Natural Arch Identification
For accurate reference, as well as indexing purposes, every cataloged
natural arch must be uniquely identified. Normally this is done by assigning
each arch its own separate reference label. A standard
for natural arch reference labels is proposed below, along with
a set of guidelines for
assigning them. However, before the standard is defined, the need for
it is briefly discussed.
Reference Labels and Catalog Numbering
All serious natural arch catalogs published to date include reference
labels for the arches they contain. Reference labels are frequently
called catalog numbers. They should not be confused with the name(s)
given to an arch. All of the various authors of natural arch catalogs
have abandoned the use of names to uniquely identify arches. The reason
for this can be found at the link "How
do natural arches get named?" Of course, any names that exist
for an arch are included in the information provided by these catalogs.
But such names never supplant the catalog number (reference label) as
the primary way of indexing the arches in the catalog.
Most authors have chosen to provide catalog numbers (reference labels)
consisting of two parts – a designation for a geographical area and
a separate number for each arch cataloged within that area. The format
for these labels is usually some variant of A-N, where A designates
the geographical area (it may be a number or group of letters) and N
uniquely numbers each arch listed by the catalog in area A.
The geographical areas found in current natural arch catalogs are usually
defined by political boundaries, e.g., the borders surrounding a country,
a state, a county, a park, or some group of these. Other geographical
criteria are occasionally used to define an area, e.g., a mountain range,
plateau, or continent. In some catalogs, the second part of the reference
label, N, orders the arches within area A in some way, e.g., by longitude,
latitude, or significance (as defined by the catalog author). In other
catalogs, N is assigned in an arbitrary sequence, e.g., the order correspondents
reported the arches to the author. Not only does the method for sequencing
N vary from author to author, it even varies within the work of the
same author. In one major catalog, arches are ordered by significance
in some areas, longitude or latitude in other areas, and arbitrarily
in still others. This is a reflection of the fact that most catalogs
have been published in parts over a period of several years.
Problems with Current Approaches
Although the A-N structure for reference labels might be considered
a de facto standard, there are reasons why a new standard is needed.
These are discussed below.
Catalog overlap. Natural arch catalogs by different authors
have overlapping coverage. As a result, many natural arches have been
assigned multiple reference labels, one from each catalog in which they
are included. For example, Landscape Arch in Arches National Park is
labeled as 2-2 in Vreeland's catalog (reference
1), SD-56 in Stevens/McCarrick (reference
2), and 1.1612 in Tom Van Bebber's World Arch Database. Although
cross-references have been developed, it is clear that the A-N structure
has resulted in different authors defining different areas and then
numbering arches differently within those areas. This in turn creates
potential confusion. A standard method for uniquely identifying natural
arches that all catalogs followed would help avoid that confusion. The
proposed standard is easily adopted by anyone wishing to develop or
revise a natural arch catalog, or just contribute a single entry in
a catalog.
Catalog maintenance. When things change, the A-N structure gets
broken. Catalog maintenance becomes problematic. For example, consider
the impact of a change in political boundaries that define an area within
a catalog. This happens with surprising frequency. The boundaries of
Arches National Park have changed several times since it was first created
in the 1920's. Counties change regularly, too. Not only have there been
several boundary adjustments, at least three new counties have been
created in western US states in the past 20 years. Finally, compare
the current borders of Balkan countries with those of 1990. Each time
something like this happens, some arches that were in area A aren't
anymore, and some that weren't, now are. Should these arches have their
reference labels changed? If so, it's usually clear what the new A should
be, but what number N should be assigned? Inserting a new arch into
an ordered sequence could be difficult. Also, taking an arch out of
a sequence will leave a gap. Should the remaining arches be renumbered
to remove these gaps? Under the A-N structure, leaving the reference
number unchanged might mean the arch is cataloged in the wrong area,
the ordering of arches in an area might be incorrect, confusing gaps
in catalog number sequences might be introduced, or all of the above.
For this reason, it would be better to have a reference label that either
never has to change, or when changes must be made, no maintenance issues
result. This is a major advantage of the proposed standard reference
label.
