The Height of Accuracy
Mount Rainier is a cultural icon to modern-day residents of Washington State and a spiritual icon for many of the native tribes of the Pacific Northwest. Though its massive white dome and physical make-up has changed very little over time, surveyors have spent centuries debating the actual height of the mountain and perfecting various tools to create an accurate measurement.
In celebration of Mount Rainier National Park's 100th anniversary in 1999, a group of surveyors from the Land Surveyors Association of Washington (LSAW) set out to measure the mountain using the most accurate surveying method available to date--the Global Positioning System.
First GPS Measurement
One of the first GPS surveys of a summit at Mount Rainier was organized by LSAW. In honor of the state of Washington's centennial in 1988, LSAW set in motion a plan to re-measure the elevation of Mount Rainier using the new technology of GPS.
Today, the initials GPS are commonplace--from the latest James Bond movie to casual day hikers. But in 1988, the LSAW survey was the first that used GPS technology to establish the elevation of a major summit in the world by placing a receiver on the summit. By conducting this survey, LSAW hoped that it would advance the production level technology for determining horizontal positions and elevations, and also raise the publicÃs awareness of this emerging technology. In July 1988, LSAW volunteers from across the state of Washington took part in the project. At the time, there were only seven healthy GPS satellites in orbit and on any given day the window of opportunity for this type of survey was only four hours. Despite these challenges, the project was a success. The summit was measured for the first time using GPS, and the resulting measurement was a summit elevation of 14,411'.1"Ã'1'.1" greater than the accepted 1956 measurement by the USGS. (The details of the expedition were published in the August-September, October-November 1988 issues of POB. For copies, E-mail firstname.lastname@example.org.)
As GPS technology advanced, it was only a matter of time before the challenge of gaining an even more precise measurement became too great to ignore. That opportunity came in 1999 with the celebration of the 100th anniversary of Mount Rainier National Park. Several things had changed since the last measurement. First, the accuracy, method and documentation in the use of GPS had been questioned by those responsible for placing the elevation on maps in the United States. Additionally, there are now more than 24 satellites available, a new Geoid model, improvements in receiver technology; and a high-precision network of control points is in place. The 1988 survey team did not have these advantages.
History of Mount Rainier Surveys
The fascination with measuring Mount Rainier began long before Washington became a state. In the early years, many debated the mountain's actual height and which methods produced the most accurate results.
1842 Triangulation from a baseline 12,330'
1856 Triangulation 14,444'
1888 Barometer 14,524'
1889 Triangulation 14,519'
1897 Barometer 14,528'
1914 Triangulation 14,408'
1956 Triangulation with reciprocal angles 14,410'
1988 GPS 14,411'.1"
Planning and Execution
The plan was to surround the mountain with receivers located on the new high-accuracy control points and tie into previous monuments used in the 1988 GPS re-measurement. During the 1988 survey, the monuments that LSAW used included Nisqually Entrance (Sunshine Campground), Paradise, Stevens, White River and McClure Rock. These points had been measured in horizontally from past geodetic control established by the National Geodetic Survey (NGS) using GPS in the 1988 survey. However, the vertical controls to these base stations were established by LSAW using second-order, differential leveling from NGS vertical benchmarks in order to provide high-accuracy vertical control for summit measurements.
To accomplish the 1999 survey, the first step was to recover the original, previously used base station monuments around the mountain. The monuments at Nisqually Entrance (Sunshine Campground), Paradise, Stevens, McClure Rock and Camp Muir were found in good condition. The monument at White River had been destroyed by road construction. To compensate, a new high-precision point established by the Washington State Department of Transportation was located in the vicinity of the previous White River base station and was used in the 1999 survey.
In order to place the receivers on the summit for the GPS measurement, the team had to first climb the mountain. LSAW planned to use as many of the original 1988 climbers as possible. The overall mountain leader from that survey, Noel Gilbrough, also an employee of the U.S. Army Corps of Engineers, Seattle District, was an experienced climber who selected the climb team and prepared the task's logistics.
Because the GPS equipment available today is much lighter and less bulky than the 1988 equipment, fewer climbers were needed to place the receivers on the summit. Of the LSAW climb team members from the original GPS survey, Gilbrough was able to enlist the services of Norm Brones and Terry Brenneman--both highly experienced climbers and surveyors. Additions to the LSAW climbing team for 1999 were Jerry Retzlaff, Mark Retzlaff, Dave Wolf, Tom MacDonald and Larry Selden. For safety reasons, all team members were thoroughly trained in glacial climbing techniques.
