- Data:
- Race results from
- Vermont City Marathon website
- Coolrunning website
- Weather data
- noaa.gov National Ocean and Atmospheric Administration
- Article:
.
This article was written for Level Renner
It appeared in the July/August 2016 - Race results from
- Notes on Analysis
- Measuring WBGT
The WBGT index is measured with a device with three different types of thermometers.- Dry Bulb (Td) - normal ambient temperature
- Wet Bulb (Tw) - sensitive to humidity
- Globe Bulb (Tg) - a black bulb which is sensitive to direct sunlight / radiant heat.
The index is then a weighted average
A good source of information on WBGT is: here
WBGT = 0.7 Tw + 0.2 Tg + 0.1 Td
Two WBGT monitors. - Estimating historic WBGT
A major difficult for this analysis is that historic weather data only includes Dry Bulb temperature and Relative Humidity. From Relative Humidity you can calculate the Wet Bulb temperature. That gets us 80% of the way to WBGT, but doesn't include the effect of direct sunlight.
Fortunately there are standard methods for estimating WBGT from just Dry bulb temperature and Relative Humidity. The technique many place including the Australian Bureau of Meteorology
WBGT = 0.567 x Ta + 0.393 x Vp + 3.94
where:
- Ta = Dry bulb temperature (deg-C)
- Vp = Water vapor pressure (hPa) [humidity]
The vapor pressure can be calculated from the temperature and relative humidity using the equation:
Vp = rh / 100 x 6.105 x exp ( 17.27 x Ta / ( 237.7 + a ) )
where:
- rh = Relative Humidity [%]
- Figures:
In the following figures appeared in the article.
Click on thumbnail to get full figure.
Fig 1. Histogram of number of finishers every 5 minutes for 20 years of racing.
Fig 2. Sum of 20 years. Men blue, women Red. The spikes are just before the 4:00 and 4:40 mark.
Fig 3. Average finishing time vs. WBGT heat index.
Fig 1x. Winning Times (men) vs. Temperature The colder the weather, the better the times, but a correlation coefficient of 0.34 is not too strong. I expect a fast race means that two fast runners showed up that day and battled it out. (Yes I could look at the data to figure this out.)
Fig 2x. Winning Times (men) vs. WBGT Index The correlation is about the same as in Fig 1x. This is due to the high correlation between temp and WBGT seen in Fig 5x. But the major difference between this figure and the one above it is the range of the temperature scale.
Fig 3x. Average Time vs. Temperature 
Fig 4x. Average Time (men) vs. WBGT Index 
Fig 5x. Temperature (men) vs. WBGT Index What this
- Measuring WBGT
