Point of Beginning

Locating the Bluegrass Via Satellite

May 28, 2002
Surveying aids in unifying a network of control for the Derby City.



U.S. Army Corps of Engineers survey marker.
Survey stakes popped up like wildflowers recently across the softly rolling glades of the city of Louisville in Jefferson County, Kentucky, home of the Kentucky Derby. Famous for thoroughbred racing, basketball and a mix of urban areas and beautiful countryside, Louisville is located along the Ohio River, with bluegrass meadows merging into busy interstates.

Situated in a river valley, the area is surrounded by a low plateau and Jack-and-Jill hills, with a ground elevation that varies between 397 and 905 feet above mean sea level, creating a challenge of multi-faceted proportions for any survey project.

Geoid 99 map used in the LOJIC project.

Spatially Sporadic Control

Prior to 2002, Louisville/Jefferson County had a myriad of existing horizontal and vertical control systems, causing a major headache for surveyors seeking consistency. Monuments were spatially sporadic and varied in currency, datum and accuracy. Control existed from the United States Coast & Geodetic Survey (USC&GS), National Geodetic Survey (NGS), United States Geological Survey (USGS), Kentucky Transportation Cabinet, U.S. Army Corps of Engineers (USACE), the historical city of Louisville, Louisville Metropolitan Sewer District, as well as other local utilities, and numerous public and private projects.

The Louisville/Jefferson Information County Consortium (LOJIC) decided to tackle the problem by creating a two-mile grid of monumented geodetic control across Jefferson County. LOJIC, a partnership of the City of Louisville, Jefferson County, the Property Valuation Administrator, the Louisville Water Company and the Metropolitan Sewer District, was formed in the late 1980s for the purpose of building and maintaining a comprehensive automated GIS for the community. The Consortium is an example of cooperation, communication and coordination among various public agencies and utilities in an attempt to develop a shared GIS to the benefit of not only the participants themselves but the entire community. The group set out to establish a unified network of monumented horizontal and vertical control and reference marks at a consistent datum and level of accuracy that would serve the various needs of the LOJIC GIS partners, as well as the community’s public and private sectors.

LOJIC selected GRW Aerial Surveys Inc., a subsidiary of GRW Engineers Inc. (GRW), a nationally recognized engineering, surveying, GIS and mapping firm of Lexington, Ken., as the prime contractor for the project. The main tasks involved recovering selected existing control monuments, locating and installing new monuments, and surveying the selected existing and newly installed monuments using Global Positioning System (GPS) surveying techniques to provide a uniform countywide Geodetic Control Network. The project area covered approximately 390 square miles and 350,000 land parcels.

Aluminum survey disk.

Field Reconnaissance in the Derby City

GRW and subcontractor ClasSickle Inc., of Louisville, Ken., conducted field reconnaissance across horse country and highways to verify the suitability of incorporating existing control in the primary/secondary control grid. Like a survey survivor team setting out on a mission, each team was equipped with essential supplies, including a LOJIC/Jefferson County street atlas and road map, description sheets for the existing monuments, a map of the tentative locations for new monuments overlaid onto a digital USGS quadrangle map background, and a digital camera to instantly capture and record necessary data.

As existing monuments were recovered and new monument locations were selected, the survey teams staked and marked the sites with survey stakes and flagging. The monuments were marked on a map, photographed and then hand-drawn sketches of the sites were produced. Notations were made of areas where digging could pose a potential threat to underground utilities. The teams recovered existing NGS HARN stations and bench marks in and around Jefferson County to be used primarily as project control.

Design sketch for new monumentation used in the LOJIC project.

Monumentation Aided by Concrete Buggy

Once the preliminary reconnaissance was finished, GRW personnel were dispatched to install the new monuments. Over 100 new survey control monuments were installed and measured. Monument pairs were set in county parks, on church grounds, in road rights of way and medians, across from the county courthouse, on college campus grounds and public school grounds, in river levees, by bridges, and even at the Henry’s Arc Zoo entrance for the purpose of providing easy access to the monuments.

Power tools came in handy during the backbreaking stage of digging the 40-inch deep monument holes. The GRW team used several different tools to ease the strain: a two man auger, a skid loader with an auger attachment, a Dingo with an auger attachment and manual clam shell post hole diggers. The Dingo and the post hole diggers proved to be the most effective tools for digging the holes for the poured-in-place monuments.

In areas where the terrain was agreeable, ready mix concrete was transported to the monument sites using a Concrete Buggy provided by Chaz Concrete Co. LLC of Louisville, Ken. In other areas, it was mixed on location by the company’s Mobile Mix concrete truck, or transported in bags and mixed by hand.

All of Louisville/Jefferson County’s new poured-in-place monuments are equipped with a 30" long piece of #4 rebar to serve as a magnetic locator to aid in future monument recovery. A few of the new monuments consist of an aluminum survey disk set in massive concrete structures. In these instances, a gasoline powered portable generator with a 1⁄2" hammer drill was used to drill a hole into which the disk shaft was epoxied.

Kentucky Department of Highways marker.

Network Design

Once the new and existing monument locations were finalized, existing NGS HARN stations and bench marks were selected as project control. Control was selected in a configuration optimized for both horizontal and vertical support of the GPS network(s) and in accordance with the FGCC guidelines with sufficient redundancy to allow for quality control evaluation and blunder detection.

This project involved two distinct phases. Phase I consisted of existing NGS horizontal and vertical control and was designed to evaluate the existing control, independent of the new and existing monuments to be controlled. Phase II was designed to establish GPS positions and elevations for the new and existing monuments to be controlled.

