Activity Feed › Discussion Forums › Strictly Surveying › Astronomic Azimuths
And don’t leave out the software. Back in the day most of the time spent was in reduction.
So with lat long, precision time, quartz stop watch, defection of the prime vertical database, and software, that makes azimuths: Better Faster Cheaper. A good reason to keep the skill alive.
Really learning Solar azimuths teaches a lot.
Doing Polaris shots in daytime requires an extended sun hood. One can be made with a toilet paper core. Cut it the long way and use two with overlapping gaps for larger diameters. Flopping the scope requires removal of the hood. Flat black paint inside the hood helps. A permanent black marker is OK as the cardboard surface breaks up reflection. Set a North reference stake on an early shot so one can reacquire Polaris quickly after any observation break. It speeds up observations having a timekeeper/notekeeper.
Question, assuming astronomic tables are in real time, GPS time requires further adjustments?
Paul in PA
- Posted by: @paul-in-pa
Doing Polaris shots in daytime requires an extended sun hood.
I tried a 11″ tube made from black paper and couldn’t see any difference at the time Polaris was fading away in the morning. So the clarity of the atmosphere is also important.
Question, assuming astronomic tables are in real time, GPS time requires further adjustments?
You hardly ever see GPS time displayed. Most devices will convert it to UTC for display (using the leap second count). Precise astro work requires UT1 (that takes into account more subtle effects), which is within 1 second or less of UTC.
. You almost never see a true gps time except in a rinex file. Receivers display UTC.
For sun shots, UTC needs to be amended to UT1. That correction +/- few tenths, can be found online, Bulletin A. The current DUT correction:
“DUT1= (UT1-UTC) transmitted with time signals = -0.2 seconds beginning 02 May 2019 at 0000 UTC”
https://www.iers.org/IERS/EN/Publications/Bulletins/bulletins.html
A guy who posts here once in a while wrote a program for your HP calculators that we used as DC’s way back when. He would update them yearly with a new ephemeris and all that was needed was to push the enter button when you were on the leading and trailing edges. The program would then calculate the AZ, of course it worked best when you did multiple observations. It was quick and accurate. Big improvement over the hand calculations.
Although, after some years we stopped using the HP calculators and if I wanted to do a Solar I had to sit down and hand calculate them. Kinda going backwards.
It really is something all surveyors should know and do at least once, but I’m probably never going to do another one, been there, done that.
I was brought up around solar observations. Dad wrote a basic program back in the 80’s to reduce them. We hand entered the programs in our HP calculators from the back of the Elgin, Knowles, and Senne books in the 90’s. Later we had a TDS data collector that would reduce them and directly read the angle from the instrument and time from the internal clock making them very fast to acquire.
In my opinion it is a tragedy that celestial observations have not been fully embraced by surveyors everywhere. All of our surveys refer to “North” but few seem to know how to determine North and what it actually is if they did (geodetic, astronomic, grid, etc.)! So most references to “North” are vague at best. Even today, with GNSS being ubiquitous, many surveyors don’t seem to understand North given their metadata statements.
I don’t know why celestial observations are not even more widespread today than they were in previous generations. Technology is available right now that should allow celestial observations to be acquired and reduced with high accuracy in seconds. GPS provides time and position and accurate engines can be installed on instruments and collectors. Robotics can track objects in bitonal images, making even then act of sighting automatic.
I realize I’m a salesman for Javad, as my profile indicates, but I’ll offer that the J-Mate (currently in beta-development) is designed to automatically find the Sun, make observations, and reduce the observations in seconds, with virtually no input from the user. The hard work once required to get accurate directions can be eliminated with current technology, but it seems few ever realized the significance of them.
E&K AstroROM module for the 41 is available for i41CX+.
I shoot the sun regularly for kicks. I wrote a comprehensive spreadsheet. Reduction made easy. I have a little GPS receiver with a pps strobe for UTC time. It’s easy peasy and incorporating Xi Eta it’s a whole lot better than before. Obs still take close to an hour, but reduction is 25-20 minutes.
However, I guess it’s mostly nostalgic
This site may be of interest to those seeking ephemeris information. I haven’t tried it myself.
https://wordpress.nmsu.edu/kwurm/ephemrides/
.Surveying in the 1970’s and 80’s i would do a Sun shot on every survey. Sometimes 2 or 3 a day. Even had a Sunshot calc on my PLS test. I tried to teach others how to shoot the Sun and Polaris but they never got the hang of it. I still shoot Polaris sometimes just for fun.