A Proposed Standard Reference Label for Natural
Arches
The proposed standard for labeling natural arches is based on their
locations within the Universal Transverse Mercator (UTM) system and
the World Geodetic System of 1984 (WGS84) datum. The complete reference
label for any natural arch in the world is its full UTM coordinates
together with the prefix 'NABSQNO'. This prefix is an arbitrary character
string used to distinguish the reference label from a location. Thus,
reference labels have a format of NABSQNO Z-E-N, where Z is the UTM
zone, E is the UTM easting, and N is the UTM northing. For example,
the standard reference label for Landscape Arch is NABSQNO 12S-620979-4294446.
Comprehensive information about the UTM system and the WGS84 datum
can be found at several websites such as Using the UTM Grid System and Using
the UTM/MGRS map coordinate system. The reader is encouraged to
become familiar with these before proceeding. It is especially important
to understand what map datums are. A lack of understanding about map
datums can lead to position discrepancies of hundreds of meters. For
example, current USGS topographic maps give the UTM location of Landscape
Arch as 12S-621040-4294243. That's because most US topographic maps
use the North America Datum of 1927 (NAD27) rather than WGS84.
Advantages of the Proposed Standard
The location of a natural arch is its only property that is intrinsic,
fixed, and not shared with any other arch. It is not assigned arbitrarily
by a human observer, it does not change, and no two arches occupy the
same exact space. Furthermore, for most arches, it can be determined
easily and objectively by anyone with basic skills using a GPS receiver
and topographic map. Thus, location is an ideal candidate for uniquely
labeling and indexing arches. Since UTM is by far the easiest location
system to use, the proposed standard is based on it.
The standard is based on WGS84 rather than NAD27 for several reasons.
First and foremost is that it is a world-wide datum. NAD27 is only for
North America. Natural arches on other continents could not be located
correctly using NAD27. Also, WGS84 is the default datum used by Global
Positioning System (GPS) receivers. Finally, the USGS is converting
all future mapping efforts to use WGS84. TopoZone.com
now provides map locations using WGS84 as the default option, and offers
conversion between WGS84 and NAD27 for coordinates within the U.S. Converting
between datums worldwide can also be done at this coordinate conversion site.
Remember also that the proposed standard is for the reference label
only. It is not a standard for how catalogs should present location.
Publishing a location for a North American arch using NAD27 for convenience
in no way contradicts the use of a reference label that is based on
WGS84. For that matter, the location could also be published using longitude
and latitude rather than UTM coordinates.
If all natural arches were indexed by their UTM coordinates, they could
be directly sorted by east-west or north-south position. Moreover, they
could easily be grouped by area. For example, the natural arches within
the boundaries of Arches National Park could be easily determined and
listed. Furthermore, if those boundaries changed, the arches in the
park could readily be re-determined and re-listed without changing the
reference labels of the arches.
When a natural arch became newly cataloged, the Z-E-N (UTM) part of
its reference label would fit easily 'between' the reference labels
of those arches already cataloged. Should an arch be deleted from a
catalog, striking its Z-E-N reference label would not leave an apparent
gap in those remaining. Finally, if all catalogs referenced arches by
their UTM coordinates, the arches that were listed in multiple catalogs
would have the same reference label in all of them. Cross-referencing
between these catalogs would not be necessary. Of course, it would still
be desireable to have cross-references to older, historic catalog numbers.
Finally, having the character string 'NABSQNO' as a prefix enables
reference numbers to be handled as such by computer applications, especially
Internet search engines. To get a demonstration of this, go to a search
engine such as Google
and do a search on [Landscape Arch]. You will get back several hundred
thousand sites, only a very few of which have anything to do with the
natural arch we want information about. Now repeat the search with quotes
to search for the exact phrase, e.g., on ["Landscape Arch"]. This reduces
the returned pages to a few thousand, but still only a very few of these
are on our subject.
Next, remove the quotes and add NABSQNO to the search, i.e., search
on [NABSQNO Landscape Arch]. Only those few pages that pertain to or
reference the natural arch are returned. Now imagine using the Internet
to search on [NABSQNO Arizona], or [NABSQNO Algeria], or ["NABSQNO 12S"].
If we set up Internet pages about natural arches in accordance with
the proposed standard, the first query would return all the pages referencing
natural arches in Arizona. The second would return French arches. The
third would find all the arches in the UTM zone 12S, including Landscape
Arch. Finally, imagine searching on [NABSQNO 12S-620***-4294***]. In
the future, when Internet search engines support searches on wild card
characters (*) within a word, this search will return all the arches
in the same square kilometer that Landscape Arch is in. Of course any
square kilometer could be entered. Furthermore, this capability would
be available without any special database software, just a browser and
access to the Internet. The value of the standard reference label should
now be very apparent.
