AscenDB Metastatistics

The recent release of AscenDB provides an interface to a database of over seven million peaks. To be exact, there are currently 7,798,709 active peaks available for users to query and examine. Even better, each peak is guaranteed to have an elevation, prominence, and geocentric distance estimate tied to its location. With all of this data at our fingertips, we can ask some interesting questions about the distribution of these stats across the world's summits.

Elevation

AscenDB stores its elevation data in feet for finer-grained accuracy, which is useful when ranking peaks as there are fewer peaks with the same height. However, much of the world uses meters when measuring the height of mountains, and has divided the world's summits into nine classes: below 1,000m, between 1,000m and 2,000m, between 2,000m and 3,000m, and so on until only peaks above 8,000m are left. There are no peaks with elevation greater than 9,000m above sea level; even Everest is off by about 500 ft.

How many peaks are in each of the classes described above? The chart below should give us an intuitive idea:

Peak Distribution by Elevation - All

As we expect, the number of peaks increases in the lower elevation brackets, and quite dramatically at that. In fact, AscenDB currently recognizes only 19 peaks with elevation higher than 8000m. To get a better comparison of the distribution without the 1000m peaks skewing the results, lets look at peaks only higher than 3000m:

Peak Distribution by Elevation - 3000+ m

Here we have an unexpected find: there are several thousand more 4000m peaks than there are 3000m peaks.

Prominence

The distribution classes for prominence are not as evenly defined as those for elevation. In this article, some commonly defined classes of prominence are used (ultra-prominent, P2K) but others have been invented to get a better distribution. The full set of class divisions used to create our analysis is thus:

Prominence >= 4921 ft. (1500m - Ultraprominent)
4921 ft . -- 2000 ft. (P2K peaks)
2000 ft. -- 1000 ft.
1000 ft. -- 600 ft.
600 ft. -- 300 ft.
Prominence < 300 ft.

In most lists of mountains, individual peaks are recognized only when they have 300 or more feet of prominence. AscenDB currently ranks peaks with 100 or more feet of prominence, and as we shall see, this again skews the results quite heavily:

Peak Distribution by Prominence - All

There are over six million peaks with less than 300 ft. of prominence in the database. This makes the rest of the graph difficult to visualize, so lets check out only the peaks with 600 or more feet of prominence:

Peak Distribution by Prominence - 600+ ft.

Even over a smaller subset, the prominence distribution of peaks seems to follow a nice exponential curve.

Geocentric Distance

AscenDB is unique among peak databases in storing geocentric distance estimates for every peak in the database. A forthcoming post will include more interesting details about geocentric distance and its relation to peaks across the world, but for now we will look at a distribution much like the one for elevation. However, the classes for geocentric distance are harder to differentiate visually, because the distance between highest and lowest is much smaller than the actual values themselves, by several million. Elevation does not suffer this problem.

Another interesting point of note: the 'lowest' geocentric distance estimate for a peak in AscenDB is 20,856,553 feet from the center of the Earth; the 'highest' geocentric distance is 20,946,210 feet, almost 90,000 feet higher than the lowest! If we took the lowest peak's geocentric distance as 'sea-level', then the highest peak on Earth would be Chimborazo, and it would be three times higher than Everest!

The classes for geocentric distance are evenly distributed, much like the classes for elevation, except they are split into intervals of 15,000 ft. instead of 1000m:

Peak Distribution by Geocentric Distance - All

This distribution is quite different from all the others, and in fact seems nearly reversed! The majority of the peaks are at the high end, which means peakbaggers have to work harder in order to find mountains that are closer to the center of the Earth. So should peaks with less geocentric distance be prioritized because they are rarer? I'm not convinced: even though the majority of peaks have a relatively high geocentric distance compared to the overall range of distances, the highest class still has over 200,000 peaks to choose from. That's few enough to be considered 'rare' relative to the total number of peaks in AscenDB, but more than enough for any lifetime.

Stay tuned for upcoming releases with more features, and additional informational posts like this one. Until then, see you at the summit!