++++++-85.91667&sll=28.927383,-98.749489&sspn=0.002423,0.004206&ie=UTF8&ll=32.679524,-85.911938&spn=0.00466,0.008411&t=h&z=17
dirt trees?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=+35.1766+-105.9438+&sll=34.744434,-111.414692&sspn=0.009098,0.016823&ie=UTF8&ll=35.176641,-105.942761&spn=0.002263,0.004206&t=h&z=18
Big dirt circles pretending to be trees
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.82417+++-106.03444&sll=36.437099,-115.359109&sspn=0.008908,0.016823&g=36.4377++-115.3597&ie=UTF8&ll=34.826874,-106.035147&spn=0.018178,0.033646&t=h&z=15
I FOUND TREES!. hmm more than 100 feet away,
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.26306++++-96.63694+&sll=34.826874,-106.035147&sspn=0.018178,0.033646&ie=UTF8&ll=32.262965,-96.636592&spn=0.00117,0.002103&t=h&z=19
trees? shrubs? windblocks? sunblocks?
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=34.7433+-111.4139&sll=35.3683,-111.54346&sspn=0.018058,0.033646&ie=UTF8&ll=34.744434,-111.414692&spn=0.009098,0.016823&t=h&z=16
Roll tape!
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=32.6025++-110.73444&sll=36.43821,-115.363124&sspn=0.002227,0.004206&ie=UTF8&ll=32.6082,-110.732478&spn=0.004663,0.008411&t=h&z=17
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=35.3694+++-111.5436&sll=32.6082,-110.732478&sspn=0.004663,0.008411&ie=UTF8&ll=35.3683,-111.54346&spn=0.018058,0.033646&t=h&z=15
Oh what the hell,
From NOAA lat lons
HINT: in this I hint at what the real issue with many “rural” sites MAY BE
HINT: I started down this path when I looked at orland, CA, which warmed more rapidly than other rural sites in the early 20th century. Something changed there WRT land use.
Anyways for now its an untested hypothesis. The problem mikeC is you trusted NOAA metadata: MikeC
“I pulled that metadata for each station in Petersons UHI study… and all but one of his rural stations had tree problems…â€
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=39.9578+++++-111.7794&sll=40.405956,-111.527828&sspn=0.004216,0.008411&ie=UTF8&ll=39.958093,-111.779033&spn=0.004243,0.008411&t=h&z=17
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=40.7908+++++-111.4077+&sll=39.958093,-111.779033&sspn=0.004243,0.008411&ie=UTF8&ll=40.790927,-111.407767&spn=0.002096,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=46.96+++++-116.855+&sll=40.790927,-111.407767&sspn=0.002096,0.004206&ie=UTF8&ll=46.962213,-116.85992&spn=0.007557,0.016823&t=h&z=16
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.3667+++++-106.105+&sll=46.962213,-116.85992&sspn=0.007557,0.016823&ie=UTF8&ll=42.368703,-106.106091&spn=0.002045,0.004206&t=h&z=18
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=42.6338++++++-106.3775&sll=42.368703,-106.106091&sspn=0.002045,0.004206&ie=UTF8&ll=42.634764,-106.37774&spn=0.008146,0.016823&t=h&z=16
http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=28.9275++++++-98.7494+&sll=42.634764,-106.37774&sspn=0.008146,0.016823&ie=UTF8&ll=28.927383,-98.749489&spn=0.002423,0.004206&t=h&z=18
Hmm might be issues there. see what you think
steven mosher, I see a whole bunch of links which look like the first three links you listed which I caught you on. So if you need me to catch you on some more of your angus pie, then put a station name or ID number. But I can already tell everyone your problem; you are using the 6 digit lat longs provided by NCDC which are from the NWS station manager estimating the location of the station from a topo map. That’s why you look for a station on Google Earth and your pins are in the wrong location… like in dirt or 100 yards from trees.
steven mosher, calling steven mosher, are you going to provide the station id numbers for those google earth photos or is this just another of your fake stunts… the first stations you linked the mms, now we dont get anything,… or have you finally stopped acting like a bug on a windshield and pulled your head out of your ass?
Station ID numbers mr mosher, lets see how bad you screwed up this time…I’m ready to looki up all of them, mr mosher
MikeC can you please post a picture of your father so we can do a Dad Size Comparison (DSC) analysis.
Wouldn’t work Mark, I’m always solid and he’s always gas. 😉
oh boy mosh, of the 51 links you posted 38 are duplicates… using the “big lie” strategy?
mikeC. no stuff caught in the spam filter, tried to chop it up. paranoid much.
dirt trees
easy Mike, do what I did.
go to the climate audit page.
copy the name
Go to NCDC. do the rest.
Or post all the correct lat lons since you claim to have dtermined thatonly ONE station is tree free.
Easy.