radial-vs-traverse surveys
Mark Williams
What realistic diference in acuracies can be expected when comparing a
good balanced traverse and a carefully performed radial survey
performed with current insturments?
Mark
S. R. Sheffield
mar...@hpnts.net (Mark Williams) wrote:
Without getting into the Theory of Error, the fact is, any
observation has a chance for a "bust" and a small amount of error
built into the instrument and operator, in my opinion, makes it
necessary to have a check of some kind. It this were a perfect world,
then the difference between closed and radial staking would be nill.
We have found that the big problem is not so much the small error
incurred but the "blunders"; ... back sighting on wrong point,
Entering wrong point no. for station or backsight, Bumping "0 set"
in the middle of a series of shots (topcon), etc.
Unfortunately, the is not a perfect world so I would suggest,
checking, checking, checking.
The problem is not so much whether you have a 1:10,000 to a 1:15000
closure, but if you have a 1:5 closure.
Steve,
Take good advice...... even if it comes from a fool.
Kent McMillan
In article <335199c2...@news.nts-online.net>, mar...@hpnts.net (Mark Williams) says:
>
>What realistic diference in acuracies can be expected when comparing a
>good balanced traverse and a carefully performed radial survey
>performed with current insturments?
GIVEN points 1 through 13 with the following approximate co-ordinates:
Pt. No. N (ft.) E (ft.)
1 7325 2000 Boundary Corner
2 7168 3580 CP
3 7135 5791 CP
4 7325 7194 Boundary Corner
5 5236 7403 CP
6 3575 7106 CP
7 2045 7240 Boundary Corner
8 2200 5546 CP
9 1903 3632 CP
10 2000 2000 Boundary Corner
11 3628 2099 CP
12 5663 1890 CP
13 3650 4310 CP
D I A G R A M:
*1 *2 *3 *4
*12 *5
*13
*11 *6
*10 *9 *8 *7
If you use an instrument which can measure angles with a standard
error of +/- 6 seconds and distances with a standard error of
+/- 0.01 ft. + 3 ppm, how would you expect the accuracy of the
calculated courses connecting the Boundary Corners to compare if:
(a) they were connected by the closed traverse 1-2-3-4-5-6-7-8-9-
10-11-12-1, with the azimuth of 10-11 being determined by
observation with a standard error of +/- 4 seconds ? or
(b) they were surveyed radially from CP 13 in the interior of the
tract with the azimuth of 13-10 being determined by observation
with a standard error of +/- 4 seconds ?
THE ANSWER that Star*Net Least Squares Adjustment software provides
is (assuming instrument centering of +/- 0.003 ft. and target centering
of +/- 0.01 ft.):
(1) The accuracy of the azimuths of the calculated courses between
boundary corners would be nearly identical (standard error of +/- 6
seconds) for both methods.
(2) The distances between boundary corners calculated from the traverse
data would have standard errors of +/- 0.03 ft. whereas the distances
from the radial survey would have standard errors ranging from +/- 0.05
ft. (course 7-10) to +/- 0.15 ft. (course 1-4)
CAVEAT: the above assumes that the data is without blunders, that the
actual accuracies of the instrument have been attained, that no member
of the survey party was recovering from an encounter with the brewer’s
art, and that none of the many unpredictable things that go wrong in
the real world of human endeavor in this case did.
IN ACTUAL PRACTICE a surveyor would need to use an error analysis program
such as Star*Net to decide what accuracies had most likely been attained
on a particular survey. The very great difference between the two
methods is that the radial survey method has the greater potential to
produce a big fat blunder that goes undetected.
Numbers aside, here are some of the PROs and CONs of radial surveying
as they strike me:
PRO
1. Faster, especially in rough terrain
2. Accuracies can be comparable to conventional traverse or in
some cases better
3. Instrument stations can be chosen to improve seeing on
lines surveyed (higher ground clearance, faster work in good
seeing conditions, such as later afternoon).
CON
1. Tendency to overreach, overextend sight lengths
2. Illusion of exactness (no traverse closure error)
3. Takes the survey away from the boundary lines and evidence to
be sought,
4. Complicates setting out new stakes if sparse, distant control
points are all that are in place,
5. Magnitude of potential blunders,
6. Analysis of errors becomes critical in rational design of survey
7. Requires more sophisticated understanding of measurement errors
by field party, more difficult to give general instructions in
advance without knowledge of actual conditions.
