4 replies to this topic

[mikehall02]

Hi All,

I'm new to cartography and am slowly making my way through a GIS certificate program at Northeastern University.
I'm also interested in someday self-publishing trail maps but have a very limited budget.

I'm wondering if its realistic to produce/self-publish maps for sale with a recreational GPS?

I did a fair amount of research and found that most Garmin units are probably 95% accurate to within 10 meters in open conditions and 20 to 40 meters under canopy. I also came across the NMAS standards that say maps at 1:20,000 or larger should be accurate to within 1/30 inch and a handy table that relates this to GPS accuracy.
(not sure why 1:24,000 requires more accuracy than 1:20,000)

NMAS Standards 90% Accuracy
Map Scale 1:x Feet +/- Meters +/-
10000 28 8.38
12000 33 10.06
20000 55 16.76
24000 40 12.19

Do recreational/local interest maps adhere to this standard? Are there legal/ethical considerations when publishing a map like this? Should maps have disclaimers with regards to accuracy?

Thanks,
Mike

[Eric Wolf]

Mike,

The accuracy of your GPS is really unimportant when making maps. The GPS tracks just become a general guide.

Check out OpenStreetMap and go to your closest Mapping Party. You'll learn much more about the mechanics of creating maps from scratch using a GPS than any GIS class will teach you.

[mikehall02]

Could someone elaborate on this a bit? I guess I'm still confused. If your data for a trail or some other feature is inaccurate, doesn't that make your map inaccurate too? More so if it's large scale?

Speak slowly and use little words and hopefully I'll catch on..

[Hans van der Maarel]

Could someone elaborate on this a bit? I guess I'm still confused. If your data for a trail or some other feature is inaccurate, doesn't that make your map inaccurate too? More so if it's large scale?

Speak slowly and use little words and hopefully I'll catch on..

You can make a bad map with good data, but you can't make a good map with bad data. The accuracy and reliability of your map will never be higher than your source data, unless you spend a lot of time checking it. I.e. if you take GPS readings along a trail with a 10 yards uncertainty in your position, the resulting data can deviate as much as 10 yards from the real situation. The same thing goes for your map. If your sampling frequency is too low, say 1 point every 5 minutes, you have no data for whatever happens in between, so you'll have to interpolate. All of that has an effect on the output map.

Now the point is, will you actually notice this? In college, we were told that the human eye can distinguish objects of 0.2 mm in size. On a 1:25k map that amounts to 5 meters. So if you're planning a map on 1:25k any errors smaller than 5 meters are not going to be noticable.

I know I'm overexaggerating these parameters, but that's to prove my point

[Do Minh Phuong]

Approximately, you can calculate map scale from accuracy by multiplying your accuracy (in metre) with 2000. For instance, if your accuracy is 5m, it should be OK to produce 1/10,000 scale map. In reverse, you can check if your GPS accuracy is enough or not by dividing map scale for 2000. Let's say 1/50,000 scale requires 50,000/2000= 25 m of spatial accuracy. If you remember Landsat TM (30 m resolution) is not applicable for 1/50,000 map, but OK for 1/75,000 or 1/100,000?

Garmin GPS's error is ranging from 5-20 m or more. To this, I think it's suitable to produce maps with scales ranging from 1/15,000 - 1/50,000. Usually, if accuracy is 5m, I don't make 1/10,000 scale map, 1/15,000 is a safe solution.

Some thoughts to share with you all.

Do Minh Phuong