PHOTOS-See GA FBN, 1998 (Part 1)

I. INTRODUCTION

A. Authority

Project GPS-1273 was conducted complying to project instructions titled "GEORGIA FBN, 1998" dated March 26, 1998 signed by Charles W. Challstrom, Acting Director, National Geodetic Survey.

B. Purpose

In order to meet America's accelerating positioning and navigation needs, the existing coordinate reference system must be continually enhanced to provide the accessibility and high accuracy required for use with the Global Positioning System (GPS). The digital revolution in mapping, charting and surveying requires a National Spatial Reference System (NSRS) consisting of, among other components, a network of monumented points having four-dimensional positions. The Federal Base Network (FBN) fulfills the requirement for this component. NGS is charged with the Federal responsibility for establishment, observation, monitoring, and maintenance of FBN. FBN provides the critical network foundation for an accurate, consistent and reliable NSRS.

NSRS, in turn, provides the common geographic framework for America's spatial data infrastructure. As such, NSRS serves as the basis for mapping, charting, navigation, boundary determination, property delineation, infrastructure development, resource evaluation surveys, and scientific applications, including crustal motion monitoring, modeling of flooding, storm surge, pollution trajectories, and agricultural runoff. A modernized, accurate, consistent, reliable NSRS is of enormous benefit to state, county, tribal, local and Federal authorities, as well as the private sector.

The survey of Tidal station FORT PULASKI was included with the FBN porrtion of the survey.

C. Time Period

Reconnaissance of the Federal Base Network and the tide station was performed by a number of personnel of NGS Section A including Darren Aug, Scott Lokken, Steve Randall, personnel from Georgia Power and Georgia Department of Transportation beginning in 1996 and continued until May 18,1998.

The actual observations began on march 9, 1998 (Julian Day 068) and concluded on May 13, 1998 (Julian Day 133) with the last of the re-observations. Data processing, bluebooking and adjustment were completed on May 13, 1999.
 

II. LOCATION

A. Locality

The project consisted of ninety-four stations located in Georgia, with the exception of CORS sites in NC, SC, FL and AL. A field office was established in Macon GA at the Residence Inn for the duration of the project.

B. Limits

Survey operations were roughly within the boundaries of:

North 34 deg 53 min North Latitude
South 30 deg 40 min North Latitude
East 089 deg 08 min West Longitude
West 094 deg 18 min West Longitude

III. CONDITIONS AFFECTING PROGRESS

Several observing sessions were affected by strong thunderstorm activity causing some missed and short sessions. Four days of reobservations were included at the end of the project to compensate. Otherwise, observations proceeded according to schedule.
 

IV. ORGANIZATION OF PARTY

A. Personnel

The following personnel from NGS were involved in the project's operations:

Joe Lindsay.................... Project Director
Steve Holdorff. ............... Computer
Ralph Harrell.................. Computer
Steve Breidenbach.............. Observer
Eric Duvall.................... Observer
Dennis Lokken.. ............... Observer
Scott Lokken................... Observer
John T. Moore.................. Observer
Kerry Nowlin .................. Observer
Steve Randall.................. Observer
Dave Rigney... ................ Observer
John Ward...... ............... Observer
Darren Aug.. .................. Reconnaissance, Marksetting and Observations

Communications were maintained with William McLemore, Juliana Blackwell and Nancy McCary of the Field Operations Branch.
Communications were maintained with Nancy S. Doyle, GPS Branch, Spatial Reference System Division.
Communications were ongoing with M. Greg Johnson and Don Dryden of Georgia Power Co.
Communications were established with Lamar Stone, Georgia Department of Transportation.
Communications were maintained with receiver operators and state employees associated with the project.
Communications were maintained with property owners and Airport managers.
 

V. FIELD WORK

A. Chronology

See Attachment D, Observation Schedules, for a summary of daily events and see Office Processing for a summary of conditions affecting data collection or quality.