Studied like hell back in early 90??s because there was a chance that it would be on the state (Ca) license test. So I once could do it in theory but never did one for real. Jp
Starting about middle of the year 1990, every survey I did was oriented to grid north using a sun shot and calculating the theta angle. I used the program from the Elgin Knowles and Senne book keyed into my HP 41 and also used the 41 for time. I could set the time from a radio shack time radio and adjust for the positive or negative adjustment based on the number of ticks on the radio. I always did 8 sets, 2 direct leading edge, 2 direct trailing edge, and the same thing inverted. I had the 41 programmed to stop time with the push of a button so I could watch the scope with my finger on the button and mark time very accurately.
With the 41 app for smart phones, I’m guessing anyone with a smart phone could still do the same thing. It might be hard to get time.
James
My memory’s fading but didn’t the original HP41C factory Survey Module have a solar reduction routine based on hardwired Chebyshev polynomials for the sun instead of contemporary Ephemeris data which needs regular updating? That meant the computation was accurate when the module was new (for a coupla years or more), but as the decades have passed the hardwired polynomials for the sun’s position drifted to where the module is now useless. ??? It could be i’m mis-remembering and I was using a keyed in (card reader) program from Joe Bell’s excellent HP41 Journal.
Concerning accurate time an HP41 with a time module or HP41CX with integrated time module, after a few months of the correction routine for crystal the 41C was the finest autonomous time standard available for civilian applications, like surveying; assuming you kept it reasonably temperature stable.
Observation of the Sun’s limbs instead of its center by using a Roelofs prism introduces several errors, most notably pointing error and to a lesser extent semidiameter error dependent on solar activity, and the extended duration of the observation as the sun moves involves atmospherics which cannot be corrected for.
That being said, even with meticulous technique I doubt astronomic north can be determined by solar observations using tripod mounted optics to within 10-15″. Yes, precision to within 5″ has been anecdotal, but really determining north by observing the Sun is essentially shooting a fast moving target as time passes.
Much better is to observe Polaris. The target is moving slowly, or briefly not at all for elongation observation. Polaris is a pinpoint in the sky so pointing is not a problem. It’s night (although observation during the day is possible), so atmospherics are minimal. Tripod, instrument thermal stability is maximized. I’d guess accuracy can be 5″ or less with careful procedures. It’s as if you have a target at near true north millions of miles away. The stinker is you’re observing a target at a high elevation, OK @ 20° latitude, useless @ 70° latitude because unavoidable instrument levelling errors degrade the observation. But 99% of us are in CONUS so it’s not a problem.
- Posted by: @mike-marks
My memory’s fading but didn’t the original HP41C factory Survey Module have a solar reduction routine based on hardwired Chebyshev polynomials for the sun instead of contemporary Ephemeris data which needs regular updating?
I never used any of the survey modules. I wrote my own except for the sunshot program from Elgin Knowles and Senne.
I never used any of the survey modules. I wrote my own except for the sunshot program from Elgin Knowles and Senne and I modified it to give me North and then Grid North in my zone.
- Posted by: @mike-marks
a solar reduction routine based on hardwired Chebyshev polynomials … That meant the computation was accurate when the module was new (for a coupla years or more), but as the decades have passed the hardwired polynomials for the sun’s position drifted to where the module is now useless.
No idea what it used, but I have a picky point about Chebychev polynomials and the word drift. Those polynomials are used to provide as good an approximation to a function as their degree allows over a selected period of time, wiggling between a small plus or minus bounded error. Then outside that period of time the error doesn’t drift, it shoot rapidly to huge values.
. A few years ago, I observed a radio tower strobe light 5-12 mi distant, from 9 widely separated points. Using opus derived coordinates and dozens of solar observations, plenty of star observations, over many weeks. I have a coordinate on the strobe. Starnet returns residuals on the azimuths less than 8”, and only 8-10 rejections, leaving 40+ azimuths in the adjustment.
Now to check the coord on the tower I set a new opus point (not included in the LS adjustment), take a solar azimuth and compare that to inverse. On one such point, I have 5 solar azimuths observation. 3 return the same second and hit inverse to 1.5 seconds. And 2 obs are 4 and 5 secs off inverse. That’s not a one-off. However, there obs that look good but don’t fit for 10”. I repeat those, and find the first obs to be anomalous.
