[This started as some graphs, accompanied by an explanation, which grew into this: An object lesson in overkill.]

Edit: Graphs added

Hypothesis:
I was interested in finding out how much hotter the temperatures on the road are for bikers than are reported by local weather services. I predicted that at about 3 feet off the ground, the temperature would be around five degrees Fahrenheit warmer than weather service readings.

Setup:
For four days (June 20-23, 2005), I biked my normal 12.6-mile route to work, a route through Phoenix, Paradise Valley and Scottsdale, Arizona, that consists mostly of asphalt streets with some concrete, asphalt and dirt paths along irrigation canals, and passes through both urban and residential areas, including some flood-irrigated areas. Attached to the rack of my bike at about 3 feet off the ground was a data logger that measures temperature and humidity levels. The data logger was shaded by a stiff piece of paper wrapped around its base, so as to prevent direct sunlight from striking the thermometer. To estimate weather service readings, I used a weather station at work that is about 20 feet off the ground, is shaded, and is at a substantial distance from asphalt. A problem with using this weather station is that it is situated in an open, agricultural area, which skews the temperatures both higher and lower than in the areas biked through. Past comparisons with local weather service readings, however, show the weather station to be a reasonable estimate of weather services for the areas biked through.

Findings:
My prediction was too high for the morning rides and incorrect for the evening rides. The attached graphs show readings from my morning and evening bike rides, along with readings from the weather station taken at the same times. Rather than use clock times, duration of the bike ride in minutes is used, so as to allow comparison between day and evening rides. Morning rides begin at a distance from the weather station and get closer as the duration increases; evening rides are the opposite.

The graphs show that, in the mornings, the road temperatures are around two to four degrees higher than the weather station temperatures. Also, because different terrains receive sun at different times, and have different rates of heating and cooling, there is a large amount of fluctuation as the bike passes through different areas. However, in the evenings, there is little difference between the bike's and the weather station's readings, and little fluctuation. These results suggest that the roads retain heat better than surrounding areas, so the morning road temperatures are higher. But, by late afternoon, temperatures of the road and the surrounding areas have reached an equilibrium, and the fluctuations disappear.

The evening readings from the June 21 graph show a large gap. This exception is the result of a sudden drop in temperature (about 10 degrees in 10 minutes) in the surrounding air just prior to the start time, so that the results look like the morning readings, with the road radiating heat while the air around it has cooled. The inverted character of the morning readings from June 23 (higher surrounding air temperatures, lower road temperatures) are likely the result of substantially different conditions in the areas distant from the weather station. Note that as the bike approaches the weather station (toward the end of the ride), the bike and weather station temperatures assume their normal gap.

Conclusions:
The findings show that, in the late afternoon, when temperatures are hottest and conditions for bikers most extreme, the temperature on the road is likely to be close to the reported temperature, in spite of bikers' experiences of feeling hotter. Bikers' perceptions are likely due to their exposure to the sun, humidity, and high activity levels, rather than a heat radiation from the road. In the mornings, the road radiates heat absorbed the previous day, so that it is hotter than reported temperatures, while heating and cooling rates for different terrains leads to a high fluctuation of temperature.

Such a geek-tease! where are the graphs? I need graphs and data! I had been thinking of running a similar experiment through thermodynamic computer simulation, and since I got hit by a car and can't currently ride (despite the fantastic freaking weather...), this may be the time to do so. I'm curious to see your data/statistical analyses.

Ack! They didn't upload. I'll edit and add attachments.

thank you. I thought I was going to geek-explode.

I think my post was long-winded enough without data. If you want it, I'll send you the Excel file. I didn't do any statistical analyses to see if the road temps vary significantly from the wx station temps, but the data's there if you want to.

boy, seeing that data makes me REALLY love the poconos, despite the harsh winters... People here are kvetching since we've had 3 days over 90!

Nah, I don't think that I could find anything conclusive from this... Your data is sort of all over the place, and there are many factors that affect the relationship in question, many of which are probably functions of one another.

I had thought of doing an engineering analysis of the relationship between several environmental factors, ground temperature, material, etc. as a building systems engineering study. In order to isolate one variable at a time to see the effect on the other variables (and to be consistent with my technology budget of $0.00), a computer analysis is probably the best choice, given that it's the only available method within my means.

I skimmed through your post. Neat idea. So, was the data logger insulated from the asphalt below by your rack? (does it matter?)

No, it wasn't insulated, and no, it doesn't matter. Direct sunlight is one thing that matters, because the unit gets hotter than the surrounding air.

Thats great! I gotta give myself some time to think about it all.

Al

Moved to General Cycling Discussion as per request by the OP.