Vertical Reference Frames

Having a reliable (and accurate) Vertical Reference Frame with concomitant geoid model will allow GPS Leveling operations anywhere in the country without reliance on monumented elevation benchmarks. The cost of providing reliable elevations will essentially be evened out anywhere to the benefit of the general public. Implementing that is another story ...

As both Cliff and Gavin have pointed out, maintaining a classical network of passive vertical control points (bench marks) is a losing proposition - no country has ever been able to figure out to keep them from being disturbed or destroyed. In addition, without very expensive and time-consuming repeat geodetic leveling campaigns it is impossible to maintain the vertical integrity especially in areas of active subsidence or uplift. In addition, geodetic leveling as good as it can be still is subject to error build up the further you go. The only way to resolve these issues is through a significant effort of combined space-based, airborne and terrestrial gravity data collection. The U.S. National Geodetic Survey is currently engaged in such an effort to produce a purely gravimetric vertical reference frame by 2022 - Gravity for the Redefinition of the American Vertical Datum )GRAV-D.

The (very) long term goal of this effort is for a common vertical reference frame for the entire northern hemisphere. Because of the lack of a high quality geoid model in most areas the vast majority of countries in the world basically lose the potential of the height component available from GNSS.

Neither option is cheap. Developing high accuracy gravimetric vertical systems will be expensive (although less than leveling) and require consider partnering and cooperation between countries.

Best of luck with your study.

Thanks for the comments.

Have you any thoughts on how rising sea levels are likely to impact a national datum (in Ireland's case MSL at Malin Head) or if there is likely to be significant geoid displacement in a non-glacial island like Ireland.

Really appreciate the comments so far.
Ireland currently has a continuous geoid model, developed in conjunction with the UK.
There is also a VORF model for the UK & Irish waters.
I'm interested in thoughts on how this infrastructure could be improved further, such as linking internationally, fully integrating terrestrial and maritime VRFs or simply improving resolution of existing infrastructure.

In and of itself, sea level change (up or down) should not have a significant impact on a nation's vertical datum. Every vertical datum I've ever seen is related to local mean sea level at one or more points, derived from some period of observations and in many cases the leveling defining the geodetic datum is carried for thousands of km away from the origin. Since tidal datums are very local the change in sea level over time will vary from place-to-place - sea level trends in the U.S. -- http://tidesandcurrents.noaa.gov/sltrends/sltrends.html

What is very important is to have a monitoring system that will accurately relate changes in the local tidal datums to the geodetic reference frame, either by leveling or GNSS. The National Geodetic Survey (NGS) in collaboration with the National Ocean Service, Center for Operational Oceanographic Products and Services (CO-OPS) has developed a transformation tool called VDATUM -- VDATUMthat allows surveyors and engineers to accurately compute the relationship between various heights systems both geodetic (orthometric and ellipsoid) to local tidal datums for almost everywhere in the country. This is possible because NGS and CO-OPS have collaborated for years to make sure there are accurate connections between the geodetic datum - the North American Vertical Datum of 1988 (NAVD 88) and virtually all of the 200+ long term tide gauge sites that are part of the National Water Level Observation Network.

No vertical datum really uses sea level, since that is hard to measure at best and ill-defined at worst due to currents, tide cycles, salinity/density variations, etc. A reference may be chosen to be around some estimate of sea level, but the real definition is in terms of one or more solid bench marks.

Melting of ice caps could have a measurable impact on the geoid.

The definition or realization of the new vertical datum to replace NAVD 88 (and the other island datums) in 2022 will not include any bench marks - it will defined by a geoid model plus your GNSS receiver. One of the major problems with NAVD 88 today is that the vast majority of bench marks have not been releveled in over 40 years.

Interesting that you would be tasked with exploring "non-traditional" reference sources (what is that anyway?) when there is a wealth of tradition resources available discussing this issue. See

http://192.106.234.28/Meetings/GGOS%20planning%20meeting2010/session%203/3_Sideris_GlobalUnifiedHtSysFin.pdf

The IAG working groups have lots of material on the issues.

Once again, what is a "non-traditional reference source" and how does it differ from merely statements of opinion like found on internet forums?

Perhaps by non-traditional your instructor means outside the surveying/geodesy community. This would include hydrologists and the others who face height issues when attempting to understand storm surge, evacuation routes and the like.

Argumentatively,

DMM

Problems with using tide gauges for sea level analysis

While there may be some entities, like the TCOON run by TAMUCC--CBI, that do the job well, i agree with Professor Hannah's views in this FIG paper: http://www.fig.net/pub/monthly_articles/july_2010/july_2010_hannah.html

Enjoy,

DMM

A technical point does need to be raised though:

A vertical datum can not rely on "the geoid" as its surface of zero elevation unless you define what "the geoid" is. Unfortunately, in this era, there are now many ways to do so. For instance two reasonably valid, yet mutually conflicting definitions are:

1) I could say that it is that surface of constant gravity potential energy equal to 62,636,856.00 m**2/s**2, no matter what, and no matter when. That means that as masses move around (including the water masses between ice sheets and sea level; including the lower mantle build up under Hudson bay; including the accretion of space dust onto our planet; etc etc) and both the size and shape of every equipotential surface changes, I will not stop using that surface whose numerical value is 62,636,856.00 m**2/s**2.

