Global Slice 🌍: Land



I’ve been trying to find a reliable source for the figure of 25% that Chris quoted from Chartbin, of how much of the Earth’s surface is mountain. Unfortunately, Chartbin has been deleted, and most of what’s on normal Google regarding this is just unattested values which one spammy site copies from another spammy site, without attribution or justification. An important point is that it doesn’t make sense to have a pie chart like that, with a slice for mountain and a slice for desert. That’s because the categories often overlap, e.g. the Great Basin Desert, , contains mountains.

However, looking for people who are actually working to refine such measurements, I found a useful paper from the US Geological Survey, Nature Conservancy and others (2020). This is “An assessment of the representation of ecosystems in global protected areas using new maps of World Climate Regions and World Ecosystems”, (PDF) An assessment of the representation of ecosystems in global protected areas using new maps of World Climate Regions and World Ecosystems ( .

This is based on much previous work (Roger Sayre has a lot of self-citations). The authors set up four independent ways of classifying land regions:(1) Temperature: tropical, subtropical, warm temperate, cool temperate, boreal, polar.(2) Moisture: desert, dry, moist.(3) Landform: hills, mountains, plains, tablelands.(4) Vegetation and land cover: forest, shrubland, grassland, sparsely or non-vegetated, cropland, surface water, snow and ice, settlements. In the paper, we see these in Maps 1, 2, 4, and 6.

The authors combine these classifications to make a high-resolution database of ecological regions, shown in Figure 7 and Table 4.

Note that these classifications are consistent with work on conservation, e.g. page 11 says they’re compatible with the FAO and IPCC classifications. Note also that, as implied at the start, mountain and desert are terms in different taxonomies. The first is a kind of landform; the second, a degree of moisture.

Interesting note: determining mountains is hard, and boils down to automatically working out how wrinkled the Earth is. Similar algorithms have been applied to measuring the wrinkledness of fabrics. They’re also being used on moons and other planets, e.g. Europa.

Other papers and links are below, but the Sayre et. al. one is recent and looks one of the most useful.

The UN page “United Nations Decade for Deserts and the Fight against Desertification”, has useful basics on the definition of dryland and desert, their subcategories, and the amounts of each. Unfortunately, it expresses these as percentage of land area, but doesn’t say what its definition of that is. For example, does it include icesheets?

There is a section at the end on land degradation, which makes me think we should do a page on that.

There is a 2021 UN page on rangelands, “New atlas reveals rangelands cover half the world’s land surface, yet often ignored despite threats”, . We should mention these. They’re not explicitly listed in the land-cover types I used from the Copernicus project: its categories probably overlap them.

Yet another categorisation, which cuts across that, is by landform. That is, by morphology, see , though that one is not a good article. There’s an interesting paper “Classification and Characterization of Planetary Landforms”, , which explains why their classification is difficult. It also shows how badly the taxonomy differs between researchers, making results hard to compare. Kenneth Gregory’s “Recognizing the Land Surface”, , has some useful info.

Yet another classification is by biome, .

Total Land Area

Total land area in 2017 was 147,135,404 square kilometres. This includes inland lakes and other permanent bodies of water. Estimates vary slightly: this one is from the source noted below. [ Note: the land-area percentages in those results, which I use in the next section, sum to only about 90% of this total. The source, the Copernicus Project, is reliable, so I need to check this with them and ask why. ]

For each person in the 2017 population of 7.509 billion, this is 19,594.5 square metres, or 0.0196 square kilometres. For a population of 10 billion, it would be 14,713.5 square metres, or 0.0147 square kilometres. For 12.5 billion, it reduces to 11,770.8 square metres or 0.0117 square kilometres.

To visualise this, here is what each of ten billion people would get, if set around the Brandenburg Gate in Berlin:

It is 121 metres on a side, just under twice the width of the Gate, 64 metres:

The slice for 7.5 billion would be 139 metres on a side; that for 12.5 billion would be 108 metres.

Slicing out Cones, Walking Your Area

We can think of these slices as cones. Each person gets a cone with its point at the Earth’s core, and a big end which — depending on the population — is not very much bigger on each side than a famous monument in Berlin.

Assume that the average person walks 1.4 metres per second. Then it would take 6.7 minutes to walk round the 2017 slice of land; 5.8 minutes to walk round the 10-billion population’s slice; and only 5.2 minutes to walk round the 12.5-billion’s slice.

Land Cover by Area

TypePercentage Area
Herbaceous vegetation21.16%
Bare or sparse vegetation 15.33%
Permanent water bodies2.05%
Herbaceous wetland1.54%
Snow & Ice1.44%
Moss & Lichen1.06%
Calculated from the 2017 land-cover statistics given by the Copernicus Land Cover Viewer at . Percentages have been multiplied by 1.1 so they add up to 100: see note in previous section.
Charted from the above data.

Countries by Area

countries by share of earth's surface
From .