Which scale factor to use

> We just got a primary GPS control report from our sub-consultant on a project. I have not done a lot of this type of work and I have a question on what scale factor to use.
>
> 1) This is strictly a 2D project. Just Boundary, no elevations needed. Should I use the combined scale factor or exclude the elevation factor?

Use the Combined Scale Factor. If you have Star*Net, let the software do the reductions to grid and then you generate surface distances by choosing a Project Combined Scale Factor that will best fit the actual extent of your survey. (Surface = Grid / CSF)

> 2) This is the fully constrained adjustment. Notice that there is one point (NYHS) which is affecting the project average quite a bit. This is a CORS Station that is further towards the edge of the zone than the project and pretty far from the project site. Would you include this in determining your project average scale factor to use? Points 111-114 and NYKT are all within 3 miles of each other. NYNB is far away but on a similar line with the zone meridian as NYKT.

I would use only the stations at or immediately surrounding the project in selecting a value for the Project CSF. Usually, I just find the max and min values of points within the project area and take the mean of them. For a small project of under a mile in extent, unless there is quite a bit of change of elevation across it, I usually see a spread of only a few ppm in the CSF between max and min, with the average approximating all project values within less than that.

Elevation factor is a vital component of combined scale factor, which applies to horizontal (not vertical) distances. Elevation factor is affected by, but has no effect upon, elevations.

:good:

It looks like a localized project with little elevation change. Use your four control points to compute mean CSF. If your traverse goes beyond the limits of the four points, you might also factor in the CSF at nearby NYKT but it's not going to make much of a difference.
Regards,
Mike Moran

Fattire,

> 1) This is strictly a 2D project. Just Boundary, no elevations needed. Should I use the combined scale factor or exclude the elevation factor?

You need to use the elevation factor. Grid coordinates are measured on the reference ellipsoid of the projection. Surveyed distances are measured on a plane above that ellipsoid (unless per chance you are in Death Valley). The elevation factor is used to project distances from the surface of the earth to the reference ellipsoid of the projection. A good way to think of it is as two concentric arc with the same internal angle. The arc length will be great for the larger diameter arc. The elevation factor can have the greatest relative affect on your project combined factor over a small area. While your project may be 2D in nature, you need to have an idea of your elevation to determine the proper grid distances. Luckily you have this info in your report.

> 2) This is the fully constrained adjustment. Notice that there is one point (NYHS) which is affecting the project average quite a bit. This is a CORS Station that is further towards the edge of the zone than the project and pretty far from the project site. Would you include this in determining your project average scale factor to use? Points 111-114 and NYKT are all within 3 miles of each other. NYNB is far away but on a similar line with the zone meridian as NYKT.

Choosing the combined factor is not a simple as averaging all of the control. You need to review the location of the control and determine if its location are relative elevation matches your project area as a whole. If you have control point on a high peak while the rest of your project is in a valley, I would remove the point from the average. Again, the elevation factor has the greatest relative affect on the combined factor over small areas. At a minimum, I would remove the three CORS from your project combined factor leaving you with something like 0.99991704.

Just remember, scale about 0,0 and truncate the resulting values so no one confuses them with grid values (for all you nay sayers, I have the perfect example of an F'd up project). Better yet, just work in the grid.

Good luck

> ...Use your four control points to compute mean CSF. ..
Which yields a CSF of right around 0.99991700. I agree that would be a good number to use.

Thanks Mike! Hope Frank is working out well for you guys up there!B-)