Heat Impact on Run - An Update and a Vision
If you are short on time, go to the end of this post to see the Race Day Run Pace Modifier application that I think we can create.
This is an update to the prior post on heat impact on IM run times. The first post is here.
Some people asked to update the numbers with the data from the hot IM LOU 2010 race. I have done that and taken a couple of different views of the data and made some added observations on the impact of heat to the run.
First let’s see how hot IM LOU 2010 was on the run. I used the following site to capture the hourly temperature for the run for each of the 3 years. So we have a consistent set of data to compare. The heat index in the following table is a result of the temp and dew point for each hour.
So we had a slightly warmer run in 2010 than 2008. For all finishers, the average run times for the three years are: 2008 5:26:22, 2009 4:48:51, 2010 5:25:52. So the overall field for the two hot years had the same average time. Calculating the average run times by division shows that the two hot years tracked very close to each other. The following charts show average times for the divisions by year. We see similar impacts of the two hot years for both the men’s and women’s field.
I decided to take a look at the percentage change in times from the hot to the cool years. My hope was the being able to apply percentage factors to a base run pace to offset for heat would be more useful than just adding a defined time element. The table below has the numbers followed by a chart of the percentage change by divisions. At the aggregate level the heat impacted average pace by ~13% for the two hot years.
In the first post on this topic I saw a tendency for less of a slowdown in the older divisions. Combining the two hot years of data for both men and women and tossing in a trend line we get the following.
Ok what are the potential explanations for this? One thought was that a higher percentage of racers in the older divisions had more IM race experience, more years of base, and fewer first timers. So potentially they may execute better on average in the challenging conditions than other age groups. This could be real, but hard to validate.
The second thing that occurred to me was that form the post on age impact on speed. I found that once we hit that 50 year number we get significantly slower. So for us older dudes, even when running at max capability, we are slower than our younger buddies, no surprise here. So what does running slower mean on a hot day? It means they are producing less muscle power per unit of time than someone running the faster pace. Therefore less heat is generated which has to be eliminated from the body. We all have the same rough body temperature cut off. Once our core temperature hits ~40 degrees C we stop! Those running slower, ie producing less heat will take longer to reach the shut down limit all else being equal. I can get my head around this as a viable explanation for why the older divisions slow down less in the heat.
OK let’s go to the pointy end of the spear. The pros they slow down significantly less than the normal mortals. Why?
Hang in there with me I am headed somewhere I promise!!
Let’s get back up out of the data to the 10,000 foot level and ask the question of what are the primary factors that impact IM running performance in the heat. Here is my list based on what I have seen on the topic. Let me know what I missed!
- The temperature / humidity during the event.
- Relative level of training / fitness appropriate for the event.
- Level of acclimation to the heat.
- Hydration and electrolyte levels.
- Body size, mass to surface area factor. (smaller peps cool better do to more surface area)
- Core body temperature when climbing off the bike to start the run.
- Age > 50.
Looking at these factors, the pros come into the hot day with most of these factors stacked to their advantage. They should slow down less that others in the field that are not as well prepared. My bet is for the pros that blew up, they cooked themselves on the bike and got their core temps too high and were behind on hydration. Once they hit the run it was all over. The only way to get the core temp back under control on the run on that day was to stand down for ~30 min.
Ok for all of us trained EN Ninjas which factors above can we control? We come into the event well trained so 2 is off the list. We have and follow our proven nutrition/hydration plan so 4 is off the list. We pace and execute the bike smartly so 6 is a diminished factor.
So what is left? Temp of the day, our level of heat acclimation and our body type and age. Plus how well did our bike go. I think with some work we can create the following Race Day Run Pace Modifier!
This application will enable a person to start with the base pace from the EN run pacing charts and by selecting from the drop downs calculate a suggest race day pace based on the factors selected. With this in hand they will also have suggested paces pre calculated to take into account a bike that had complications. An example:
First does any of this make sense??? What did I miss???
If the concept of the Day Run Pace Modifier resonates with you, I have some ideas as to next steps at getting the needed data to flesh this model out. I am a strong believer in the power of teams. I would like to get the involvement of say 3-5 others to help make this a reality. Once we get through the process, I would suggest it be submitted to RnP for a complete review and modifications before publishing to the team.
If this concept makes sense to the team – Then if you are so inclined to help with this effort, PM me. I will coordinate this in our off season.
One big wet sock against the wall – let’s see what sticks.
Matt
0
Comments
As well thought out as your race was !!
One cool observation: Trained for the Marine Marathon in Aug/Sept in West Palm Beach in '08. On race day (in Washington DC) race day temps started off at 42 and I ran a PR marathon that day...coincidence? I think not.
In any case, heat can and does play a huge role in performance, so having a modifier is a great idea...now just need to tap into the power of da haus to collect some data then build a model.