B. Instrumentation

A total of 24 GPS receivers were used to complete the Georgia FBN survey: Fifteen Trimble 4000SSE receivers, five Trimble 4000SSI receivers, three Ashtech Z-XII3 receivers and one LEICA SR950 receiver. A list of the receivers, as well as a list of the agencies that owned and operated them follows:
 

(1) National Geodetic Survey.
(2) Georgia Department of Transportation.
(3) North Carolina Geodetic survey
(4) U.S. Coast Guard.
(5) Georgia Power

The Trimble 4000SSE and 4000SSI receivers measured phase data in units of whole cycles. Trimble microstrip dual-frequency antennae P/N 22020-00 and P/N 14532-00 were used with all Trimble receivers.

Ashtech Z-12 receivers measured phase data in units of whole cycles. Ashtech antennae P/N 700829.3 were used with the Ashtech receivers.

Leica SR9500 receiver measured phase data in units of whole cycles. Leica Dorne Margolin w/Choke Ring antenna P/N AT303 was used with this receiver.

Antenna heights were referenced to top of antenna ground plane and recorded as such on the GPS Observation logs. Antenna constant values for top of ground plane to L1 phase center and L1-L2 phase center offset as determined by Dr. Gerald Mader, Geosciences Research Division, were defined in a Stephen Hilla memorandum titled "New versions of MERSET.AIF and ant_info.001" dated 18 February 1998 (see attachment I). The antenna heights from PAGE4 savit files are entered in the B-file.

Ten meter cables were used with the GPS equipment to connect the antenna to the receiver. Antennae were approximately oriented to true North.

The SECO 2.0 meter fixed height pole was used at all but one antenna set-up during the project. The fixed height poles required no height measurements. The height of the fixed pole was added to the antenna constant to determine the overall height of ground plane above the station mark. The SECO poles can be rotated 360 degrees to determine perpendicularity and adjusted if needed. A 1.8 meter fixed height pole was used at BM K175, GA DOT equipment.

Data was collected at 15 second epoch intervals with an elevation mask of 15 degrees.
 

VI. DEVIATIONS FROM PROJECT INSTRUCTIONS

In the project instructions on page 4 under project network it was stated that a list of stations (Table 1) and observing scheme (Table 3) would be provided. This information was never provided.

Part of this survey project was initially processed with the OMNI software per project instructions. The project was then processed using the PAGE4 software per oral instructions from Juliana Blackwell of the Field Operations Branch. Results generated by OMNI were never completed and are not submitted with this project.

Horizontal and Vertical adjustments were not run as per verbal instruction from Silver Spring HQ. A Free adjustment with accuracies was computed.
 

VII. DATA PROCESSING PERFORMED

A. Software Used

Trimble's datalogging program 4000 was utilized for logging and downloading Trimble data. A list of Trimble firmware versions follows:
 
 

NAV Firmware	Survey Support (Signal Software)	Downloading Software
7.26	3.07	3.11.00 (1)
7.24	3.07	3.11.00 (2)
7.23	3.07	3.11.00 (3)
7.25	3.07	3.11.00 (4)
7.01	3.00	3.11.00 (5)
6.12	2.12	3.11.00 (6)

(1) Receiver 461, 520, 835, 838, 488
(2) Receiver 285, 286, 289, 960, 993, 997, 201, 693
(3) Receiver 513, 571
(4) Receiver 667, 642
(5) Receiver 817
(6) Receiver 033

The raw receiver data files were downloaded to 3.5-inch diskettes using various computers.

The field office transferred the data to personal computers using standard DOS commands. The raw field data was placed in the DATA directory under the project ID and Julian day. NGS programs and OMNI programs were used to reformat, merge and reduce the data.

