Perhaps so, but that's not the right way for them to do it. The corrections should be applied to the slope distances before reduction to horizontal and vertical.
Maybe it's just curvature that is not applied...
I recently had the touchscreen replaced on my data collector. In the process, the manufacturer also loaded the latest version of SurvCE, and shipped it back to me. I didn't notice that the units for temperature had changed from Fahrenheit to Celsius, and as a result, on two separate occasions before I noticed the error, I had surveyed with the temperature set to 60 degrees Celsius instead of 60 degrees Fahrenheit.
I use a Kestrel pocket weather meter to measure temp, pressure and rel. humidity.
From the instruction manual for my instrument (a Sokkia SRX 3" robot), I've got the formula for calculating the PPM correction. Using this formula I have calculated that the incorrect PPM value (using 60 deg. C) was about 47 PPM, whereas it should have been about 8 PPM (if I had correctly entered 60 deg. F).
So, are my measurements too long, or too short? I'm thinking my measurements are too long.
I'd love it if someone could check me on this. My thinking is that the instrument measures length on the basis of some number of wavelengths of light, and calculates a raw distance therefrom. Then, as a second step, it calculates the true or adjusted distance by multiplying the PPM factor to the raw distance. Therefore, if the PPM factor was about 39 PPM larger than it should have been, the distances would be consistently longer by about 39 PPM, and therefore to correct for this error, I would divide the measured distances by 1.000039. Does that sound correct?
Thanks for any help.
Al
rfc
The rule of thumb is for a change of 1.8 degree F or 1.0 degree C you will change by 1ppm. Your 500 foot distance might be off by 0.0125 feet
NO big deal I would think.
FOR: Moe Shetty:
The EDMI that NGS uses are the Wild DI2002 and the manufacture states 1mm +/- 1ppm, NOT 0.1mm +/- 1ppm. Also note
that NGS does base lines NOT baselines.
When NGS did the Corbin base(after they put in pillar monuments) the City of Los Angeles loaned them its Kern Mekometer ME5000,
rated at 0.2mm + 1ppm.
JOHN NOLTON
Big Al, post: 373582, member: 837 wrote: I recently had the touchscreen replaced on my data collector. In the process, the manufacturer also loaded the latest version of SurvCE, and shipped it back to me. I didn't notice that the units for temperature had changed from Fahrenheit to Celsius, and as a result, on two separate occasions before I noticed the error, I had surveyed with the temperature set to 60 degrees Celsius instead of 60 degrees Fahrenheit.
I use a Kestrel pocket weather meter to measure temp, pressure and rel. humidity.
From the instruction manual for my instrument (a Sokkia SRX 3" robot), I've got the formula for calculating the PPM correction. Using this formula I have calculated that the incorrect PPM value (using 60 deg. C) was about 47 PPM, whereas it should have been about 8 PPM (if I had correctly entered 60 deg. F).
So, are my measurements too long, or too short? I'm thinking my measurements are too long.
I'd love it if someone could check me on this. My thinking is that the instrument measures length on the basis of some number of wavelengths of light, and calculates a raw distance therefrom. Then, as a second step, it calculates the true or adjusted distance by multiplying the PPM factor to the raw distance. Therefore, if the PPM factor was about 39 PPM larger than it should have been, the distances would be consistently longer by about 39 PPM, and therefore to correct for this error, I would divide the measured distances by 1.000039. Does that sound correct?
Thanks for any help.
Al
John didn't answer YOUR question, so I'll try. I don't think you can just subtract the factors ("39 ppm LARGER than"). Wouldn't you first divide by 1.000039 to get things back to "uncorrected", then multiply by 1.000008? In otherwards, multiply your distances by .999969? Slightly different that what you get doing it your way (.999961).
Moe Shetty, post: 316414, member: 138 wrote: I was recently involved in the measurement/recording of a new EDM calibrated baseline. It was fascinating to see NGS's EDMs. ...snip ...
It is ironic to me that finding a site to build a baseline can be more difficult than anything else.
Thanks for the link. Does this new CBL include the pillar monuments? Mr Malcolm Archer-Shee, formerly of the Md State Hoghway Administration, was instrumental in achieving the high density of HARN points and a good number of CBLs.
