Which Coordinates are more accurate?

[reaper895]

I sure would like to hear folks thoughts on this please have at it.

[+JeremyR]

Technically DegDec (decimal degrees) is the most accurate because you can extend the precision as far as you can accurately measure.

 

For example:

 

53.667133453462, -2.493874353463 are perfectly legitimate coordinates that refer to an area that is considerably less than 1 millimetre 'square'. In fact, the area referred to by those coords is probably microscopic (I haven't worked it out). That is obviously way beyond the possible accuracy of even military-grade GPS, though.

 

Still, though, six-figure DegDec is probably the most accurate measurement your GPSr can take and that is what the GPSr works with internally. All other coordinate systems are simply conversions from decimal degrees.

[+The Jester]

Actually you can get the same precision from decimal degree, decimal minute, or decimal seconds. The decimal place just occurs in different locations.

[+Cardinal Red]

I sure would like to hear folks thoughts on this please have at it.

 

It is an effort in futility to choose a coordinate format to optimize accuracy. If you need pinpoint accuracy, buy a much more expensive professional unit.

 

This topic comes up often enough. Someone can post the math to show accuracy potential. It's not a true picture. Just use the format that suits your needs.

[+JeremyR]

Actually you can get the same precision from decimal degree, decimal minute, or decimal seconds.

Except that decimal seconds and minutes are almost always rounded to 2 or 3 dp's...

[+mrbort]

Actually you can get the same precision from decimal degree, decimal minute, or decimal seconds.

Except that decimal seconds and minutes are almost always rounded to 2 or 3 dp's...

 

That's a rounding question and for most programs, you won't get that -- you'll get decimal rounding to a similar precision to what makes sense.

 

For instance Mapsource does:

 

N36.99903 W109.04519

N36 59.942 W109 02.711

N36 59 56.5 W109 02 42.7

 

All will give you a similar accuracy (down to less than half of the final decimal point). Any more precise isn't within the resolution of a handheld GPS. Go several more decimal places in and you're out of the precision of what makes any sense on a geoid model (since the geoid is more complex than that)....

[+Lil Devil]

Using the examples above, you have potential accuracy as follows:

 

N36.99903 W109.04519 - 3.64 feet N/S - 2.92 feet E/W

N36 59.942 W109 02.711 - 6.07 feet N/S - 4.87 feet E/W

N36 59 56.5 W109 02 42.7 - 10.11 feet N/S - 8.11 feet E/W

 

All of those values are more accurate than consumer grade GPS receivers, which is 30 feet without WAAS and 10 feet with WAAS. So for geocaching, it doesn't matter which format is used.

 

This website uses DDD MM.MMM because that is way most GPS receivers are configured at the factory.

[+Markwell]

Something I wrote back in Feb 2003 for Geodashing:

> Do the number of digits mean more accuracy?

 

Not necessarily. They mean more precision. It's kind of like saying I'm 35.90793977 years old. How much more information do you need than the fact that I'm 35 (.9 means just about 36) years old? It's an old mathematic concept called significant digits.

 

Standard GPS units aren't more accurate than 15-30 feet. If the unit is displaying 4 decimal points instead of five, it means that it is rounding, and could therefore be off by 0.00005° Latitude and 0.00005° Longitude. At my location (N41.6° W088.2°) being off BOTH latitude and longitude by 0.00005° would mean a difference of 6.94 meters, or 22.77 feet - just about within the range of error on the GPS units.

 

In decimal degrees, the fifth decimal place is JUST BARELY in the realm where precision will impact accuracy. The third decimal place in minutes is about the same, riding that fence of necessary precision for accuracy.

 

Bottom line: changing the display units on your GPS will only slightly affect the precision (IIRC, decimal seconds are the most precise), and will in NO WAY affect the accuracy of the unit. Clear as mud?

 

Markwell

 

All of this states exactly what Lil Devil said above. The precision of any of the formats on the display of the unit is more refined than the accuracy of the unit, so changing a setting and inputting a desired coordiante is not going to make the unit more accurate.

[+Renegade Knight]

Actually you can get the same precision from decimal degree, decimal minute, or decimal seconds.

Except that decimal seconds and minutes are almost always rounded to 2 or 3 dp's...

They are all rounded at "close the heck enough".

 

GPS receivers from the era when Selective Availability was an issue did round a bit shy of current rounding. On those for whatever reason they didn't normally round UTM. When they turned UTM off and the rounding was an issue, UTM was a work around becasue it was to the nearest meter which was much closer to the GPSs post SA accuracy.