Program SSE2RNX was used to reformat Trimble raw field data from a binary sequence to an ASCII sequence.
Project observation data was processed using the PAGE4 version 9810.06 software suite.
The B-file was created using CR8BB, Version 3.21 (January, 1990).
The G-file was created through PAGE4 execution.
Programs OBSCHK Version 2.4.2 and COMPGB Version 1.2 were run to compare the project's B-file and G-file.
Program NEWCHKOB Version 1.4.2 was run to validate the project's B-file.
Program OBSDES Version 1.8 was run to compare the project's B-file and D-file.
Descriptions were digitized using DDPROC version 4.43.
Program QQRECORD was used to build the QQ records of the A-file.
The free adjustment was computed using program ADJUST version 4.10.
GEOID96 was used to compute geoid heights.
Corel WordPerfect 7.0 was used for this report and related documents.
The raw field data and processed data were transferred from hard drives to 100-megabyte Zip disks using standard DOS commands.
Static processing was performed during the project.

B. Rejected Data

Three vectors were rejected in the free adjustment.

On JD 098, the data at station FAA TMA A was rejected. Residuals of about 6 cm in DY, and 7 cm in DU are present here. Severe thunderstorm activity on this day probably caused this discrepancy. This leaves station TMAA with only two good occupations and no FBN ties easterly to stations HOME or MULG, but BM ties are adequate. Also on JD 098, the vector to station OCP2 was rejected. Residuals of 3 decimeters in DY were present. On JD 120, the data at station FAA 53A B was rejected. Residuals resulting in 10 cm are present in DU. Again, severe weather activity toward the end of the session was most likely the cause of bad data. This station still has three good occupations with good ties.
 

C. Office Processing

Precise ephemerides determined by the IGS, referenced to the ITRF96, Epoch 1997.0 coordinate system were used in reduction of data collected from March 1, 1998 through the end of project. The project was then processed with the PAGE4 version 9810.06. CORS data used in all sessions contains data from ASHVILLE CORS ARP (ASHV) in Ashville, NC, ATLANTA CORS ARP (ATL1) in Atlanta, GA, CAPE CANAVERAL 1 CORS ARP (CCV1) in Cape Canaveral, FL, CHARLESTON 1 CORS ARP (CHA1) in Charleston, SC, and MILLERS FERRY 1 CORS ARP (MLF1) in Millers Ferry, AL.

PAGE4 Processing

The project data was processed with the menu-driven suite of PAGE4 software. PAGE 4 is designed to run in a batch mode; once input files are created, little user interaction is normally required. The first step is to place RINEX observation files, the IGS precise ephemeris file and a broadcast orbit file in a processing subdirectory. The PAGE4 menu is then entered and the input files are created by selecting various menu items under Input Files, Station Information, and Setup Options for processing. Program default values are read from a file named DEFAULT.TXT, residing in the subdirectory containing the PAGE4 software. The user has the option to accept or modify these values while in the menu system. Once all input files are completed, the run menu can be accessed.

The seven processing steps are Merge all RINEX files (MERGEDB), Do Triple Difference Solution (TDSOLVE), Synchronize Observations (MERGEDB), Choose Baselines (STREE), Choose Reference PRNs, Fix Cycle Slips (REFPRN/EDITDB), Do the Baseline Solution (PAGE4) and Fix Integers (FIXINT/PAGE4). An eighth step in the run menu is Clean Up Output, which will delete unneeded output files that are specified by the user in the DEFAULT.TXT file. This step is executed only after the user has examined and accepted the processing solutions.

The MERGEDB execution creates a combined data base containing GPS observations from the RINEX observation files of all stations to be processed in a session. MERGEDB synchronizes the observations to common epochs using receiver and satellite clock information and reformats the pseudorange and phase observations into a standard binary format to facilitate further processing.

Program TDSOLVE computes an L3 or L1 triple difference solution and updates the Position, Offset, Meteorological data (.POM ) files with improved coordinates. Once the .POM files are updated, an additional MERGEDB execution is performed.

Program STREE selects an independent baseline network by finding the "minimum spanning tree" that connects all stations. The tree is based on the length of baseline and the amount of data available. The program then modifies this tree by examining user specified items such as which stations to use as hubs, which stations are unreliable and must be connected by a spur baseline, and which baselines are to be forced to be part of the spanning tree.

