Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Wednesday, October 15, 2008

Chicago Marathon 2008

The data do not lie: The actual environmental conditions from the course

Apologies for the absence lately, it has been a hectic time at work and for other endeavors lately, but our deadlines have now passed and we can return to a more regular posting routine that we are accustomed to. Just in time, too, as the NYC Marathon is just over two weeks away, so watch for our previews of that one as Paul Tergat and Paula Radcliffe try to add to their previous victories in that race.

But for now let's look back to the Chicago Marathon, where much was said about the weather conditions on the day. All the fuss was due to last year's oppressive conditions which forced the organizers to close the course early and send people back to the finish. It was a cooking day in Chitown last year, to be sure, but this year it was a stunning day---and we have the data to prove it!

The historical record

A look at the past three runnings of the race reveal three vastly different days. In 2006 it was a miserable day---cold, in the 30s or 40s F, if I recall, and overcast the entire time. Generally a dreadful day to run a marathon! 2007 was quite literally burned into our memories - it was already in the 70s F at the start, with no wind, and glaring sun that baked an already hot city into the 90s F by day's end.

This year was cooler at the start, and much less humid as the day went on, producing a warm but dry day. Here is how the conditions stacked up from last year and this year, according to the data from the weather website Weather Underground:

These graphs show the "official" historical data on the Weather Underground site. Temperature data are on top, while relative humidity data are at the bottom. We took values from the same station, so there is no bias in that sense. The big difference is the starting temperature---it was over 10 F lower this year, and so even though the humidity was similar at the start I can tell you it was a vastly different day. And from 11:00 the humidity was consistently 10% lower than last year.

This year's weather was something runners know about. It was one of those days where you step outside in your running kit and think, "Hmmm. . .maybe I should bring a long-sleeved shirt with me to the start because I might get cold standing around before the gun." Last year, on the other hand, was muggy, hot, and oppressive. The preceding month was unusually warm for fall in Chicago, and every morning was hot and humid even before the sun came up. You never felt cool last year in September during your morning run or ride, I can tell you that much.

The "official" vs. the "actual"

But the official data tell only half the story, and after last year the race organizers knew they needed something more to inform them about the conditions on the course. The problem last year was that around 11:00 it was clear to the officials that they were going to have problems if they kept the course open as their "peripheral" resources, i.e. ambulances on the course, were all deployed and transporting runners to local hospitals. Therefore it was the best decision they could have made to close the course, because had anything else happened they would not have been able to respond to it.

Part of the solution was to invest in technology and purchase four portable devices to measure the temperature and relative humidity on the course, rather than relying on the data from weather stations. We placed these devices (together with faithful graduate students!) at strategic points on the course, namely the northern-most, western-most, and southern-most points and the finish area. Then we took readings every 15 min, although for simplicity we have included only hourly measurements here:

Again, temperature data is on top and humidity data at the bottom.

The first noticeable detail is the the readings we took on the course are quite different from the "official" data. This might not be a surprise as that weather station is likely distant to the race course, and weather is a very local phenomena. But still, as far as we know this is the first this this kind of discrepancy has been shown, and the important implication is that if you organize an event and want to know what is happening on your course, you must collect the data yourself and analyze in real time to see how things are changing and exactly what is happening.

During portions of the race the difference between the "official" values and the data we measured was as much as 40% for the humidity and 11 F for the temperature. In addition, the difference between the northern-most station, located near mile eight, and the finish was 6-7 F. The reason is probably because that part of Chicago, the neighborhood of Lakeview, consists of heavily shaded narrow streets. Compared that to the finish in Grant Park, which is totally exposed to the sun, especially for the first half of the day as it is eastern border is Lake Michigan.

So based on the data, one has to ask, "Why not run the course in the opposite direction?" After all, temps in Lakeview were 5 F lower when we stopped collecting data at 11:00 there! Of course we did not measure into the afternoon, and by then the temps on the northside might have been similar to the finish. But this is a good example of how science and data can drive critical decisions that might improve future events.

Was the weather a factor?

This question is being debated in the comments section for the race report, and we have people weighing in on both sides. Was it hot? Yes, it was, and the air temperature was close to what it was last year, in fact. The humidity was much lower, though, and the real difference was probably the starting temperature and the fact that the weeks prior to the race this year were generally much cooler. Overnight lows have been solidly in the 50s F since the beginning of September, and so the city never really heated up like it did last year.

The conditions were far from "ideal," but then again they were not dreadful by any stretch. It was a stunning day in Chicago, and normally wuhen the environmental conditions are considered "hot," it shows in the winning time. Evans Cheruiyot's 2:06:25 was incredibly fast, and not indicative of a winning time in "hot" conditions. That said, the conditions were not ideal for world record-type performances. On the day, in 'real-time', we wrote somewhat arbitrarily that the elites were probably slowed by about a minute as a result of the temperature, which seems reasonable, even now.