Edited September 2, 2009 by Renegade Knight

[+PirateKatz]

For plotting points, it doesn't matter if you use D-M-S or decimal degrees.

 

However, if you're interested in any kind of analysis (lengths of lines, areas of polygons, bearings, etc., etc.) you will run into the situation where geographic coordinates are incredibly cumbersome and lead to complex math (ie, determining area in degrees squared). It's very rare to ever see a map that's actually in geographic coordinates (even though lat/long might be displayed) because of the sheer level of distortion brought on by the different X/Y scales, which vary depending on the location.

 

There's an entire branch in geography dedicated to projected coordinate systems for that reason.

[reaper895]

thank you for the responses they have been helpful and educative

[+StarBrand]

It matters little for Geocaching for not only all the reason stated above - but just as a matter of fact - all the coordinates on this site are reported as decimal minutes with 3 decimal places.

[+fizzymagic]

I sure would like to hear folks thoughts on this please have at it.

They all have the same accuracy; some have more precision than others.

 

<pedantic>

The accuracy is how close the coordinates are to the (unknown) true coordinates.

 

The precision is how close the indicated coordinates are to the measured coordinates.

</pedantic>

 

For the case of consumer-grade GPS units, the accuracy is always less than the precision, so the format does not matter.

[+Harry Dolphin]

It is good to see that there are some people who understand the difference between precision and accuaracy!

[+NYPaddleCacher]

Using the examples above, you have potential accuracy as follows:

 

N36.99903 W109.04519 - 3.64 feet N/S - 2.92 feet E/W

N36 59.942 W109 02.711 - 6.07 feet N/S - 4.87 feet E/W

N36 59 56.5 W109 02 42.7 - 10.11 feet N/S - 8.11 feet E/W

 

All of those values are more accurate than consumer grade GPS receivers, which is 30 feet without WAAS and 10 feet with WAAS. So for geocaching, it doesn't matter which format is used.

 

This website uses DDD MM.MMM because that is way most GPS receivers are configured at the factory.

 

As I wrote in a post in another thread, although most GPS receivers and the web site are configured to *display* coordinates in Degrees Decimal Minutes (N36 59.942), the two file formats (GPX and LOC) express lat/long values as Decimal Degrees (N36.99903). I assume that the Groundspeak database stores them as DecDeg but converts them to Degrees Decimal Minutes for the cache page listings. In any case, if you download waypoints to your computer then transfer them to your GPS, there is going to be a coordinate conversion if you view them on your GPS in any other format than Decimal Degrees.

[+Viridios]

I personally use the Universal Transverse Mercator (UTM) format. The majority of maps I use are in UTM, and is accurate down to the meter. I find that is is really easy to pinpoint my location on a map, and for triangulation with a compass if my GPS conks out.

[+sTeamTraen]

It is good to see that there are some people who understand the difference between precision and accuaracy!

Wikipedia has a great definition:

Accuracy is the degree of veracity while precision is the degree of reproducibility.

[+tomfuller & Quill]

Before geocaching was invented, I used a very expensive 3.8 lb. GPS usually with a telescoping rangepole antenna. On the waypoint, I would take 3- 3 minute readings. I would then do some digital magic to Differentially Correct my waypoints. The one that was entered as final was the one with the best PDOPS.

Your standard sport handheld GPSR won't read to more than one decimal point on seconds.

on seconds 1 minute is divided into 600 parts. Decimal minutes to 3 decimal places divides the minute into 1000 parts, so other factors being equal decimal minutes is the way to go.

BTW, last week a USGS technician set up a Trimble GPS with the antenna on a short tripod over a BM near my workplace. It sat there for TWO DAYS (solar powered). I'll bet his horizontal data was within 1 CM and the vertical within 3 CM.

The benchmark was set in 1936 and this was the first return visit by the USGS.

The number and orientation of "seen" satellites has a great deal to with how accurate your GPS is at any particular time. If your GPS is seeing 6 or more satellites and one of the WAAS satellites, you are doing good.

[+Harry Dolphin]

BTW, last week a USGS technician set up a Trimble GPS with the antenna on a short tripod over a BM near my workplace. It sat there for TWO DAYS (solar powered). I'll bet his horizontal data was within 1 CM and the vertical within 3 CM.

The benchmark was set in 1936 and this was the first return visit by the USGS.

 

Guess again. Try a hundredth of an inch or less. A benchmark I found recently lists the coords as: 40 43 41.47800(N) 074 03 00.66732(W)

Edited September 4, 2009 by Harry Dolphin