After STREE is executed, program MERGEDB creates individual data bases for each baseline. Program REFPRN chooses the reference satellites to be used in the double difference processing of individual baselines. The reference satellites are selected based on the elevation angle of satellites and availability of satellites. EDITDB is an automatic data editing program. It uses the reference satellite scenarios created by program REFPRN to determine if any cycle slips exist in the data. Edit instructions to fix cycle slips or delete bad data are written to edit command (.EDT) files. Program BDATA is used to update the data bases with these edit instructions before the execution of the PAGE4 program.

Program PAGE4 first executes an ion-free (L3) double difference float solution on the combined data base. Program L3OK is run to update the coordinates in all data base header information (*HD.DAT) files with the best float-solution estimates. Program RUN_INT is then called. RUN_INT fixes integers while working on individual baselines. PAGE4 L3 solutions are run on each individual baseline and L3 ambiguities are written to the corresponding integer (.INT) files. PAGE4 wide-lane (L4) solutions are then run and the L4 ambiguities are written to the same .INT files. Program FIXINT_1 then uses the L3 and L4 ambiguities to determine the L1 and L2 ambiguities. Ambiguities meeting program criteria are then fixed. Additional PAGE4 L3 and L4 solutions are run on the individual baselines. FIXINT_1 is then executed to determine if additional integers can be fixed. The fixed integers are written to the .INT files and a final PAGE4 combined data base solution is executed. Post-fit residual plots and other output files are then examined to determine if the solution is acceptable.

File GAFBNCBN.HA contains the project's bluebooked station descriptions.

Programs COMPGB, OBSDES, OBSCHK and NEWCHKOB were executed to compare the project bluebook files. Output of the program executions were directed to computer files and printed. Explanations for any file inconsistencies are detailed in Part F.

D. Problems Encountered

On JD 092, the data at CTCORS site DNLB was not useable because of a problem trying to download the data from the receiver. Also, the data at station MULG was lost. This data, unfortunately, was nowhere to be found. Station MULG still ended up with three good occupations. The remaining stations on JD 092 were tied from CORS sites ATL1 and CHA1.

On JD 105, the data from station MOOD was also lost. Every effort was made while in the field office to back up all data, and how this happened, I just don't understand. This leaves MOOD with only two good occupations, but there are good BM ties, adjacent stations are tied, and residuals look good here.

On JD 078, BM B 388 was scheduled, but found to be unuseable. Re-observations were later scheduled in this area for BM ties. BM S 185 was found as a replacement.

On JD 110, the Global Navigation file was not available. Also, data at station K175 had a 51 minute gap due to a low battery, from 20:09 to 21:00 UTC.

On JD 114, there is a gap in data at station ABY2 from 04:00 to 14:00 UTC due to power failure.

E. Final Results

A minimally constrained Free Adjustment was computed holding the position of CORS Station ATLANTA CORS ARP held fixed and the elevation of station ZTL A held fixed. NAD83 (EPOCH 1997.0) coordinates and NAVD88 elevation were used. Relative length accuracies were computed.

Results:
Degrees of Freedom       =    990
Variance Sum             = 113191.8
Std. Dev. of unit Weight =     10.69
Variance of unit Weight  =    114.34

Overall, most residuals were in the mm range, with some up to around 3 cm. Most of the larger ones were between CORS stations, or CORS to CTCORS. A few other stations had DU residuals around 2.5 cm. Horizontal comparisons look good.  Vertical comparisons look good except for some of the BM's that were farthest from the constrained vertical station. Perhaps this was the reason for the difference between computed and data base elevations of around 7 cm at stations MOOD, T074, and D234, and 9 cm at station 1410.

Length relative accuracies show 2 vectors below one part in 10 million but the vectors are short, 0.15 KM and 1.8 KM in length.