Issues with the Proposed Standard
There are two possible concerns about using UTM coordinates as the
reference label for natural arches. The first is convenience. NABSQNO
Z-E-N is much longer than any reference label based on A-N. The second
is accuracy. How is the Z-E-N determined? Where on the arch is it measured?
And just how accurate is it really? Let's discuss these in turn.
A 23-character reference label is certainly less convenient than the
4 to 7 characters that the A-N format usually requires. Fortunately,
in practice NABSQNO Z-E-N will be mainly used by computers. People will
still use names and shorter references. But the standard gives computers
what they need for precise and rigorous identification. Of course, it
can be used for the same purpose by people when they want. A key point
is that the standard reference label does not prohibit the use of shorter,
more convenient labels as supplemental reference aids. For example,
if a standard reference label were added to all the natural arches in
Vreeland's catalog, there would be no need to delete his original catalog
numbers. Indeed, these should be preserved for historical reference.
The Stevens/McCarrick catalog adopted the practice of attaching an English
name to all of its arches for the convenience of people. Adopting the
proposed standard reference label does not require that such practice
be discontinued.
Now let's consider questions of accuracy. First, how is the Z-E-N determined?
Currently, the best readily available method is to use a combination
of a GPS receiver and a topographic map to determine the location of
the arch as accurately as possible, and then to transform that location
as needed to the WGS84 datum. As with any method, inaccuracies creep
into the result. We consider the impact of these inaccuracies in the
discussion below.
Where on the arch is the Z-E-N (location) measured? Natural arches
are not points. They have horizontal extent. But reported locations
are points. What convention should be followed for picking a
point on an arch where its location measurement is taken? Ideally, the
convention should be to report the location of the arch's center point
as the location of the entire arch. Unfortunately, it is probably impossible
to exactly define a center point for every type, shape, and size of
natural arch. Even if an exact definition existed, it would be very
difficult for field investigators to correctly apply it every time.
Consequently, there will always be a spread or imprecision in the location
of arches that is introduced by their horizontal extent. This spread
can be estimated to be about 10% of the horizontal extent of an arch.
For the very largest arches, this can be as much as 10 meters. A 10
meter inaccuracy means that the last number in both the easting and
the northing are pretty meaningless.
How accurate are location measurements? With a typical GPS receiver
and map it is usually possible to pinpoint your location, or the location
of a point that is visible to you, with a precision of about 10 meters.
Thus, if several independent observers measure the easting and northing
of the exact same point and report its location to the nearest meter,
even if they are all using similar GPS receivers and the same topographic
map, they will report several different values. If the observers take
the measurements correctly, almost all of their reported locations will
be within about 10 meters of the average of the measurements. This average
is the best estimate of the actual location of the point. The 10 meter
spread is an indication of how accurate a location can be determined
when no mistakes are made. We'll look at the consequences of mistakes
in a bit. For now, note again the implication that the least significant
digit of the easting and the northing are not really meaningful.
But so what? Do we really need an accuracy of better than 10 meters
for the location? In the vast majority of cases, no. If you are told
that an arch is at a certain point, and you draw a circle around that
point with a radius of 10 meters, odds are you will find only one arch
inside it. Even if it is a little bit off, the Z-E-N has uniquely identified
the arch. That's all we need. It's kind of like saying that the Z-E-N
uniquely identifies the one arch that is close to that location.
What about when there is more than one arch in the circle, i.e., when
we are trying to identify a closely packed cluster of small arches.
Even this case is not usually much of a problem. The errors that show
up in the measured location of one arch in a small cluster will impact
the measured locations of the other arches in that cluster in the same
way. For example, the measured locations might all be wrong by 10 meters
in the northeast direction, but their locations relative to each other
will be preserved. If we are only talking about 2 or 3 arches in the
cluster, identification will not be a problem. In the rare cases where
many very small arches are clustered in a 10 meter circle, unique identification
using the Z-E-N is not practical. But in those cases there is probably
little value in unique identification. Instead, it is probably more
valuable to identify the cluster as a single feature. Of course, that
can be done using the Z-E-N for the cluster.