As you undoubtedly noted when observing the measurements, atmospheric corrections are taken seriously. Two Thermistors at different heights were measured at both end points of measurements as well as barometric pressure, notes about wind speeds and cloud cover. At least that was the way it was done when I was involved (last time about 2000).
I would always prefer these discussions of ppm corrections to start with a statement that the instrument's performance was validated on a CBL. Better still would be a reference to making a complete set of measurements rather than a mere " I set up on one point and measured one or two of the other points and they checked pretty good."
Of course many areas do not have easy access to a CBL. As you note, the NGS has a program to help groups interested in establishing a CBL. The policy statement is in the document you linked.
Thanks again for the link. I had lost track of developments in this area. I was a little disappointed that the list of references did not include Reuger's text although it did include an article. Australia has an excellent EDM CBL program. It also uses a variety of configurations. See the State of Victoria's handbook here:
I helped out on the Sacramento CBL a few years back. My role ranged from mark setting to recording on the HP-200LX (or whatever it was). Mostly what I remember is that it was 108å¡F the day we took the measurements, and I was wiped out by the end of the day.
I would say that a tribrach out of adjustment would cause equally significant errors if not rectified.
Moe Shetty, post: 316414, member: 138 wrote: I was recently involved in the measurement/recording of a new EDM calibrated baseline. It was fascinating to see NGS's EDMs. They used two identical EDMs, with accuracy specs that I didn't previously know existed: 0.1mm +/- 1ppm.
Prior to coming out to Maryland, Steve Breidenbach, from NGS, ran a calibration on the Corbin baseline in Fredericksburg VA at the NOAA NGS geomagnetism lab.
Drive out to Hancock MD to the new line, run the measurements. Occupy each of the four markers, measuring each of the other three legs. We used thermistors and barometers and recorded temperature and pressure before and after each block of distance measurements. Mr. Breidenbach briefed us on many of the particulars; one of which was that a one degree celsius temperature difference can cause a one part per million change, and it is an EDM like theirs that can 'notice' that.
He then drove all his gear back out to Fredericksburg to run the calibration routine again. Buckets of work.
It is ironic to me that finding a site to build a baseline can be more difficult than anything else.
That is tight- I think the best production edm in a total station is around .5mm +1ppm
pmoran, post: 373657, member: 8922 wrote: That is tight- I think the best production edm in a total station is around .5mm +1ppm
yes, i'm scouring my memory of what model it was. i do remember it was a leica, and that NGS had bought two for calibration projects. i think that particular instrument was built and marketed for industrial metrology, but not a production edm. we had umbrellas on edm and targets and edm stations had 'thermistors' (not sure this is appropriate name) set up for monitoring temp and pressure, measurement for measurement
GeeOddMike, post: 373616, member: 677 wrote: Thanks for the link. Does this new CBL include the pillar monuments? Mr Malcolm Archer-Shee, formerly of the Md State Hoghway Administration, was instrumental in achieving the high density of HARN points and a good number of CBLs.
As you undoubtedly noted when observing the measurements, atmospheric corrections are taken seriously. Two Thermistors at different heights were measured at both end points of measurements as well as barometric pressure, notes about wind speeds and cloud cover. At least that was the way it was done when I was involved (last time about 2000).
I would always prefer these discussions of ppm corrections to start with a statement that the instrument's performance was validated on a CBL. Better still would be a reference to making a complete set of measurements rather than a mere " I set up on one point and measured one or two of the other points and they checked pretty good."
Of course many areas do not have easy access to a CBL. As you note, the NGS has a program to help groups interested in establishing a CBL. The policy statement is in the document you linked.
Thanks again for the link. I had lost track of developments in this area. I was a little disappointed that the list of references did not include Reuger's text although it did include an article. Australia has an excellent EDM CBL program. It also uses a variety of configurations. See the State of Victoria's handbook here:
Malcolm was there on that very project. He had written a new program for reducing CBL data, and was testing it as the measurements were made.
rfc, post: 316364, member: 8882 wrote: On a recent outing, I set the temperature and pressure on the gun to what it was in the am (about 45 degrees F), and forgot to increase it as the day progressed (to nearly 80 degrees F).
Wondering if there was a way to "post process" measurements and change them according to what the temperature actually was at the time, I wanted to see what the errors might be, so I set up a quick test at about 100' and sequentially set the temperature in the gun from 35 degrees up to 80 degrees.The distance did not change much at all across that range, and not linearly...it seemed more random and residuals were on the order of less than +/-.001'.