Also deflection of the prime vertical was not easily had years back. Change in semi diameter due to Solar activity is well inside the noise
I use a dedicated pps output from a GPS receiver for time. And Quartz stopwatch. I use the multi foresight method. I only obs the sun below 30 degrees vertical angle.
I repeated the radio tower intersection, new points, solar shots, 6-12 mi distant. The coord from the first run is less than 30 cm from the second run. (The strobe lights appears to be at least a meter in size.)
Sun shots are not that easy, or fast. But there’s no mathematical reason against it. True UT1 time is far more readily available as well. Time has to be curated carefully.
- Posted by: @larry-scott
On one such point, I have 5 solar azimuths observation. 3 return the same second and hit inverse to 1.5 seconds. And 2 obs are 4 and 5 secs off inverse. That??s not a one-off.
Impressive. I can’t meet such repeatability (much less precision) using T2 type or better terrestrial theodolites on stable concrete fixed mounts with optimal targetry in the 2-4 mile range, even with favorable atmospherics (dam subsidence deflection surveys). Be aware an arcsecond is approximately the angle subtended by a U.S. dime coin (18 mm) at a distance of 2.5 miles. You state the strobe tower was @ 5-12 miles distant so you are claiming solar observation repeatability better than field survey instrument capabilities by a wide margin.
OTOH, maybe you’re claiming after a least squares adjustment many observables were not adjusted by more than 1.5″, and all were within 8″ after adjustment and you chucked the +10″ observations (usually not statistically rejectable). Well done precise survey but be a little humble and certify you located the strobe tower within a few feet if that was the contract.
My experience with solars is they’re no better than 15″ even with careful procdedueres & anybody who claims better is blowing it up their ass’s clients.
I mostly used SMI’s local hour angle method with the HP41 and HP48.
With the canopy density of some of the locations around here it was nice to obtain a static position on as many points around a forest and make a sun shot from them and either traverse thru the static point and the azimuth point we had set as we measured around the project tract.
My best results for sun shots were when I would turn to a point a few seconds in time past the sun and mark the time when the sun’s leading edge touched the vertical hair of the instrument.
Mostly after 6 repetitions, the computed azimuth was ?ñ a few seconds at most from any of the single shots.
I believe the ephemeris ended a few years after SMI ended making programmed cards. Suppose I should give it another try between two points in a rather NS direction from one another that I have observed static data on and see how it compares.
Correct time and a well leveled instrument are key to that method.
0.02
To be sure repeatability 1-2 seconds cannot be counted on. I’ve worked hard on refinements of the observation procedure over the past couple of years. And a comprehensive reduction excel sheet.
Across OPUS derives points, at 11 mi, I do get repeatable at 1-5 seconds. And hit inverse to the same. Not just once
So I realize that in a commercial project in today’s gps environment solar az is not financially viable or warranted, or even a good idea.
It’s in the minutia. Eliminating micrometer scaling error, advanced leveling procedures, gps pps time source, multi foresight, practice practice practice, and multi day. USNO ephemeris. Im not pretending to get better than the instrument is capable in any one reading or set, or one-off lucky shots.
I like the art of observation. 15” ‘best case’ is not my experience. 15” is a do-over.
Some Instrument operators are better than others, that’s just the way it is. There is no doubt in my mind that Larry Scott and
A. Harris are 2 of many.
I would suggest you read ” Astronomy Applied to Land Surveying by R. Roelofs” (1950). This will give you a very good
understanding of different types of errors in the Wild T2 and T3 theodolite or other instruments that are the same.
You should also read Prof. Art Peterson papers on astro. See ACSM-ASPRS annual convention 1986, Vol. 2, pp177-185.
See also “Accuracy of the Automatic Grid Azimuth Determination by Observing the Sun Using Kern E2 Theodolite, by
Nikola Solaric and Drago Spoljaric, in Surveying and Mapping 1988. The next paper you will want to read is Accuracy of Automatic
Grid Azimuth Determination by Astronomical Methods with the Leica-Kern E2 Theodolite by Nikola Solaric, Drago Spoljaric, and Zedenko Lukinac, Surveying and Land Information Systems, Vol. 54, No. 1, 1994, pp5-19
There are many other papers but you can find them. Then go out and practice taking at least 50 sun shots (different days)
using what you have learned from all the references above and more and let us know how you did.
JOHN NOLTON
In my life I have done sunshots twice. Both in 1994. Not a once with the left, once with the right joke. Glad that I can say that I did it, glad that I never have to again.
Log in to reply.