Comparing Terrain Data

[john_stimacH3EQQ]

We have a three different methods for importing terrain data into our CAD model.

1. Google Earth Terrain data with a 20' resolution.

2. USGS LiDAR data import (https://apps.nationalmap.gov/lidar-explorer/#/)

3. Mapzen Terrain Tiles (https://github.com/tilezen/joerd/blob/master/docs/data-sources.md#list-of-sources)

How would we go about comparing these data sources to determine the most accurate? This is all preliminary (pre-sales) work prior to winning contracts, so in most cases survey work has yet to be performed.

[J-Porter]

I believe all those options provide a date the data was collected somewhere.  I'm no GIS expert though.  Another option would be to use Infraworks. 

[AllenJessup]

You can look at the Metadata for each source to determine their expected accuracies. Once you're done the Survey on a couple of projects. See what kind of volume calcs you get if you do a volume Surface between you're surveyed Surface and each imported one.

[GKNIGHT892NE]

Accuracy in the triangulated network is found in the points of the triangles. the closer the points are to each other the closer they are to the contour line that indicates the surface. Don't get all hung up on looking for accuracy in surfaces, all the points have a degree of error, anything derived or calculated from the points would have a degree of error depending on the type of calculation done on the data contained in the point data. It all comes down to it's close enough to get the job done but it's not precise as the decimal places you can calculate. This is where engineering judgement comes into play. Don't be inaccurate, don't be precise, don't be imprecise, be just right.

[brian.strandberg · Accepted solution]

Our survey team compared the accuracy of LIDAR data vs field survey data.  Where there was no vegetation (roads, gravel areas), it was very close to the field survey data.  In areas with grass or worse, heavy trees it was off by close to 4'.  On aggregate over the area of this site there was a difference of 1.12'.

I think the only way you can compare how accurate the data you are getting is would be to compare it vs a field topo survey.  You can find a site you have a survey topo for and compare it vs each of these methods.