As can happen with blog posts something contentious turns up, it won’t lie down, I was trying to move on to other things. Roger Andrews emailed me about some movement results he has unearthed, which relate to this earlier post on UHI at Heathrow Airport.
No head image here, are too large, read on to see.
Figure 1
Image from report here.
Figure 2
Plot produced by Roger Andrews. Note: right hand scale should be the same as figure 1.
Figure 3
Extracted the movement data and added an overlay plot. Sure hope I have the X axis of the whole thing spot on. The movement data is spot day figures.
So what is going on?
I can understand plotting a mean won’t show anything, including given the poor underlying data resolution. It is part day only.
What does “movement” mean?
100+ tonnes of fuel fully laden or landing empty?
Which runway.
Which direction.
Freight or passenger?
Active flight?
Easter is in there too where we know record traffic was broken and seems to be in the temperature.
A possibility is the heat is from something else. I did mention the close proximity of the passenger terminal vehicle access tunnel, for those who don’t know it is major dual carriageway. Then we have people in buildings, were sleeping there so services were very active.
Doesn’t make sense. Ideas anyone?
[update]
[/update]
[UPDATE 22nd April 2012]
Figure 5
I’ve now created new software which creates a patch plot, done entirely under my control, no guessing or assumptions by third parties. This is a work in progress so where it will go from here remains to be seen.
The above uses dark grey for missing data, lighter grey for zero, the rest is obvious. No scales. A patch is for one hour during one day. Runs 1st April to 30th April, zero hours at the bottom and 23hrs at the top . Movement data as RA, is spot values for each day.
I might now be in a position to unearth detailed runway usage information. There is a pattern of usage with direction and takeoff/land switched to a schedule but there is variance on practice.
[/UPDATE]
Post by Tim Channon, co-mod
Tim
The graph I sent you shows aircraft movements but I just noticed that the scale I linked them to doesn’t. It shows NOx concentrations (I thought at the time that only 200 flights a day in and out of Heathrow seemed a little low). The scale I should have used is the one on the right hand side of the first graph. Sorry about that.
Do you know exactly where the thermometer is located at Heathrow?
OK, I found it.
A met enclosure is linked in the original post. Absolute confirmation is near impossible to obtain.
Some time ago I spent a long time locating met enclosures as part of a project so I do know, long story.
Try pasting this into Google Earth, or http://maps.google.co.uk or http://www.bing.com/maps
51.479232° -0.450652°
Best image is Bing, just. No G. street view, might be Bing but I suspect a wall is in the way, incompatible here, ah, oh well, can sit on it. I know there were video frames taken by passengers but I have no video here, or at least not Adobe.
Fantastic, thank you Microsoft. See update, we have an image of the met site from ground level.
At the met site and to the east for several hundred metres the barrier to the road goes down to ground level.
Immediately to the east the barrier has couple of feet or so gap at the bottom which continues east for 100 metres or so, then to ground level again. Given this is often just after V2 there will be maximum takeoff thrust angled downwards. However, usage to the east will be less common.
Betcha reverse thrust on west usage will be roughly there.
Sure hope they have anemometer elsewhere. Seems likely given the usual tall mast is not present.
That looks like it will channel jet blast at the met site.
Between the building and the met site is a standby generator. (unimportant)
The building whatever it is has equipment on the roof which looks like cooled equipment, so whatever is in there is heat dissipating. I don’t think this is a big deal yet this is relatively recent, is not present on older images.
Looking at this view
http://binged.it/LLpAk5
You have a nice 3×3 meter square of asphalt right next to the station, plus a road, plus a big car park to the North, plus the tunnel cap to the west.
What you may be seeing is simply sunlight effects on these surfaces, which would have a significant impact. I’d look for cloud cover reports for those days.
Are you both using the same temperature datasets? I’m guessing not.
The big differences between Heathrow and Northolt in Tim’s plot don’t seem to show up in Roger A’s. Adjustments?
I think the dates Tim has put across the middle of the plot are shifted one day left.
Here’s an aerial photo so we can understand the layout better.
Anthony, Perhaps digging out the other data will be necessary.
Rog, just exploded trying to figure that out.
