>Technical Support>Confusion on Elevation Tools
Somewhat new to RA, so maybe this is a dumb question....but....For the past few years I've used Garmin Connect to store my run data. As of October, I have started using this site. Since my watch automatically uploads to Garmin Connect, I actually am now keeping TWO records.
Anywho...I am curious why I experience such a large elevation difference between the two sites when I upload. Do they work differently? I guess in my mind, I would think that watch data is watch data, regardless to where its uploaded to. Maybe I'm wrong about that.
Example from two trips on one of my normal courses:
Garmin Connect : 979' (Correction Disabled) 915' (Correction Enabled)
Running Ahead : 1622' (Before toggle/resample) 1245' (After toggle/resample)
Garmin Connect : 815' (Correction Disabled) 925' (Correction Enabled)
Running Ahead : 1269' (Before toggle/resample) 1207' (After toggle/resample)
So in other words, even with the little changes that are always happening with GPS, RA consistently shows between 300-400' more in elevation. (BTW, what exactly is toggle resample?) I realize that watches aren't perfect, and there will always be small differences, I'm just confused why there are such LARGE differences that are at the same time so consistent with each site!
Thanks for any info anyone can provide.........
The GPS elevation is useless, which is why you should always enable correction. Elevation correction works by looking up each data point's elevation in a reference elevation data set. There is no one official elevation data set because the data can be obtained in multiple ways. Data sets also come in different resolutions (i.e. the elevation is measured at every 10, 30, 90 meters, etc). Even if two sites use the same data set, the elevation can still be different depending on the algorithm used to interpolate the location.
Long story short, while elevation profile provides you with an idea of your route, it is by no means the definitive elevation.
Derek, I agree with eric in that most gps elevation data can be, ah, strange. Here's a somewhat more detailed explanation of why the variations of interpretation of the same data: http://www.topofusion.com/climb.php
I once did about a 1500ft in 4mi run (road, fairly steady gradient) with 3 gps receivers - and got 3 very different results for amount of climb. So I tend to take elevation data with a grain of salt - unless it's consistent with known information. Two of them had barometric altimeters, although I don't think the Foretrex 401's barometer was used with the satellite data.
That said, I was doing one of my hill workouts today (2/8/2013, I think I made the map public) so took the time to be sure that not only were the satellites locked, but the beginning elevation had stabilized before starting (gave the barometer a chance to figure out pressure and lock in with satellite data to get elevation). It's a garmin 910xt. (FR305 elevation data is frequently garbage.) The top of the hill has a benchmark, but I just had to go with gps readings with the gps at the start. Both the elevation profile from the map and the graphs were reasonable for what I did. GPS may have been about 50-60ft too high at top (gps: 937ft; benchmark 881ft, acceptable for my purposes), but the profile shape is accurate (I did some 15-sec stair reps and 100-ft hill reps on my way down).
However, I toggled the Resample elevation button, and it generated a profile that was really off (and credited me with an extra 300ft of up), so not sure what that button actually does. I could only get the original profile back, by going into the Graph page. Actually, the first toggle removed the profile, then the 2nd toggle brought back the strange profile. Maybe eric can provide a clue as to what the resample does.
The Garmin connect profile was pretty similar whether I used the corrected or uncorrected version and also agreed with the SportTracks.
I should point out that this is one of the reasons I use a gps: I can't count. I thought I only went up that 100-ft hill 4 times. I apparently did it 5. duh. (I'm usually more even in splits on those, but not arguing with it this early in season when running on snow.)
BTW, when you first posted, I had gone back and looked at some of my profiles and how RA compared with Garmin Connect corrected and not corrected. It was pretty random as to similarity.
Thank you both for the info. I'm not going to stress myself over it, I was just shocked at the massive difference.
I have a race in April that is reported as involving 3800' elevation, so I was just hoping to incorporate my training elevations into my planning.
I guess if I get really desperate, I can haul along my climbing GPS (handheld) since I can calibrate it and has shown to be accurate.
Use your handheld. That's what I did until the 910 came out.
Most of my races beyond 13 mi have over 3000ft of uphill, and mountain races typically have 2000ft up in 2.2 miles, so I definitely pay attention to the profiles. I've used my handheld to generate the elevation profiles for a couple races.
Have fun. It's nice to see another hill lover.