SUMMARY I use radial survey methods when they are appropriate, but
I would not want to without Star*Net.
regards,
Kent McMillan, RPLS
Austin TX
CoPLS
(Mark Williams) says:
>
>What realistic diference in acuracies can be expected when comparing a
>good balanced traverse and a carefully performed radial survey
>performed with current insturments?
I suppose that both are sufficient, provideing that one provides
proper filed checks.
Obviously the best and most accurate method of collecting field
data is to traverse through and occupying each found monument, turning
multible angles and shooting resipical distances.But the reality is, that
almost every land surveying firm that I know (at least in private
practice) shot everything radial, with maybe a couple check
angles/distances to previously tied to points.
Does this create any problems? Maybe/maybe not. The LS in charge
must make his/her own call concerning this. And until (if) state
regulations are passed addressing this issue a surveying firm in private
practice will operate in a method that takes the least amount of time. But
there's another issue -
who will police it? and how would they???
Yes I know what you are thinking, ALTA/ACSM has stanards concering
this methods of measurement. I would assume that every surveyor that
conducts an "Urban Survey" has an instrument that reads to less than 10";
and turns 2 sets of direct and reverse angular observations; that there is
no more than 5" of spread from the mean angle; that the angular closure
closure does not exceed (10" x square root of the number of stations);
that the linear closure is better than 1:15,000; that all distances are
measured with an EDMI or double chained w/ steel tape; and that the
minimum length of distance measured is no less than 81m.
But, like I said, I am sure that all surveyors that conduct these
types of surveys operate under the terms of the ALTA specs.
Ronald Flanagan PLS
Colorado Engineering & Surveying Inc. - Englewood, Colorado
Professional land surveying services since 1972
Kent McMillan
In article <19970501140...@ladder01.news.aol.com>, co...@aol.com (CoPLS) says:
> Yes I know what you are thinking, ALTA/ACSM has stanards concering
>this methods of measurement. I would assume that every surveyor that
>conducts an "Urban Survey" has an instrument that reads to less than 10";
>and turns 2 sets of direct and reverse angular observations; that there is
>no more than 5" of spread from the mean angle; that the angular closure
>closure does not exceed (10" x square root of the number of stations);
>that the linear closure is better than 1:15,000; that all distances are
>measured with an EDMI or double chained w/ steel tape; and that the
>minimum length of distance measured is no less than 81m.
> But, like I said, I am sure that all surveyors that conduct these
>types of surveys operate under the terms of the ALTA specs.
>
I'm personally not a great fan of the portion of the ALTA/ASCM (American
Land Title Assoc./ American Congress on Surveying & Mapping) Minimum Standard
Detail Requirements for Land Title Surveys that attempts to address survey
accuracy, the table of "Minimum Angle, Distance and Closure Requirements for
Classes of Surveys". The intent behind standard specifications is fine:
to provide a shorthand way for a person to ask for a survey of a particular
quality and scope of service. Think of how difficult it would be to buy a
truck if everyone who went to a dealer had to describe what they thought a
truck should include. "Do you want a steering wheel?" "Yes, I almost forgot,
we'll need one of those, too." But unlike trucks, there does not seem to be a
general understanding of what a Land Title Survey should include.
Since we do live in the age of cheaper and faster, without some formal statement
of what the product should be, the steering wheel (metaphorically speaking) seems
to be the first thing to go in a marketplace seeking a cheaper and faster survey.
So, generally, I think the front part of the Minimum Standard Detail Requirements
are O.K.
On the other hand, though, the table of "Minimum Angle, etc." is a weird little
concoction. I understand the intent, but what a strange road to that goal.
If I understand the specification correctly, if I am using my total station with
an EDM with a standard error of +/- 5mm + 2ppm, I should not use it to measure any
ranges less than 81m (266 ft.). I don't know when the last time a surveyor
taped across busy arterial streets and interstate highways, but I assure you
whenever it was it wasn't anyone I know.
Actually, the specification reads: an "EDM having an error of 5 mm ... ", which
strikes me as pretty funny. If I knew that my EDM distances were in error by 5 mm,
I'd change the instrument offset correction.
Since we live in an age that is producing measuring technologies faster than the
ALTA/ACSM spec gets updated, the real question is how to specify survey accuracy
in a way that is independent of the instruments and methods used, to specify the
accuracy of the results.