Of course, observers do make mistakes and incorrect location measurements
certainly have been published on occasion. Obviously, when this happens
an error of unknown magnitude is introduced in the location, and hence
in a reference label derived from that location. Normally, mistakes
of this nature are eventually discovered and corrections made to the
published location. But what about the reference label? Should it change
too? Yes, clearly it would have to. But, perhaps surprisingly, this
does not create much of a maintenance problem. One could look at the
correction as the simultaneous deletion of an arch that never existed
(one never did exist at the erroneous location) with the addition of
a new arch (at the correct location). When any catalogs that included
the incorrect reference label got updated, they would likely include
a page that listed the incorrect Z-E-N along with a note about the error
and deletion. Not much different from how an updated catalog might treat
an arch that had collapsed. It would state that the incorrect Z-E-N
had been used in the past as a reference label, but that there is not
an arch at that location.
Guidelines for Using the Proposed Standard
Hopefully, the above discussion is convincing that the UTM location
(Z-E-N) method for identifying arches is a better way, and that it deserves
to become a standard. It may not be perfect, but it is clearly superior
to the other approaches used to date. Before it can become a standard,
however, guidelines must be created and adopted to describe how these
reference labels are to be assigned and managed within the context of
natural arch cataloging efforts. A preliminary set of guidelines follows.
If the guidelines below are followed by everyone who assigns a standard
reference label to an arch, the full benefits of the standard will be
realized while any accuracy problems will be minimized. Of course, one
of those benefits is the continuing development of an Internet based,
on-line database of natural arches. The guidelines below should also
be kept in mind when using the standard to query that database. Note
that, once locations accurate to 1 meter are easily obtained in the
field, some of these guidelines will be unnecessary and will be revised.
Note also that these guidelines apply to the reference labels only,
not to published locations.
- Place the reference label prefix (NABSQNO) only on web pages intended
to convey documentation about an arch, e.g., photographs, location,
type determination, names, history, dimension measurements or estimates,
or any other standard
attributes of the arch. Do not include the prefix on a web page
that merely refers to the arch, e.g., an index to other pages that
do provide documentation. In these cases, refer to the arch
with just the prefix N. For example, an index page might include Landscape
Arch in a list with the reference N 12S-620979-4294446.
- Include, as appropriate, the country, state, province, and/or county
in which the arch is located somewhere on the page. Include any names
that are associated with the arch on the page.
- Only have one arch (and one reference number) per page unless the
multiple arches shown are in a cluster that you want to document as
a single feature. This will avoid the search engines mixing data from
one arch with the reference label for another arch.
- Only assign a full reference label if you are certain of the location
to 10 meter accuracy. If you know it to less accuracy, you may assign
a temporary label that reflects this fact. Substitute the character
'X' for the unknown digits of the easting and northing. For example,
if you knew the location of Landscape Arch to only 100 meter accuracy,
you might assign a temporary reference label of NABSQNO 12S-6209XX-42944XX.
If you knew it only to 10 kilometer accuracy, you might assign NABSQNO
12S-62XXXX-429XXXX. However, if you only know the UTM zone for an
arch, follow the zone with the abbreviation 'TBD' for both easting
and northing. For example, Landscape Arch might be assigned the temporary
label NABSQNO 12S-TBD-TBD. TBD stands for 'to be determined.' Obviously,
such temporary labels do not uniquely identify the arch, they are
merely place holders so that arches with poorly determined locations
can be partially cataloged.
- Before assigning a reference label to an arch, ALWAYS check to make
sure that one has not already been assigned. NABS endeavors to maintain
a complete set, so it is best to check with us prior to making an
assignment. This will avoid multiple reference labels for the same
arch. Inquiries of this nature, and any reference numbers you assign,
should be sent to info@naturalarches.org.
- Always assign the easting and northing parts of the reference label
to the nearest meter, even though the location is only known to 10
meter accuracy. It should be understood that the last digit of the
easting and northing in a full reference label are arbitrary.
- Always assign the reference label based upon the WGS84 datum.
- Once a full NABSQNO Z-E-N reference label has been assigned to an
arch, it should only be changed if: (1) it is discovered that the
location it is based upon is wrong by 100 meters or more, or (2) it
is necessary to avoid confusing it with another, close-by arch. For
example, if the location of Landscape Arch were found to be actually
40 meters to the north of the location that its reference label is
based upon, the reference label for Landscape Arch should not be changed,
even though the published catalog location is refined. Remember that
a 40 meter shift is less than the size of this arch.