In summary, the results look very good.
 

F. Checking Program Results

COMPGB ran clean with no errors.

NEWCHKOB resulted in 520 errors. A large group pertain to the new *71* records that were inserted for gps antenna information. Another large group pertain to gps heights being invalid and are related to the CORS sites. A third large group pertain to ellipsoid height order and class. A horizontal adjustment was not run and ellipsoid heights were not calculated. One error shows station name contains an invalid character, but this is how it is contained in the data base.

OBSCHK produced 142 errors. The first 94 are for ellipsoid height order and class which were not computed because a constrained horizontal adjustment was not run. The remaining 48 errors are because the program does not recognize IGS as orbit source.

CHKDESC contains one error. At station W 64, DEKALB county is shown to be in error. This is actually correct but the data base shows it with a space, DE KALB. The DESC program, however, will not accept it with a space.

OBSDES showed 48 errors. These are the stations that were submitted with GA FBN, 1998 (Part1).
 

VIII. STATISTICS
 

A. Stations Occupied
1. Existing NGS Horizontal Stations ................	39
2. Existing Vertical Stations ......................	13
3. Existing NGS Horizontal/Vertical Stations .......	11
4. New GPS Stations ...............................	26
5. CORS Stations ...................................	5
TOTAL.....	94
B. Vectors Observed
1. Non-Trivial (PAGE4) ............................	423
C. Total Number of Sessions and Days for Project
1. Calendar Days.....................................	66
Observing Sessions................................	41

IX. RECOMMENDATIONS AND COMMENTS

This survey project began on a rather unorganized note. As was mis-stated in the project instructions, there was no station list or observing scheme provided. Reconnaissance was still incomplete, and there still seemed to be disagreement as to the number of CBN points that were to be co-observed. This project was a combined FBN/ANA/CBN survey in cooperation with GADOT. If this was not complicated enough, a list of specifications were included in the project instructions that made planning the actual survey nearly impossible. After weeks of trying out different possible observing scenarios, the operation began.

Recommendation: With the widespread use of CORS sites and CTCORS sites used for positioning in a project such as this, and considering the data processing system (PAGES) now in use, it seems that some of the specifications required should be reviewed to ascertain whether they are still necessary. For example, Is there still a need for all adjacent stations to be directly connected in at least one observing session?

I strongly recommend that the Checking Programs be updated. Especially OBSCHK and NEWCHKOB. With a large number of errors present, it is hard to spot any legitimate errors that may be in the midst of them. A new updated version of CR8BB would also be greatly appreciated.

Phase 2 of this project will include the same stations plus CBN and PAC stations that were not included in this part of the project. The entire processing scheme will be no different, with the exception of additional stations.

Recommendation: Maybe there is something I do not understand, but why couldn't all the stations be included the first time, and save a lot of extra work?

A word of caution. There are two Bench Marks in Georgia with the same designation (E 57). The one observed in this project is in Cherokee County, PID is EE0064.

CBN station MBDR was established and co-observed with station R143 and will be processed in Part 2 of this project. R143 is not in a good location and MBDR will be used as a replacement for this FBN station.

Station EASTMAN CBL DODGE 1032 was observed in five different sessions and is a replacement FBN station for R 272, which is no longer suitable for GPS.
 

X. ATTACHMENTS

A. Project Instructions-See GA FBN, 1998 (Part 1)
B. Project Sketch
C. Station Listing/Fixed Control Data
D. Observation Schedules-See GA FBN, 1998 (Part 1)
E. Free Adjustment
F. Checking Program Results
G. Obstruction Diagrams-See GA FBN, 1998 (Part 1)
H. Summary of Recon activity-See GA FBN, 1998 (Part 1)
I. PAGE4 Redundant Occupations Comparison Listing
 

Prepared By,

Stephen F. Holdorff
Computer, Surveying Technician
NGS Section A
 

Reviewed, Submitted,

A. Joe Lindsay
Project Director
NGS Section A