Of the words, terms and labels used in the geospatial professions, geodesy may be the least understood. Sometimes, you just have to start with a definition.

Geodesy is the science of measuring positions, distances, directions, and the gravity field for the purpose of creating geometric models of the earth’s physical surface that are suitable for the positioning and geo-referencing of geographic detail thereon. [By the author] 

The word geodesy comes from two Greek words: Earth and dividing. This relates to the fact that some of the geometric models (like projections) divide the earth into “zones.”  

Words like datums, projections and geoids are commonplace, but all too often we fail to achieve a real understanding of the concepts behind them. They are part of geodesy. Unfortunately, even experts disagree on the meaning of geodesy, since there are many parts to it. To explain:  

  • There is the core science of geodesy, which includes the above concepts, and more.  
  • When practical geodesy is taught as part of a comprehensive academic program, it covers more tools and methods, such as transformations, least squares adjustments and photogrammetry.

This is an important distinction, because these practical parts are the ones that give users the most grief. For example, do you know that a least squares adjustment is done every time one performs a GNSS observation or a transformation? Having tools with push-button functionality is great, but how are we to apply professional understanding of the results if we do not know what happens behind the scenes? 


Towards Geo-References

Geodesy is the science that gives us geo-references. It is used to: 

(a) develop the mathematical models for the size and shape of Earth (ellipsoids), 

(b) build geometric constructs on top of those ellipsoids (projections), and 

(c) study gravity fields that are used for elevations.  

With all the current use of digital maps, it can be easy to assume that geodesy is also a new thing. We know that the Earth was not created with a coordinate system. “On the eighth day, let there be a coordinate system” did not happen.  So, humanity had to find a way to create a geo-reference.  

Historic documents found at the Library of Alexandria provide evidence that Eratosthenes of Alexandria (300 BC) was the first one to have described a projection placed on a sphere. That makes geodesy at least 2,300 years old.    

Since then, if man wanted to travel to China or to the Moon, or wanted to launch a worldwide global satellite navigation  system, geodesy was at the center of it. Ellipsoids, geoids, and projections are all constructs that help create geo-references, and a wealth of information was written about them. The resulting continuous emergence of new technologies and methods became an independent and comprehensive science that was called “geodesy” (from the Greek “dividing the Earth”).  


What is Good Geodesy?

Today, everyone uses geodesy. It is the basis for any mapping, GNSS, GPS and navigation technology. Without it, solutions such as Google Earth, XYO, self-driving vehicles and ordering pizza would be much less effective.  

Accompanying all this is the concept of a datum. In Latin, “datum est” means “it is given.” It comes from “dare,” or  to give. So, every time we fix a geo-reference in space, we need two things to happen. First, someone needs to put a finger on the ground and say (with authority), “here.” Second, the rest of the people have to say, “Ah, OK.” That means that every datum is based on a (hopefully) informed decision, followed by consensus. It would also have to be based on good geodesy.  

What is good geodesy? Who sets the standards? Worldwide, geodesy came to the foreground when nations started to launch large military operations or large national mapping programs. In the United States, the USGS started mapping the national territory in 1884 (that is over 130 years of mapping). Every time a new and better technology to measure angles, distances, and elevations appears, geodesy is strengthened and amplified.  

For example, as of late, the National Geodetic Survey is no longer assuming that the world does not move. Every portion of the Earth’s surface suffers tectonic shifts, even if we don’t feel them. This means that everything moves, even the all-important satellite tracking stations. We have already reached GNSS accuracy levels that allow centimeter-range results with appropriate tools and methods, and so the published coordinates for benchmark points will suffer relatively significant and continuous changes over time. To be able to handle these changes, the National Geodetic Survey had to introduce “velocities.” Point data are now published with (1) coordinates with a time stamp, and (2) a “velocity,” which is a vector expressing the rate of change in centimeters/year and the direction of that change. Therefore, datums are no longer seen as rock-solid, because rocks are less solid than thought. 

What makes this exercise more difficult is that the worldwide GNSS/GPS system is observed from tracking stations, and each one moves at different rates and directions. In the absence of enough understanding of geodetic principles and history, these key operations become impossible.  


Early Systems

Another good example is the Public Land Survey System, which was developed without the benefit of projections.  It is based on early determinations of meridians and parallels, and measurements on the surface of Earth. Therefore, there is a disconnect between parcel data expressed within townships and sections, and data that are placed geodetically (based on a projection or State Plane Coordinate System). 

It becomes very difficult indeed to bring survey data via CAD into a GIS database. The reasons for this are two big ones: (1) Township and Section boundaries do not match grid meridians (plat bearings do not line up with grid North), and (2) survey (ground) dimensions are not easily convertible to sea-level ones (creating the need for scaling). This difficulty has forced communities to issue instructions/regulations for this operation. How many people know that? This is definitely part of geodetic training.  


Understanding the Problem

To properly understand the problem with geodesy, it is enough to realize that there are no longer any comprehensive academic programs to support geodesy, and “geodesist” is no longer a recognized job description. That does not mean geodesy doesn’t exist. Consider the large amount of information available on the NGS and NOAA websites. Some people have made significant contributions to the science, such as (notably) Peter Dana. I have benefited significantly from his work.  

Expanding on the idea of geodesy, there are tools and methods that the GIS practitioner encounters, but which often are left out of discussions on geodesy. For example, transformations help bring one dataset into the georeference of another. This normally can be done in two ways, either with a Helmert or Affine transformation. The difference between the two is normally not taught, even if the use of the Affine version can actually deform a database by definition. Only a look at the math will reveal the dangers. As such, things like transformations, least squares adjustment, statistics, etc., need to be part of geodesy/GIS training.  

Where can students and geospatial professionals get a good foundation in geodesy? Most experts would agree that it takes a good part of your professional life to become proficient, and, therefore, it is not a trivial effort. Unfortunately, most universities have stopped offering comprehensive geodesy programs. It seems that some programs exist in foreign countries, but they are hard to reach. A possible approach is threefold:  

  1. Start with a basic introduction to geodesy, one that provides a good cross-section of the major components of practical tools and methods.  
  2. Then take more detailed short courses in specific geodetic topics, such as datums and geoids.  
  3. Then, peruse the vast geodetic library provided by the United States Geodetic Survey by, for example, accessing https://pubs.er.usgs.gov/publication/b1532, which is the most important geodetic publication in the U.S.

Recognizing the needs of those entering or working in the GIS workforce, GeoCounsel also developed a 15 session course. This material is not causal, but based on graduate university studies and many years’ worth of consulting. Learn more at https://www.geocounsel.com/services/training/

Then, there is a book that is seen as the bible of geodesy. "The Third Edition of Geodesy," by G. Bomford, was published in Oxford at the Clarendon Press. It was first published in 1952, and the last edition was in 1971. This 731-page book has a wealth of information that will be important for centuries. It is still available from Amazon. A multiple-semester course in geodesy can be based on this material.