How do you perform a least squares adjustment on GPS data?
I keep getting a 'Degree's of Freedom < 1' error. While the base always sat at the same point everyday, I have several different days of data where we closed/checked in on points shot the day before. Any help would be greatly appreciated.
P.S. I have attached the raw data below if that helps.
Solved! Go to Solution.
A network map would also help. I'm assuming you are using more than 2 receivers?
I don't think your data can be adjusted as it was recorded. Not enough known control points, single point initialization in undefined coordinate system. If you have base station data recorded for 2 hours send it in for an OPUS solution? Then post process RTK data with the OPUS correction for each day. I would use an existing predefined coordinate system UTM83 or State Plane this makes a map check on Topo DRG's or Aerial Photos easy. Then correct the UTM coordinates to the Assumed Project coordinates just by adding the difference to each triplet; this way they are at least based on an established projection . We do 99% of our work in UTM83 coordinate zones, if other projections are needed it is done in the office.
The Autodesk least squares adjustment routine is for Conventional Survey data (Total Station / Theodolite - Distance Meter) Information needed in addition to all traverse data ( Hz,Za,SD,HI,TH) in the Field Book file are Takeoff coordinates Reference Azimuth and Closing coordinates and azimuth.
The only software I know of that performs adjustment on GPS networks is MOVE3 which is available separately and as a module addition to Leica Geo Office. It requires WGS84 coordinates and Local coordinates on at least 3 points.
Raw GPS coordinates are WGS84 Geodetic coordinates in Latitude, Longitude, Ellipsoidal Height.
Local coordinates are usually plane coordinates (UTM or State Plane coordinates) in Easting, Northing, Elevation (Orthometric Height)
In order to get Local coordinates from GPS coordinates projection parameters, ellipsoid and a GEOID model are required to compute plane coordinates from GPS coordinates. GPS receivers have software that computes the transforms for published coordinate systems. (Look at Map Setup, Coordinate Systems in C3D for coordinate system parameters) One point initialization in undefined coordinate systems just causes problems. (unknown convergence, scale factor, grid azimuth, datum elevation.)
Suggested field procedure
Operate in the field in native WGS84 / UTM83 with a local GEOID09 model loaded.
When using local coordinate transforms use a 3D transform with 3 points at least 1 mile apart to establish the elevation of local plane ,grid, and azimuth
Base Station Setup
1. Setup on Tri station with known WGS84 coordinates. Set WGS84 control point throughout project using RTK. Collect Base Station Data and submit data to OPUS; use Opus solution to check position.
2. Unknown Location Setup: (HERE) Collect Base Station Data and submit data to OPUS use Opus solution to correct position of RTK points in post processing.
RTK Rover Setup
1. Use one of the predefined Coordinate Systems with a Local GEOID Model
2.Unknown Coordinate System (Coordinate Transform ) Use a 2 to 5 point 3-D transform with points scattered as far apart as possible in the project.
With Leica GPS 1230GG equipment multiple day different base locations 5-20km apart RTK closures have been under 5cm between referencing different base stations.
To simplify the last, what you need to perform least squares is at least one degree of freedom (redundant observation) for the adjustment package to "adjust". Otherwise, all you are performing is a minimally constrained network and propagating coordinates throughtout your network, whether or not you are setup on a known point or not. Coordinate system has nothing to do with it; simply the vector resolution between receivers.
C3D cannot adjust GPS baselines to the best of my knowledge.
Log into access your profile, ask and answer questions, share ideas and more. Haven't signed up yet? Register