Ben,
Thanks for the Glossary Link. It's a saver.
Dave
Dennis,
Thank you. Your help is much appreciated.
Dave
I'm new to the board, but I thought I'd put my two cents in. I explain it to students and peers this way:
The NAD83 datum is really just a very specific orientation of an ellipsoid (the GRS80 ellipsoid), such that the ellipsoid fits North America 'conveniently'. That is to say, the ellipsoid is centered so that it fits pretty snugly against the surface of North America, but is pretty far removed from landmasses on the other side of the Earth (where the NAD83 datum isn't employed).
Once we've decided exactly how the ellipsoid should be centered and oriented (and we have, this orientation of the NAD83 datum is fixed and will never change), we could theoretically locate a point on the Earth with 0 degrees latitude, 0 degrees longitude, and 0 meters ellipsoid height. This theoretical point isn't of much value to us, however.
What is of value, however, is a network of points for which we can establish excellent coordinates in the datum.
The trouble is that not long after we establish the coordinates of a particular monument, the slow and steady movement of tectonic plates means the true, real-time coordinates of that monument will change. Instead of updating coordinates for these moving monuments every time they move (which would be madness), we take snapshots of coordinates for these networks of points at particular points in time.
These snapshots are the reference frames (alternately called 'realizations'). The NAD83 (2011) Epoch 2010.00 datum is a snapshot of the NAD83 datum coordinates of a network of points. The snapshot was taken on January 1, 2010 (thus the "Epoch 2010.00"), and the reference frame computations were completed in 2011 (thus the "2010").
The NAD83 (2011) Epoch 2010.00 coordinates you'll find for monuments are the coordinates stemming from that snapshot. Anytime you work with those coordinates, you're working inside that snapshot.
The next time NGS releases an updated reference frame (realization) of NAD83, it will be a more recent snapshot of the coordinates of (largely) the same network of points. The Epoch 2010.00 coordinates for those points will still be reasonable for you to use, as will the coordinates for the updated reference frame. You'll just need to decide which snapshot you want to work in and go from there!
I really hope this helps. Please let me know if any part of the explanation is lacking (or wrong!).
Thanks, Eddy. Nice explanation and it helps to cement my understanding.
One question I've got is about the Epoch. If the snapshot was taken on January 1, why isn't it called, say, "2010.01"? Why the double zeroes? Or why not use the JDay for the decimal part?
Oh, and welcome to the Forum.
Dave
> Thanks, Eddy. Nice explanation and it helps to cement my understanding.
>
> One question I've got is about the Epoch. If the snapshot was taken on January 1, why isn't it called, say, "2010.01"? Why the double zeroes? Or why not use the JDay for the decimal part?
>
> Oh, and welcome to the Forum.
>
> Dave
That would make too much sense! In any case, the snapshot is really taken at midnight at the start of the new year, so it really is .00 if you look at "decimal years".
Snapshots are also sometimes taken halfway through the year, and accordingly have dates such as 2010.50 or 2010.51, but I wouldn't be opposed to use of Julian dates.
In any case, I'm glad the explanation was of use to you. And thanks for the warm welcome.
Dear eddym and Dave,
Hi, eddym! Welcome!
eddym's post represents the first definition of "datum" in the
geodetic glossary. That is, the datum is only a set of parameters
used to realize a Conventional Reference Frame. Again, the dirty
secret of "datum" is that is has two different definitions in the
geodetic glossary. And, that first definition does not conform
to the IERS Conventions.
eddym, since you asked for comments/corrections, I'll pitch in.
NAD83 was not established at the Ideal level, nor at the Reference
System level, nor at the Reference Frame level as a North American
surface fit. NAD83, at all three layers of hierarchy, was configured
as an Earth-centered, Earth-fixed, ellipsoid (using the best available
Earth fit ellipsoid -- GRS80). Your explanation here is accurate
for NAD27. NAD27 did use a North America fit -- but not NAD83.
I can refer you to the NAD83 report at:
http://www.ngs.noaa.gov/PUBS_LIB/NADof1983.pdf
Briefly, NAD83 ellipsoidal geocentricity was materialized by the
use of Doppler coordinates. Now, the Doppler coordinates were in
a particular DoD reference frame. Using all information available
at the time, the offset of the DoD frame from geocentricity was
established, and applied as a correction to the result.
We now know, given the wealth of information available to us, that
the NAD83 reference frame missed geocentricity by over 1 meter.
But, it was not for lack of trying to hit geocentricity exactly.
In comparison, non-geocentricity of NAD27 exceeds 200 meters.
eddym, you relate the existence of Reference Frames to the
presence of crustal motion. I must disagree. The Reference
Frame is defined as the realization. That realization can change
over time merely by the accretion of geodetic data, which frequently
gets more accurate over the years. A succession of datums (definition #2)
existed before the incorporation of crustal motion in the Reference
System models. The spate of coordinate sets issued by NGS over
the past 30 years has been driven by the maturation of geodetic
GPS praxis, rather than shortcomings of crustal motion models.
Dave, eddym is accurate in characterizing what occurs when one
adds crustal motion/plate tectonics to the Reference System layer.
Now, one needs physical theory/models to address point motion.
Following the practice of Astronomic Geodesy, coordinates and
data are migrated through time to materialize the datum (definition #2).
And, it is most convenient to choose a single epoch (although, there
have been deviations from that in the past -- such as NSRS 2007).
Call it snapshot or epoch, it's genesis is the Reference System.
The coordinates are materialized in the Reference Frame, the
specific date is chosen, and that choice is one of the parameters
that makes it a Conventional Reference Frame.
eddym predicts the next realization of NAD83 will be largely the
same network as the last. Based on my experience and history
I would disagree. First, it is not clear that the 2022 Reference
Frame/Datum will be geocentric (as near as ITRF can make it), or
will be non-geocentric (to retain the original NAD83(86) Reference
Frame). These discussions are ongoing as of this date. Also,
the data and points submitted to NGS data bases continue to grow.
Note my 2007 report just to show geodetic network growth leading
to the 2007 solution (sections 4 and 5).
http://www.ngs.noaa.gov/PUBS_LIB/NSRS2007/NSRS2007Analysis.pdf
FWIW, I recommend the book " Geodesy for GIS and Geomatics Professionals," by Elithorp and Findorff. It can be ordered through: http://www.xanedu.com/educators/custom-books-catalog/elithorp/ I used it in an undergraduate course in geodesy where most students were ill-prepared for the higher maths normally expected in such courses. It also cost less than $100US.
I had wanted to use Gilbert Strang's Linear Algebra, Geodesy and GPS which I highly recommend though now out of print. Its successor, Algorithms for Global Positioning, not so much. BTW the co-author, Kai Borre, has provided supporting Matlab script files linked to the text. Scripts are free.
The Elithorp book does include a nice explanation of NAD27/NAD83 differences i.e. regional versus global datums. The references provided by Dr Milbert provide more current information about later NAD83 implementations.
Hope this contributes,
DMM
It does, indeed. Thanks, Odd.
Dave
Perhaps thinking about this issue as a leveling circuit might help. Running levels between four points A, B, C and D in the following manner: A->B, B->C, C->D I can determine their relative heights easily. I cannot determine their heights in any height system without adding additional information i.e. a height for one of the points.
Of course who would ever determine the height of a point without a valid check to another benchmark?
That's a pretty good metaphor, Odd.
Dave
Your explanation is excellent. I figured my understanding was lacking in places. Thank you for setting the record straight.
Wink! 😉
All the best!