The early pace in the elite race was too fast, given the conditions. They paid for it in the second half, with the exception of Cheruiyot (who did also slow down, it has to be said), and large time gaps were the result of the super fast early pace, combined with the conditions. However, for the vast majority of the field, the conditions were not harmful, and the athletes running anything slower than...oh, about 2:08, were never, ever in danger of heating up too much, or becoming critically dehydrated.

The medical tent

I was sitting at the entrance to the medical tent, and in time we will analyze the data regarding symptoms, number of admissions, time to discharge, etc. One anecdote I can share now, though, is about a runner who must have run about a 3:15 or so, judging by what time he came to the tent. He was on the back of one of the golf carts used to get collapsed runners from the finish to the tent, and was sitting up with a Gatorade cup in his hand. As he passed us, he recoiled and spewed what must have been a beaker full of vomit. That in itself was not remarkable, but the volume was---the poor guy must have puked 500+ mL of Gatorade/water! Yet there he was, trying to drink MORE, probably because he'd been told he felt so terrible as a result of dehydration! Dehydrated, with maybe a liter of fluid sitting in his stomach... if something doesn't strike you as being "wrong" with that picture, then nothing will He had no business with that kind of volume in his stomach, and in case you missed it last year you should read our series on dehydration and fluid intake. Just click the "Featured Series" tab above for the links.

In the meantime feel free to weigh in, especially if you ran the race, and watch for more data from the medical tent as we analyze that!



Anonymous said...

I believe from the context of the text that in the first two graphs you have 2008 and 2007 reversed. Your graphs say 2008 was hotter, while your text says otherwise.

Mike LaChapelle said...

I recently ran a 3:37 marathon on a cool, sunny fall day (start temp 45 F, finish 60 F). Although I stopped to drink 1-2 cups at each water stop (every 2 miles), I still cramped up in the last 2 miles due to dehydration (measured a 7 lb water loss after the race, 4% of my body weight). I feel like if I don't drink constantly throughout the race, I will end up in trouble at the end. After I finished, I tried to rehydrate and did end up throwing up. Any suggestions on alternate hydration strategies?

JC said...

1) yes, the graphs are definitely switched and threw me off for a bit 2) sounds like you need to take in some sodium via gatorade or salt caps ect to help your body to retain water better

Chris said...

It is not important to consider wind speed? It seems convection plays an important part in keeping the body cool.

You mention in the post that there was "no wind" at the start of the 07 Chicago Marathon. However, Weather Underground's (WU) record shows a SSW 4.6 MPH wind. WU also recorded a similar wind at the beginning on the 08 Chicago Marathon. It appears from WU that for the first 2-3 hours of each 07 and 08, wind speeds were nearly the same, but became strong in 08 later.

Ross Tucker and Jonathan Dugas said...

Hi Anonymous, I think you must have been the first one to see this post! Initially I was having problems arranging the graphs they were swapped for a bit as I tried to sort it out. This was corrected shortly thereafter and now the graphs match the text.

Thanks for letting us know, though!

Kind Regards,

Ross Tucker and Jonathan Dugas said...

Hi Mike,

Well done on your recent 3:37 marathon!

We encourage you and other readers with similar questions to read our series on dehydration and our series on muscle cramps. The one-line message for you is "drink to thirst," and we explain all of the concepts in both of those series.

On another note, I wonder when you weighed yourself after the marathon? And also if you weighed yourself immediately before? Because by some rough approximations I have you ingesting a total of nearly three Liters of fluid and still finishing with a 4% loss:

Average 1.5 c per station
Average 150 mL per c
Therefore 225 mL per station
13 stations, therefore 2.925 L
or about 830 mL per hour

That means you were sweating close to ~1.8 L per hour to finish with a 4% weight loss, which even for an 80 kg man seems a bit high given the environmental conditions and your race time.

Regardless, the better advice is to forgo drinking on a schedule and rather to follow your thirst.

Good luck in your next race, I hope you do not experience any vomiting and run a PB!

Kind Regards,

Ross Tucker and Jonathan Dugas said...

Hi jc, and thanks for your comment.

In fact no one needs to ingest sodium during exercise. You body retains water just fine when it needs to, and especially during exercise.

We suggest you also check out our series on dehydration as we explain why sodium ingestion, especially in the form of Gatorade or any other sports drink, is not necessary.

Of course, ingesting Gatorade for the carbohydrate it contains is beneficial, so you should not confuse the need to ingest carbohydrate with fluid/sodium needs.

Kind Regards,

Ross Tucker and Jonathan Dugas said...

Hi Chris,

Indeed, wind speed can influence your ability to lose heat to the environment, so you are correct that this is important.

However, the wind speeds from the many weather stations around the city are incredibly variable and living here I find that in the city it is hard to get any consistent wind unless you are on the lake front.