2) I could say that it is that surface of constant gravity potential energy which best fits, in a least squares sense, to global mean sea level. As (I hope) we all know, sea level is changing globally (rising) on a magnitude of millimeters per year (global average) and if it is "swelling" annually, then so should the geoid.

[For those who want to delve deeper, there is both the direct (mass change) and indirect (geometric location of sea level) impact of ice melting/sea level change on all equipotential surfaces, but that's for another discussion...issues of changing SHAPE without changing SIZE, etc etc]

So, take these two conflicting definitions at face value, and let's look at a really simply example:

Under definition #1, if you have a house near the ocean whose floor has an orthometric height of 1.000 meters (e.g. it is 1 meter above "the geoid") in the year 2022, and sea level is rising globally at 1.7 mm/year, then 50 years later (2072) the orthometric height of that floor is still 1.000 meters (assuming no subsidence of the crust nor any change to the masses which would change the geoid's physical location near this house, etc).

Under definition #2, that 1.0 meter orthometric height becomes 0.915 meters after 50 years (1.000 - 0.0017*50). That is, without subsidence, your orthometric height has dropped, telling you that the geoid is "creeping up beneath you", because sea level is swelling globally.

The question is: What do users want? Do you want a height that has some tie to physical change of the ocean? Or do you want a height that is disconnected from global sea level change? Both are computable. Both have some usefulness.

Of course there's more to it...local sea level change is more useful than global, etc etc. But fundamentally, the definition of "the geoid" must be adopted for any orthometric height system to work, and the time has come to recognize that the definitional tie between geoid and sea level can no longer be presumed in a world whose non-static sea level is known and measurable at a scale that matters.

NGS serves the people of the USA and we've heard arguments for both definitions of the geoid to be adopted for 2022. That's why you haven't heard the "final word" on this from us yet. We're being cautious about getting this right.

Dru Smith, Ph.D.
NGS's NSRS Modernization Manager

Is there any anticipation that 2022 will include time dependent vertical translations? Are vertical displacements consistent enough to be modeled over time?

It seems to me that a system were the elevation of solid objects (like the Hoover Dam, for example) was seen to be constantly changing, year by year, would be a difficult one for the average surveyor to manage and the general public to comprehend. There would be an impression that the dam (and all similar engineering works) was unstable - and the internet would make much of that.

Not just that, but orthometric heights, at least for me, are for differentials. I don't really care how high above sea level (an arbitrary value) a point is. (I'm also hundreds of miles from the coast). I'm more concerned about the difference in elevation (related to gravity) between two points. I'm interested in the orthometric height difference, because I'm interested in how water will behave between these two points. It's not really about determining a unique location on planet Earth, such as an ellipsoid height is.

I'm interested in how water will behave between these two points

Then technically you need Dynamic Heights, although unless you are working on a very large project or have a large vertical range, the differences will be about the same.

If everyone in the future gets elevations from GPS, gravity models, and short level runs, why not define the new system with Dynamic Height?

Here in the Portland Area there are already 4 different vertical datums in play:

1. The City of Portland Datum, established in or before 1895 and based on tide gauges at the mouth of the Columbia River. Reasonable well monumented with brass.

2. NGVD29(47) datum, a 1947 regional adjustment of '29. Adopted (prior to NAVD88) by the Oregon DOT and many counties and cities in Oregon and Washington. Lots of legacy data on this datum, lots of agencies require deliverables using it.

3. The City and Port of Vancouver, Washington Datum, based on the pre-1947 adjustment of NGVD29. Differs from the 1947 adjustment by about 0.2'. The supply of brass monuments on this datum is diminishing but the C.O.V. still requires us to use it.

4. NAVD88. The supply of brass on this datum is pretty limited. Of course, offfsets to the other datums (datii?) is known so estimates can be made.

None of these will be going away in my lifetime. [sarcasm]I can't wait to add a 5th datum to this mix[/sarcasm].

drusmith, post: 344698, member: 9799 wrote: ... that surface of constant gravity potential energy equal to 62,636,856.00 m**2/s**2, no matter what, and no matter when.

I'm curious how this works operationally. How can this defining constant be connected with actual observations? One way might be to drop an object from infinity and measure its energy as it hits some reference surface on the ground. The precision would be abysmally low. There must be some procedure, many orders of magnitude better, of comparing this defining geopotential with GNSS ranging, or satellite orbital periods, or some such.