At least one GPS session baseline was directly observed between each station azimuth pair. This resulted in uniform coordinate quality on all survey stations, an overall stronger GPS network and fewer total GPS observations. Where existing NGS bench marks could not be directly occupied, a temporary eccentric point was set at a nearby inhabitable location. Conventional differential leveling techniques were used to transfer an elevation from the bench mark to the eccentric point.

In one instance (GPS86-43), a suitable location for an azimuth monument (Buddy) could not be found. A GPS occupying a temporary point was set and controlled as a part of the network. It was later used as an azimuth mark in determining a direction to a physical feature azimuth mark.

Data Processing and Quality Control

Quality control is the backbone of any successful project and a vital ingredient in data processing. To ensure accuracy, GRW utilized multiple quality control steps. During GPS field collection, the data was processed on a daily basis as it was collected. After downloading the data from the GPS receivers to the computer, GRW’s quality control technician “checked into” Trimble’s GPSurvey (Trimble, Sunnyvale, Calif.). During this process, the occupation start/stop times, operator entered station name, autonomous position and operator entered antenna height were examined and evaluated.

Two common blunders, mis-keyed station name and/or antenna height, could be detected during this operation by comparing the entered data to the observation schedule and the observation log sheet that the receiver operators completed for each observation.

Once the quality of data was confirmed, GRW differentially corrected it using Trimble’s WAVE baseline processing module, generating four key pieces of information: integer ratios, reference variances, satellite phase tracking summaries and residual plots—all of which provided an indication of the quality of the corrected baselines. Together these indicators give the GPS data processor an indication of the quality of each baseline and flagged potentially troublesome baselines. In some instances a weak solution may have been improved by adjusting the processing parameters and reprocessing the data.

The next step in GRW’s data analysis involved connecting vectors together, head to tail, forming closed loops of independently observed baselines much like a closed traverse. This was performed using Trimble’s loop closure utility. In these well-conditioned networks, defective baselines were detected and isolated. Where sufficient redundancy existed in the networks, defective baselines were simply discarded; otherwise they were re-observed.

After this preliminary data screening, the baselines were imported into Trimble’s network adjustment module TRIMNET. Here, tools for the comparison of redundant baselines and the detection of duplicate stations (misnamed points) were used to further screen the data and detect blunders. Up to this point all coordinate data was expressed either in ECEF (Earth Centered Earth Fixed) or WGS 84 (latitude/longitude). For the next step, the project coordinate system was selected (NAD 83 Ky. S.P.C.S. North Zone HARN).

A single point with known horizontal coordinates and ellipsoidal height was then held fixed and a minimally constrained least squares adjustment was performed. This gave an indication of how well all the baselines fit together within the networks, independent of the rest of the control. Several iterations of the adjustments were performed. After each iteration the results were evaluated and mis-fitting baselines were re-processed, rejected or re-observed. The network weights were also adjusted until acceptable solutions were obtained.

Once acceptable minimally constrained adjustments were obtained, GEOID heights were interpolated (using Trimble’s GEOID module) and incorporated into the networks. Additional horizontal control stations were then constrained and adjusted iteratively. This gave an indication of how well the existing horizontal control fit the GPS data. (Historically, before the existence of the HARN system, the quality of the GPS data was far better than the available existing control. However, when tying a GPS network to HARN monuments, the fit is generally very good.)

As soon as acceptable horizontally constrained networks were achieved, additional vertical control was iteratively added into the adjustments. Again network/control fit was examined and network weights were adjusted until acceptable fully constrained solutions were achieved.

A well-conditioned network with a sufficient number of well-distributed bench marks is critical in the development of good elevations using GPS surveying techniques. The networks employed in this project possessed the necessary qualities to promote accurate usable elevations of a reasonably uniform nature. In some instances they may be superior to those that may have been developed using conventional third-order surveying techniques.

LOJIC 2001 monument.

Benefits of the Project

Upon completion of the project, any site in suburban Jefferson County was within two miles of an accurate survey control monument. This project in essence became the basis of future surveys within Jefferson County, which will be referenced to the control network established for LOJIC by GRW. The new control network will ensure stricter compliance with local subdivision regulations that require plats/plans be tied to monumented survey control and include state plane coordinate references; be used in maintaining elevation references for National Flood Insurance Program (NFIP) community rating system credits; be used as pre-marked control for future aerial photo acquisitions with airborne GPS; and provide the base for consistent survey references for all local surveying, civil engineering and construction projects.

As an added benefit, LOJIC, through its interactive GIS application on its website, has made the new control network databases, descriptions and documentation available to the public from the LOJIC website at www.lojic.org. Users can view the Geodetic Control Map, which uses LOJIC digital ortho imagery in MrSID format as its base and allows users to zoom in by parcel number, address, street intersection or monument name. One or more control monuments may then be selected for display of monument description sheets and related photos.

Pleased with the outcome, users of the network have seen firsthand how the art of surveying reaps a bountiful harvest when applied to areas of great diversity and Kentucky beauty.

For more information about GRW, contact Ed Rinehart, PLS, Survey Manager at 859/223-3999 or via E-mail at erinehart@grwinc.com. GRW Aerial Surveys Inc. is headquartered in Lexington, Ken., and has offices in Kentucky, Ohio, Indiana, Tennessee and Texas. For more information about LOJIC, contact Curt Bynum, LOJIC GIS Coordinator at 502/540-6121 or via E-mail at bynum@lojic.org.