PPM - Parts Per Million .. what does it mean?

This is my understanding.....not sure if it's 100% text book accurate:

It's a way to quantify very small values using a ratio instead of a dimension or unit of measure. If you've got a measurement error of 1 ppm then for every million units your measurement could fluctuate to be off by 1 unit. So in 1,000,000 ft you're off by 1 ft. That translates to 0.001 ft in 1,000 ft.

Usually with total station measurements the ppm is not significant for the kind of work I do.

If you have any sort of temperature swings, you need to be aware of the effect of PPM variables. Every adjustment of 1° C causes about 1 PPM error in your measurement. This means that a swing of 20° or so in Celsius (Say from 40° to 75° F) will cause a 0.10 foot error in a mile long observation. Add in PPM correction from atmospheric pressure, and you can start to have some issues over the course of a traverse. At the very least, it throws your relative positioning statements out of whack. If a PPM correction isn't applied at all, then you can really some some strange results.

Just something to consider. On local jobs with short traverses, or for topographic survey, I wouldn't get too worked up about it, as you're well within tolerances to begin with.

That's true....I forgot to add that I do input the temperature correction which applies the ppm automatically on most instruments.

It's been a long time since I made an observation over 5,000 ft in the garden state....

The PPM phrase generally refers to a correction factor that needs to be applied to distance measurements because atmospheric pressure / density will have an effect. The two factors that need to be taken into account are atmospheric pressure and temperature. When these 2 variables are know, a correction factor can be applied.

One other area where this comes into play is in the measurement itself. Most EDM's will have a stated error factor of something like a couple of mm's and some parts per million. These are the accuracy limits of the instrument and they can be plus or minus.

Most measurements made by most surveyors the PPM is insignificant. As to the magnitude of the PPM error, 1 PPM translates to 0.01' in 10,000'.

Thanks everyone for your helpful information. I understand it now:-D

It's a Millimeter in a Kilometer!

Stephen

To clarify (maybe). EDM's are counting full and half waves of a certain spectrum of light. Instruments are generally set to calculate distance based on the number of waves (divided by two, out and back). The length of any spectrum of wave is affected in different ways by differing atmospheric conditions. The instrument is calibrated usually at 1 atmosphere (look it up). Infrared is used extensively because it is not affected as much as other wavelengths by changes in the atmosphere.

So anyway, the instrument is calibrated at 1 standard atmosphere (and then tweaked to or for the wavelength the EDM is using), at a certain distance. The standard error in those conditions is reported such as +-5mm. Then it is determined at further distances, which is the +-5mm ppm or something like that. In this case you always have an uncertainty of +-5mm (at least), then +-5mm more for each certain distance further that you try to measure (convert 1 million mm to meters to get a better idea). Or you can convert all of the units to feet and tenths of a foot to get a better idea in U.S. units. IN other words it's 5:1,000,000; units the same on each side, a statement of precision similar to 1:10,000.

If you are working at extreme temperatures or high altitudes you need to enter this data into the machine in order to keep measuring to the stated specifications because this would be a different environment than the atmosphere the gun is calibrated at.

Stihl,

In surveying and mapping PPM is used to reference systematic and random errors or distortions. On the System 1200 instruments, Leica uses the term "PPM" as the combination of systematic errors based on atmospheric correction and the distortion of grid versus ground measurements for a given projection. The later is zero if the project truly on the ground. By applying the correct "PPM" factors this systematic error is made to approach zero.

All measurements contain some random error which can not be reduced. Manufactures PPM along with a base error to describe these errors. The TSP1200 system has an EDM error of 2 mm +2 ppm when using standard mode. Thus a measurement at 1000 m would result in predicted random error of 2 mm base error plus 2 mm in error based on 2 ppm if all systematic error was removed as part of your stringent measurement techniques.

John

In addition to all the good explanations above, it helps my redneck brain to think of it in relation to traverse closures, kinda like:

1 in 5,000 = 200ppm
1 in 10,000 = 100ppm
1 in 20,000 = 50ppm

Sometimes I test deed closures or old plats in Star*Net by typing in all the dimensions and setting the standard error for the distances (and angles & bearings and such) at the PPM equivalent to the minimum legal closure of the era or the locale. It really helps reveal ambiguities, scriveners' errors, drive-by surveys, etc.

It's a Millimeter in a Kilometer!

not significant for most topographic survey work with a total station. becomes significant for network GPS survey.

Here is a pretty good article from Professional Surveyor on it:

How Things Work: Parts Per Million (PPM)
http://www.profsurv.com/magazine/article.aspx?i=1196

At a very basic level, and not particular to surveying, it is a RATIO. 1 part to a million parts. You might have a ratio of volumes, such as 1 part gin to a million parts tonic. Or a concentration such as 1 milligram per liter, which is nearly equivalent to 1 part per million because 1 kilogram of water has a volume of one liter. Or as often applied in the surveying world, ppm is often a comparison of distances, i.e. 1 foot to a million feet. I suppose it could be applied to angular measurements, but I don't recall seeing reference to that concept in ordinary use.

You won't see it applied to angles, because there is no likely error mechanism that makes the error proportional to the angle.

Ordinarily, the error or uncertainty in an angle measurement doesn't change much, and is not dependent in a known way, with the actual angle. In contrast, a component of the distance error (by most any measurement method) is proportional to the total distance.

Closest thing to PPM for angles is "Strength of Figure."

Actually it can be. A question on that would go like:

You have a gun specification of +- 3" at 2000', what is the angular precision?

So, tan 3" = X/2000 so X is 0.03' Then convert to precision

0.03/2000 = 1/X angular precision of gun is 1:66,700

This is similar to a textbook problem our students get. So the question I have is does the ppm apply to angular measure as well as distance? If not, is there another ppm correction that should be stated for angular measure, and if not, why not? My answer is because the optical ppm would depend too much on unquantifiable things, and is random error rather than systematic error. But I could be wrong.

>You have a gun specification of +- 3" at 2000', what is the angular precision?

The angular precision is 3". The specification does not include the 2000 ft. Angular accuracy does not depend on the distance as a primary effect (although centering error gets into it).

You can compute a distance ppm that gives similar errors to an angular accuracy, but that is only good for small errors and does not scale with the angle. So it is not ppm of the angle.

Strength of figure is a property of the figure, not an angle.

Maybe, but 3" is going to result in different error amount at differing distances. So, in effect there is something similar going on and you can determine the precision at differing distances in the form of 1:X.

> Maybe, but 3" is going to result in different error amount at differing distances. So, in effect there is something similar going on and you can determine the precision at differing distances in the form of 1:X.

I would say the issue is asking what the angular precision is. In that case, it is 3"; angular precision isn't a function of distance. Angular precision does however affect the positional precision as a function of distance, so maybe a word change in the question would make it clearer.