The weather station located about one mile off shore was reading wind speeds of 8-12 mph the entire day, while those in the city were variable. For example, I recorded wind speed data from FIVE "official" stations around the city, and at 12:00 the readings were as follows (in mph):

Station A: 0-1
Station B: ~3
Station C: 0
Station D: 1-2
Station E: 0
On the water: 7.6

It is unclear on the WU site where the historical data are measured, but my guess is that it is from either O'Hare or Midway, our two airports, which are considerably inland and not buffeted by the buildings in the city like the other weather stations are. This is the only way I can explain these differences, besides that fact that I also do a bit of sailing and can tell you that wind is an extremely local phenomena, sometimes very different at two ends of a one mile course.

Thanks again for your comment!

Kind Regards,

Anonymous said...

As per your anecdotal runner, certainly fueling stations could be improved in a manner similar to that proposed to race direction based on the temperature data.

I think too many races tend to have more fueling stations later in a race at a time when the field is thinner (making using every station easier) and at a time when concentration and judgement is lapsing.

I also think that too often people are corralled through the finish, past the food tables, and have to make a snap decision on how much liquid and food they will need in a single pass. Typically this choice has to be made within moments of finishing.

These factors promote massive fuel consumption while the heart rate is still high. It is also counter productive in promoting cool-downs.

My guess is that the 3:15 runner didn't have a litre of fuel in his stomach at 20mi, but that in the final stages and especially the first few minutes after the race ended, his consumption far outweighed his need.

Chris said...


Agreed on the wind being a very local condition. I too live and train in Chicago and am amazed how I always find the headwind :)

I did notice that the wind seemed to die later in the race this year.

Aside from the weather, another thing I noticed about this year's race is that the water and Gatorade cups were filled more than I have ever encountered--they were at least 3/4 full, if not more. I found myself grabbing a cup, pinching the top and pouring at least half of it out before trying to take a sip. That's likely part of the response from 07. Just an FYI.

Jen said...

I was shooting for a 3:55 and ended up running 3:55 in Chicago last week and I was very pleased. I thought the course was "comfortably hot" in the beginning and there is plenty of shade throughout the course that made a huge difference for me. It was noticeably hotter near the end and I could feel the difference but I was also very tired and the perceived exertion was greater in the heat near the finish. I personally do not drink gatorade. I drank Hammer's Heed in my own throw away bottle over the first 5 miles and could dink when I wanted to without relying on aid stations. After that I relied on the ample water provided but only drank what I need to from the cups and did a gel with water every 5 miles for my 30-40 g of carbs for the hour. I find that I need to take the electrolyte capsules, 2-3 each hour, and felt they helped me cope with the heat. In cooler weather (with similar long races), I do not need the endurolytes. I was still covered with salt on my black shorts at the finish but then again, the endurolytes have more in them than just salt.

J.P. said...

Gents, thanks for all your work on this site. I appreciate all the effort and accessible information.

Here's some anecdotal data for your further consideration. I've run 13 marathons between age 29 to 41. The first six or seven I had no cramping issues, running them all between 3:03 and 3:30. The last 6 or seven, I've cramped in all but one. My anecdotal experience are as follows:

- Heat does not seem to have played in role in whether or not I cramp.

- I drank too much fluids in some of my early efforts, but more recently I've have been drinking to thirst, consistent with your advice. (I started doing this because I just felt awful when I drank too much).

- In the first 6 or 7 when I didn't cramp, my training was usually not very intense, rarely exceeding 40 miles a week. I often felt completely exhaused, weak and fatigued late in the race. I was out of gas in these efforts.

- The last six or seven marathons, I've generally trained more and smarter (mileage up 6o miles a week a few times). I've not run out of energy in these. I've run these between 2:59 and 3:19.

- So for me, cramping has occured as I have trained more, gotten older and run consistent splits late in the race (as opposed to slowng). Do you think age or additional wear and tear on the legs has anything to do with cramping? I note that with additional training, I've been much more comfortable running faster times, and I don't feel as weak and fatigued late in the race.

- In the end, I don't know if there is a causal relationshp, but I've been more vulnerable to cramping as I've gotten older, trained more and run faster miles at the end of the race. The additional training has made me much more comfortable during the race and overall much more able to run faster at the end of the race. Of course, it doesn't matter much if you are running faster mile splits at the end if you have to stop for 3 minutes to let a cramp unclench!

Nuke Runner said...

I think it would be much more usefull to display dewpoint rather than relative humidity. Dewpoint gives a much more meaningful understanding of the amount of moisture in the air.
The differences between the official NWS observations and the course observations could be due in part to measurement techniques. The lower relative humidities indicate the course readings could have been influenced by background radiation.

Sarah L said...

I collapsed with heat exhaustion and dehydration and was in the medical tent at the end of Chicago... (temp of 105!! How?!) I'd love to send you my story - but I can't find a way to email you or get in contact! I'm on slintern@hotmail.com - love to hear from you