I’ll have a think.
TB:
I took my data off the spreadsheet Tim provided:
http://tallbloke.files.wordpress.com/2012/05/heathrow-uhi-1.xls
The only tweaking I did was to combine each set of 24 hourly readings into a daily average for direct comparison with the aircraft movement numbers. I didn’t use any hours when one or other of the two stations was missing a reading.
Hi RogerA: Thanks, I guess the cooler nights at Heathrow offset the higher days during April 2010 then? Could you plot mins and max’s for the two locations?
Ulric just sent me a link to some data, does this match?
http://www.tutiempo.net/en/Climate/Londres_Heathrow_Airport/04-2010/37720.htm
Reached the one-foot-ine-the-grave state. Anything else going to go wrong, get in the way, break?
I’ll go and see if the paint is dry. (where he will discover yes, and a rash of flies tested first) One of those weeks.
Wassup Tim? Hardware hassles?
TB:
http://oi45.tinypic.com/23i9ch4.jpg
Average max. temp April 2010: Heathrow 15.2, Northolt 15.2
During Heathrow shutdown: Heathrow 15.3, Northolt 15.3
Average min. temp April 2010: Heathrow 5.6, Northolt 4.9
During Heathrow shutdown: Heathrow 4.4, Northolt 3.9
Average max-min April 2010: Heathrow 9.7, Northolt 10.3
During Heathrow shutdown: Heathrow 10.9, Northolt 11.4
Thanks again RogerA. Hmmm, the temperature differences are pronounced during flying times, so we need to think about where the important differences lie in the longer term records.
Something just dawned on me. Is everyone sure “movements” is being interpreted correct here?
Movements went negative just after the no-fly days. I would assume that this “movements” simply means many more flights were “out” than into Heathrow during these days. A flight “in” (+) should have much lower effect of the temperature station when landing than when “out” (-). Am I interpreting this correct?
I think fights in and flights out definitely need to be taken into account.
Massive reduction in ‘movements’ because of airborne ash from the Icelandic volcano with the unpronounceable name: Eyjafjallajökull ( praise $DEITY for cut and paste
May have had something to do with insolation levels too…
DC: Yes, I’m wondering if the grounding of flights led to a reduction in cloud amount over the airport. That would lift temperature inthe daytime. The no-fly nights were cold. As Tenuc said on the other thread, all the water vapour from the engines creates a microclimate. I think Anthony Watts earlier suggestion to try to find cloud data is a good one. Photos taken around the airport over the period would help a lot.
I could swear I saw Shergar on one of those photos
[Reply] Being ridden by Lord Lucan I think.
I apologise for the delay. How long I don’t know.
Questions are being asked about what exactly constitutes an “aircraft movement”. It works like this:
1 takeoff = 1 aircraft movement
1 landing = 1 aircraft movement
1 takeoff + 1 landing = 2 aircraft movements
These relationships hold true for all values of takeoffs, landings, aircraft and movements.
The above is a patch image created by new C code from the April 2010 XY data in the spreadsheet copied into a text file and interpreted accordingly. The colour mapping is jury rigged minimal. Key features in the XY data check against image XY.
Missing data is grey. Zero is light green. Positive is increasingly red. Blue accordingly.
I’ve had to reverse out of a variety of technical problem here until I am sufficiently certain I can trust stuff. It’s been horrid, discovering bugs in things which I assumed were solid production (not my code).
Does this match the previous plots? No. Bizarrely it does kind of match, some exact, some not.
I am pointing the finger at unknown behaviour of gnuplot and I tried to figure that out, no dice. Pure guess, it is somehow interpolating strangely.
I’m not particularly worried about possibly having misled anyone because this is not a new work and the spectrogram tells a tale.
What next?
Not sure right this moment.
I’ve unearthed some flight archives which seem to include runway usage.
Added Figure 5 to the article. This corrects earlier plots, without substantial difference.
I think I now have the data needed to reconstruct runway usage to individual flight resolution. This would be a large and difficult task so I am not sure whether I am going to do it. Perhaps it is a “fun” exercise proving it can be done from public data.
Doubt there is enough data available to go a step further and estimate fuel burn, would need load factors.