I love exploring the outdoors, hiking up mountains, photographing white snow at sunset or the dark misty floor of old New Zealand podocarp forest, sitting by the side of wide boulder-strewn rivers and setting up my tripod in the middle of a creek to take long exposures surrounded by deep green mosses.
I also love data. I hated maths at school and spreadsheets bore me to tears, but the wonderful things you can do with data not just to make it pretty but also to see the raw numbers in a new way … much like the difference between describing a new animal to someone, and showing them a picture of it.
When I go hiking I collect a lot of data. I capture the full track log with latitude and longitude, elevation and timestamp; I capture weather information including temperature, humidity and take wind speed readings at intervals. I also take azimuth readings and get position fixes on points of interest to later triangulate them on a topo map so I can accurately caption my photos and describe my experience.
I’ve also started recording audio logs to capture additional information about the experience to help recall and convey the story later (with a secondary purpose of being useful to a coroner investigating my death should I fall off a cliff, again)
This is a Google Earth plot of yesterday’s tramp up to Lewis Tops in Lewis Pass, New Zealand:
This is the elevation plot, DEM-adjusted from Google Earth (above) and the straight GPS elevation data as shown in ExpertGPS (below):
Unfortunately the track data from the ascent up the hill to the bushline was corrupted and lost by my Garmin eTrex 30 (which we’re discussing on the Groundspeak Forums), so I managed to duplicate it for the Google Earth plot but fixing it for an elevation plot is a bit trickier.
The trip computer is separate from the track data recording, so this covers the whole trip:
I’ve found the elevation stats from my Garmin’s trip computer are usually way over (although apparently this has been corrected in the last 2.9 firmware) and that from Google Earth’s DEM data way under so those two extremes plus the track data usually give me a moderated result. In this case though the trip computer is way out, as according to my calculations the elevation gain was closer to 1,200 metres and at least 909 metres according to the Google Earth DEM data which is usually conservative. At the very least the the interval between valley floor where I started and the summit was 690 metres so … I think my Garmin is having some issues at the moment.
For those unfamiliar with elevation gain, it’s just the accumulated vertical distance on a track; so in this case if the trail was a smooth and straight concrete ramp from bottom to summit the elevation gain would be exactly 690 metres. But it’s not smooth and straight; it’s up and down and even after reaching the summit and going along the ridge then looping back and coming down through the forest accrued another 187 metres elevation gain according to the track data (barometer + GPS).
Anyway, some more graphs; this is the west-most part of the track as I looped around a tarn and headed back off-trail:
I didn’t get a photo of that particular tarn as there were some people swimming in it, but I took one of a similar scene though a different tarn, just to the east:
… which I also got a photo of from below when I looped back downhill:
The timestamps of these photos are compared to the timestamps of the track data by ExpertGPS and geotagged to about 5 metre accuracy so I can review on a map exactly where the photos were taken:
Now for the weather data; this is the temperature and relative humidity data sampled at 10-minute intervals, straight off the Kestrel and generated by Kestrel Communicator (smoothed):
Interesting, but more useful when plotted with the track and terrain data; here I’ve shaded green the timeframes I was below the bushline in the sub-alpine forest and the white section is where I was above the bushline in the alpine meadow. The red line marks the highest point of my climb at an elevation over 1,560 metres.
Now it starts to give some context and meaning to the data.
Another fun thing to do is annotated illustrations of scenes, often done on signs at viewpoints though with much fancier graphics, but some rough lines to reproduce the terrain with sufficient fidelity to correlate it to the photo is fine.
This is a photo from Lewis Pass looking north-east up Cannibal Gorge along the St James Walkway to Ada Pass and Gloriana Peak where there’s still some snow down to 1,800 metres:
More to come in later posts, with overlaying photos with topographic data, showing fields of view of photos on a map taking into account obstacles like mountains, taking high-accuracy azimuth readings, position fixing and triangulation from multiple locations with a short baseline with and without a GPS; and more.
In closing, a photo of a white gentian in the tussock herbfields of Lewis Tops:
I left my DSLR at home on this hike and just took the point-and-shoot to lighten my pack for the 4,000+ foot elevation gain.
If you enjoyed this article please share it with your friends and colleagues.