When I checked the ppm, it did indeed change (+ above 68 degrees or so, - below)as it should, but what's the use of setting it frequently if it doesn't make a difference?
What am I missing?
You can't test it at 100 feet. EDM's can't determine atmospheric correction at less than 200 feet. So at 100 feet you can expect to get fluctuations in the measurement greater than the specs. of the edm. as it tries and fails to resolve properly. Or so it used to be stated in the literature. At least one set of old ALTA instructions required steel taping distances less than 200 feet. One of the reasons the infrared spectrum is used so much is that it is least affected by atmospheric conditions. Temperature isn't going to affect it much, but pressure does have measurable affects if you're working at 10,000 feet rather than 500.
Moe Shetty, post: 373677, member: 138 wrote: yes, i'm scouring my memory of what model it was. i do remember it was a leica, and that NGS had bought two for calibration projects. i think that particular instrument was built and marketed for industrial metrology, but not a production edm. we had umbrellas on edm and targets and edm stations had 'thermistors' (not sure this is appropriate name) set up for monitoring temp and pressure, measurement for measurement
When I was at DMA we established a baseline at Patrick AFB in FL- right along A1A. We had 1000's of measurements on the line between a half dozen pour in place monuments. We used a K&E Ranger VA and HP 3808A and a K&E Rangemaster II for the prelims and then a crew from our detachment at White Sands Missile Range came out with a geodolite- that thing was the beast. of course when ever possible we observed mets at both ends. That baseline was super tight.
rfc, post: 373600, member: 8882 wrote: John didn't answer YOUR question, so I'll try. I don't think you can just subtract the factors ("39 ppm LARGER than"). Wouldn't you first divide by 1.000039 to get things back to "uncorrected", then multiply by 1.000008? In otherwards, multiply your distances by .999969? Slightly different that what you get doing it your way (.999961).
Sorry just saw this response. I've got to figure out how to get notice to my inbox.
In any case, no I wouldn't first divide by 1.000039, I would begin by dividing by 1.000047 and then multiplying by 1.000008, which I think in the end means that you can simply subtract the factors. Maybe I'm wrong?
Big Al, post: 378106, member: 837 wrote: Sorry just saw this response. I've got to figure out how to get notice to my inbox.
In any case, no I wouldn't first divide by 1.000039, I would begin by dividing by 1.000047 and then multiplying by 1.000008, which I think in the end means that you can simply subtract the factors. Maybe I'm wrong?
Yes. That's what I meant: Divide by the wrong scale factor and multiply by the right scale factor.
Still trying to wrap my brain around it, but i don't see how any number multiplied by (SF2/SF1) can ever be equal to one multiplied by (SF2-SF1) (SF being "scale factors").
rfc, post: 316364, member: 8882 wrote: On a recent outing, I set the temperature and pressure on the gun to what it was in the am (about 45 degrees F), and forgot to increase it as the day progressed (to nearly 80 degrees F).
Wondering if there was a way to "post process" measurements and change them according to what the temperature actually was at the time, I wanted to see what the errors might be, so I set up a quick test at about 100' and sequentially set the temperature in the gun from 35 degrees up to 80 degrees.The distance did not change much at all across that range, and not linearly...it seemed more random and residuals were on the order of less than +/-.001'.
When I checked the ppm, it did indeed change (+ above 68 degrees or so, - below)as it should, but what's the use of setting it frequently if it doesn't make a difference?
What am I missing?
To answer your first question...I use Trimble Survey Controller and Trimble Access (both store data the same way). The raw data file (.job or .dc) stores the distance data UNCORRECTED. That is, it is the raw distance. Also stored in separate records are the temperature and pressure (the S6 has an on-board barometer), and the prism offset (all input by the user). When I process conventional data, I apply the corrections in the office, many times I can input better (i.e. more accurate) atmospheric parameters. Also, for high accuracy work I can input the prism offset for each prism that I determined on a calibration baseline. I often get weather data from a nearby weather station later in the office, and use it. Barometric pressure does not vary much over an area EXCEPT for the well know variation by altitude. Most weather reporting gives sea level pressure, so if you know the elevation at your instrument and target stations, you can correct that value to give a station pressure. Also, I use the reported relative humidity.