Best regards,
McKenney
I don't survey to these standards but I do survey to our state standards
(NH). Seems to me the Alta standards place lots of emphasis on how to
measure but not on where to measure from and to. A famous surveyor/author
correctly states that it is better to have faulty measurements between the
actual corners than precise measurements between points that are not
corners. Seems like these standards get hung up on the precision of
measurements and not the accuracy of measurements. The real art of
surveying is to make accurate measurements as well as precise measurements.
I see lots of precise surveys which lack basic information about whether a
monument was found or set; its condition; its size and descriptive details.
I'd rather know these things with measurements to the nearest foot than
lack this info but know angles and distances to seconds and hundredths.
Mind you, I survey in a rural area and many surveys are done in remote and
rugged terrain. It's my practice to mark witness trees at each monument and
to blaze and paint the lines between monuments. Since I also manage many
tracts for timber and forest products, it's important to know where the
line is on the ground not just on paper. Most surveyors think if it looks
good on paper then the job's done; landowners should know where on the
ground their property lies.
Well, this is just my humble opinion. The surveyor is a professional and
therefore should be able to use or modify methods to do the work, not just
follow a cookbook.
Kent McMillan
In article <01bc5743$82aafac0$670320ce@mckenney>, "McKenney" <d_mck...@conknet.com> says:
>
>
>
>I don't survey to these standards but I do survey to our state standards
>(NH). Seems to me the Alta standards place lots of emphasis on how to
>measure but not on where to measure from and to. A famous surveyor/author
>correctly states that it is better to have faulty measurements between the
>actual corners than precise measurements between points that are not
>corners. Seems like these standards get hung up on the precision of
>measurements and not the accuracy of measurements. The real art of
>surveying is to make accurate measurements as well as precise measurements.
I think that the distinction that I would make is between ACCURACY and
CORRECTNESS. A surveyor who reports a distance of 5278.02 ft between
monuments that are actually 5281.85 ft. apart has made an INACCURATE survey.
A surveyor who reports a distance of 5278.02 ft. between monuments that are
indeed 5278.02 ft. +/- 0.25 ft. apart has made a fairly ACCURATE measurement by
most commonly circulated standards for land survey measurements, but may not
have made a CORRECT survey if the monuments are reported as being the corners of
the estate surveyed when in fact the corners may be shown to be in some other
significantly different location.
I do not though view CORRECTNESS and ACCURACY as an either/or proposition.
In modern practice, surely what we are after is both. I agree with you that the
ALTA/ACSM accuracy spec is ill-suited to accomplishing both with a reasonable
economy. It seems to me that it is a good bit easier to develop a general
specification that addresses survey CORRECTNESS by describing the scope of the
survey investigation than it is to specify survey ACCURACY in a way that deals
only with the results in a way that most land surveyors would readily understand.
In Texas, our licensing board has addressed the problem by defining
three different classes of surveys, Urban, Suburban, and Rural, and by requiring
that survey measurments (bearings and distances) between any two monuments positioned
by the survey not have a relative uncertainty greater than certain different limits
established for each class or survey. The board wisely decided to remove itself from
the practice of telling land surveyors how to measure the positions of monuments,
and addressed only the actual reported results of the measurements.
As a matter of professional practice, however, the analysis of a survey to determine
with a reasonable level of certainty that the accuracy standards have been acheived
is a bit tricky. A program such as Star*Net will analyze the uncertainties in the
inversed courses between any and all points on the survey, but for a boundary with,
say, twenty monuments the number of courses to be examined is in the hundreds.
Kent McMillan
In article <01bc5743$82aafac0$670320ce@mckenney>, "McKenney" <d_mck...@conknet.com> says:
>
>
>
>A famous surveyor/author correctly states that it is better to have
>faulty measurements between the actual corners than precise measurements
>between points that are not corners.
A.C. Mulford actually wrote "For after all, when it comes to a question
of the stability of property and the peace of the community, it is far
more important to have a somewhat faulty measurement of the spot where
the line truly exists than it is to have an extremely accurate measure-
ment of the place where the line does not exist at all."
I apologize if it seems annoyingly pedantic to distinguish between
Accuracy and Precision, but the two terms do have fairly well established
meanings in the modern American surveying textbooks (Moffitt/Bouchard,
Davis/Foote/Anderson/Mikhail).