I'd love to join you and play along on Oct 9th of this year as we will probably have a rather warm run to test this out on.
P - Doubt the temps in Madison will deviate enough from the baseline to allow for testing. But we do have a race in five weeks which will serve well. I intend to take two new elements with me to that race: cool packs for T2 and the Energy Lab, and a realistic pace based on the temps I'm actually encountering.
Matt - some random thoughts. First, don't you have a race to get ready for - now less than five weeks away??! Hope this goes on hold until end of Oct!
Second, regarding factors which affect performance in the heat: I'd add a few more: amount of sunlight - e.g., cloud cover, daylight vs night, elevation of the race (thickness of the atmosphere). Intensity of radiation will affect skin temp and thus need for and speed of cooling. I don't know how to quantify this, but I do know from experience it has a big impact. Also, one's tested sweat rate - bigger sweaters may need slower pace to allow for increased hydration. WInd, and thus the speed of fluid loss from evaporation, might also be a consideration.
Third, riffing on the age factor, I got to wondering if looking at a subset of AG times, say the top XX% in each AG instead of all comers, might be helpful. Presumably, we want to know what people who were actually RUNNING were doing. Taking all comers might skew the projection to suggest going slower than the successful racers do.
Fourth, I presume one of the anticipated tasks is to look at other races which have several years of data in significantly different temps to compare, e.g., Canada, CDA, Wisconsin. Hawaii is not a helpful data set, cause it's always the same friggin' weather there!
Fifth, when this comes out, it should be included as part of a strategic plan for racing in extreme conditions.
I will for sure PM you, but I truly hope this goes on the back burner (heh-heh) until after you get back from Hawaii.
Al, Once I got all my logistic needs taken care of for the trip to Kona, I had time on my hands this week being in recovery mode. Yes further heavy lifting on this topic will wait until the off season. In the next few weeks, I will be spending time practicing my transitions!
This is amazing. Nice work, Matt! I will look at this more closely after IMWI.
Matt, This is great stuff and has potential to be extremely useful to ironman triathletes. My preliminary thoughts are borrowed from Noake's Lore of Running. The environmental factors influencing an athlete's ability to lose heat are humidity, air temperature, wind speed and radiant energy load. Radiant energy load is warmer objects (sun and hot pavement)putting heat into cooler objects(the runner). So I would definitely add humidity to the estimator. Consider adding expected wind speed and whether there is expected cloud cover. Anyway, great stuff! I look forward to following your estimator's evolution into becoming another useful tool in EN's armamentarium of triathlon knowledge.
"Armamentarium" is secret doctor talk. I bet Bret's a physician. Maybe even a surgeon.
@Al. Use of any word greater than two syllables rules me out as being an orthopedic surgeon! Having to look up the spelling of the word armamentarium rules out me being an internist. I did take a semester of latin in high school, though.
I took 4yrs of Latin and 4yrs of Greek (ancient type) in high school and I'm kinda stumped
.
Matt, great work! Very excited to tighten all of this up and add it to the wiki as another component/tool in a s ystem for racing Ironman.
ar·ma·men·tar·i·um/?ärm?m?n'te(?)re?m/Noun
1. The medicines, equipment, and techniques available to a medical practitioner.
2. The resources available for a certain purpose: "the entire armamentarium of electronic surveillance".
Matt, you are a awesome. This is brilliant. This goes well beyond Daniels table for temp adjustment where a ball park VDOT 52 experiences a 3sec/mi slow down for every 5F heat index above ideal running temp of 60F.
I will be re reading your post over and over to slowly consume all of it. I don't think that you can address all variables simply due to complexity, but I think that you are well on the way to account for the most significant factors and come up with something very practical. Now, I can test any data you want verified and provide you with Garmin files. For every workout I do, I document water temp if swimming, temp, heat index, wind for bike and run. I do notice a wide variation in run paces affected by heat/humidity combo.
Nice work. Keep it up.
i had missed this thread previously and just glanced through it now....
looks awesome!!!!! matt is 'the man', in the good sense of the term vs. the bad sense (i.e. the guy who keeps a brotha down).
coaches: keep me in mind for guinea pig status, whenever you begin accumulating data. i live in sunny socal and have medical knowledge, so i won't die during testing.
GH
This is just the way my engineer brain looks at stuff. Drives my wife nuts!!!
3 days after I get back from Kona I get on a plane to India for 9 days OH fun!!!. In the 25 hours it takes to get there I will have some time to outline an approach to the data collection, validation and interpretation process we can go though to get to our end game.
When I get back, I will pull together a small team of 3-5 to bounce the next steps against. I hope from this we will have a direction of how to use large sample date (many race results by age group) and inputs from single performances ( aka the EN team).
I am confident that WE can produce a next generation pacing advisor for IM marathons.
Matt