To do this you need to know the parameters for your instrument. In the case of the high accuracy S6, the values used are included in the dc file.
In the past, I used a Zeiss S10. This instrument stored corrected values in the on-board data file, along with T and P that was used (in the case of the S10, it had on-board temperature and pressure sensors). Knowing the parameters, I am able to reverse apply the atmospheric correction to get raw values, and then apply "more correct" values if needed.
Moe Shetty, post: 316414, member: 138 wrote: I was recently involved in the measurement/recording of a new EDM calibrated baseline. It was fascinating to see NGS's EDMs. They used two identical EDMs, with accuracy specs that I didn't previously know existed: 0.1mm +/- 1ppm.
Prior to coming out to Maryland, Steve Breidenbach, from NGS, ran a calibration on the Corbin baseline in Fredericksburg VA at the NOAA NGS geomagnetism lab.
Drive out to Hancock MD to the new line, run the measurements. Occupy each of the four markers, measuring each of the other three legs. We used thermistors and barometers and recorded temperature and pressure before and after each block of distance measurements. Mr. Breidenbach briefed us on many of the particulars; one of which was that a one degree celsius temperature difference can cause a one part per million change, and it is an EDM like theirs that can 'notice' that.
He then drove all his gear back out to Fredericksburg to run the calibration routine again. Buckets of work.
It is ironic to me that finding a site to build a baseline can be more difficult than anything else.
We calibrate our instrument every year (at least), I wonder how many others do that? I also measure with all of our prisms. Most are a single type (seco mini), but we have several others as well. Not all match the published offset, usually only varying by 1 mm, but I have seen some that were up to 3 mm different.
I believe it is written into New Jersey law that an instrument must be calibrated once a year.
We now use a new baseline in Ohio that has pedestals rather than ground points...eliminates centering error. NGS has a downloadable program that you prepare a data file for, and it will compute the instrument scale and offset parameters. Very useful information, and these can be used in processing to correct the measured distances.
Temperature differences have relatively little effect on the PPM settings and therefore relatively little effect on distance measurements. A difference of 35 degrees (F) while you're working is enough to be significant, but as was pointed out earlier in the thread, less significant that if you have a tribrach or target rod with a bubble out of adjustment.
Likewise, I've seen field crews far more often introduce far more significant error into their work by mixing target glass, or by simply putting in "0" or "30" a the system offset because that's what's printed on the prism shell rather than testing to determine the actual system constant (the internal offset inherent to the instrument + the prism offset) rather than just entering the reported prism offset.
The adjustment of 1 PPM per 1.8 degrees (F) sounds about right, making a difference of just a smidge less than 0.01' over a 500' distance. Depending upon the scale and purpose of your survey, it may be worth making the corrections as you process the data. If you are doing a lot survey, shooting topo where few if any of your observations are more than 100', or construction staking for most building applications (where high precision is not critical), don't worry about making any corrections.
Barometric pressure and elevation however, will have a fairly significant effect. If weather conditions don't change much throughout the day and if you are working at roughly the same elevation all day, the effect on your PPM setting and effect on distances is of no concern. But if you are working in the mountains, you may change elevations of a few hundred feet or more within a job, and weather can change pretty quickly, so you do need to keep these things in mind and may need to adjust your PPM setting throughout the day.
My modus operandi was pretty simple:
1. Check Thermometer AND Barometer, calculate ÛÏppm correction,Û EVERYTIME you setup the ÛÏgun.Û
2. IF the instrument stayed in the same place for a while, repeat Temp/Pressure/ppm every hour (or sooner if conditions changed).
When traversing see 1 above.
When doing trilateration, set ppm correction to 0 (zero), and measure Temp/pressure at BOTH ends of each vector (and vertical angles if possible), and make the ÛÏmeanÛ ppm corrections unique to each vector afterwards. Our trilateration vectors usually ran from ~5,000 to >25,000 feet, and several thousand feet of vertical was not uncommon.
Control traverses were observed the same as trilateration sets (temp/pressure @ both ends of each leg).
It WILL add up...(ie non-trivial)
Good thermometers (Taylors in metal cases) were pretty inexpensive, good barometers (Thommen) were not. I usually carried 3-4 sets, and they were checked against each other every morning, and every night.
Loyal
It also helps to not have both the Total Station and Data Collector applying the corrections.!
