# Elevation of a point on Earth

What purpose is served by taking sea level as the reference point for finding the height of any point on Earth when we all know that measuring the sea level is a very complex task and it differs from place to place and also keeps on changing? Why can't we just take Earth's centre as the reference point for finding the height of any location?

• how do you propose we measure from the center of the Earth? Sure would be a lot of significant digits to discern one elevation from another if they used the center of Earth as the reference.
– f.thorpe
May 7 '17 at 2:27
• I don't exactly know the method. But I think that whatever that method is, it must be a lot simpler then finding the mean sea level. You can also check the link astronomyforbeginners.com/astronomy/howknow.php May 7 '17 at 2:43
• i can think of a bunch of different ways to find the mean sea level... finding the distance to the center of the Earth is much more complicated
– f.thorpe
May 7 '17 at 3:42
• One reason is that the Earth is not a sphere. Have a look at this very related question: Farthest point from the center of the Earth May 7 '17 at 5:07
• Which center? the center of mass? or the center of volume? and remember both of which move.
– John
May 8 '17 at 1:51

There are many reasons.

First, measuring from sea level is traditional. It started before we had an accurate way of measuring distance from the center of the Earth, and probably before most people knew it was (approximately) a sphere. But anyone living by the shore could determine sea level to within a few feet.

Another reason is that measuring from sea level tells you useful things like pressure altitude, which are important if you're a pilot or mountain climber. The polar and equatorial radii differ by about 14 miles (21 km), so depending on your location, a distance of say 6,365 km from the center could be deep underground or halfway to the stratosphere.

• You have a typo in your 2nd last line, 6,3565 km. Is there a decimal point missing?
– Fred
May 7 '17 at 8:49
• @Fred: Thanks for catching that. It's not a missing decimal point, just was caught by the @\$#%! default insert mode (rather than normal overtype) of the browser. May 7 '17 at 18:12

Why can't we just take Earth's centre as the reference point for finding the height of any location?

While there are many uses for Earth-centered, Earth-fixed coordinates, they're not particularly useful or meaningful in most human-centric endeavors, which happen on or near the surface of the Earth. Consider two points on the surface of the Earth, one at a distance of 6377.12 km from the center of the Earth, the other at 6359.59 km. At which location will you have to worry about altitude sickness? The answer is the one that is 17.53 km closer to the center of the Earth. The first location is Lake Asal in Africa, which is 155 meters below mean sea level. The second location is the Amundsen-Scott South Pole Station, which is 2835 meters above mean sea level.

Because those ECEF coordinates are not particularly useful or meaningful, the first thing a GPS receiver does after computing position in ECEF coordinates is to convert that position to latitude, longitude, and elevation above a reference ellipsoid. There's an issue with this, which is the use of a reference ellipsoid. Some places at sea level have an elevation of 85 meters above the reference ellipsoid, others as much as 107 meters below the reference ellipsoid.

While there certainly are challenges in the concept of mean sea level, it remains extremely useful. It tells us, for example, which parts of New Orleans are subject to coastal flooding, which parts of Fargo are subject to springtime flooding, and which parts of Colorado are subject to altitude sickness.

• Why does the mean sea level tell us which parts of Fargo are subject to springtime flooding? Wouldn't that be the elevation difference between riverbed and floodplain, such that whatever reference point one chooses cancels out anyway?
– gerrit
May 8 '17 at 11:18
• @gerrit -- Some reference is needed to tell you which way water will flow. Distance from the center of the Earth will not do the trick, nor will elevation above the reference ellipsoid. A geoid (any geoid; there are an uncountable many of them) will do the trick. However, the standard geoid is essentially a modernized version of mean sea level. May 10 '17 at 7:57
• Why would you need a reference point to measure the elevation difference between riverbed and floodplain (which is what you need for Fargo flooding modelling)? You don't classically measure the height of a building with a barometer either. Certainly with small, local differences, it should be more accurate to measure the elevation difference between the two, than to measure the elevation of each compared to a distant reference point? More of a surveying question than an Earth science question though.
– gerrit
May 10 '17 at 10:58
• @gerrit -- Surveying is at the heart (or better, surface) of Earth science. As an Earth science, it's arguably older than geology or geography. (Except for the surveying aspects, it's hard to call pre-18th century geology a "science".) The two Mason-Dixon line was surveyed by two English astronomers, one of whom later motived another surveying experiment, the Schiehallion experiment. See this answer to How is the mass of the Earth determined? May 10 '17 at 11:44
• I did not mean to imply that surveying is off-topic, perhaps it deserves a meta-question to consider the boundary between Earth Science, surveying, and GIS.
– gerrit
May 10 '17 at 11:57