Brones, Brenneman, MacDonald and Wolf comprised Team 1. Gilbrough, and Jerry and Mark Retzlaff comprised Team 2. Selden operated the base station at the 1988 Camp Muir monument at the 10,000' level.
The gear carried by each of the climb team members 10 years ago weighed more than 80 pounds. Technological advancement in GPS equipment reduced the weight for the 1999 survey team by 50 pounds. For the mountain operations, we selected Trimble 4800 receivers (Trimble Navigation Ltd., Sunnyvale, Calif.) for their compact size and built-in antenna systems, which reduced the need to carry separate, bulky antennas. For the base control points, we selected Trimble 4000 series and 4800 receivers.
An additional advantage was battery power. In 1988, Gel-cell batteries--state-of-the-art at the time--were heavy and susceptible to power reductions caused by cold climates. Today, receivers operate on light, one-pound lithium batteries that operate well in cold conditions. We used three 4800 receivers: two for the summit and one for Camp Muir--and provided each receiver with three lithium batteries.
There were two requirements for selecting the control to establish horizontal and vertical for the positioning of the summit. The first was good horizontal and vertical ties to the high-precision network and, secondly, good satellite visibility for data collection. Also, for comparison, the recovered 1988 control would be part of an overall GPS sub-net. (Good vertical control is essential, along with placing control around the mountain, to determine the Geoid slope and accurately calculate elevations.)
Nine high-accuracy points made up the surrounding network. This inside network was also tied to more distant points in Seattle, Spokane, Wenatchee and Port Angeles to establish research into long baseline computations and to determine measurement statistics. All recovered monuments from the 1988 survey finalized the network. All members of the LSAW mountain and ground operations were briefed on the project schedule, operations preparations and procedures. All team members were experienced in GPS data collections and operations. The project schedule called for an August 25, 1999, start date with the summit team members climbing to approximately 7,000' and making camp to help acclimatize the climbers to the lower levels of oxygen at Rainier's higher elevations. Then, on August 26, the climbers would arrive at Camp Muir for the night prior to the assault on the summit.
Early morning, August 27, the team would reach the summit and start collecting data on the ice cap. Data collections were to start at noon and run for four hours. During this window of time, the ground stations and base station at Camp Muir would also be collecting data. We planned to use two receivers at two different measure-up heights on the ice cap. During the GPS session, the surveyors at the summit would attempt to locate two monuments placed and measured on the crater rim in the 1988 survey.
Let the Climb Begin!
On August 25, the team prepared for the first day's climb. All equipment was checked and double-checked. Safety procedures and survey procedures were reviewed. The team started the climb toward Camp 1 above Pebble Creek. On the morning of August 26, Jerry and Mark Retzlaff and Selden proceeded to climb to Camp Muir and rendezvous with the other climb team members. (Scheduling problems had prevented them from starting their climb on the 25th.)
The day before the start of the climb, the weather conditions were very poor. Blowing snow caused lack of visibility, forcing other climbers to turn back. However, as a good omen, the weather suddenly cleared on our start date and remained good for the duration of the project. Because of the previous winter snowfall and cold conditions, the climb route remained in good condition although it was late in the summer. Anyone involved in performing mountain climbing expeditions knows the relief experienced when all things go right.
On August 27, the team climbed to the summit and placed the two receivers on the ice cap. One receiver was attached directly to a lithium battery, which was attached to an ice screw; the other was placed on an insulated pad on the ice cap. Both measure-ups were made. The receivers started collecting data at 11:49 a.m. A check was made with Camp Muir and the ground stations around the mountain--all were up and running. Data was collected for a period of four hours at one-second epochs. During the summit session, base camp was set up and water was made for the night's stay. The team searched for the monuments set in 1988 but never found them. The team also made a tape measurement between the two receivers and the points referenced on the summit.
The morning of August 28, the receivers were up and running at 7:59 a.m. Camp Muir and all ground stations were operating and collecting satellite data. GPS data was collected at the summit until 9:20 at which time the Summit 1 receiver was moved down to Register Rock and placed on a flat spot on top of the rock in order to establish a permanent horizontal and vertical reference on top of Mount Rainier.
After collecting the data at this position, the climb team prepared and initiated the descent to Camp Muir. Selden remained at Camp Muir, collecting GPS data until the summit team arrived. The LSAW project was on schedule, and all was going well. Then, all the climbers continued to Paradise. The summit and Muir receivers were packed up with the utmost care for the trip home. The day ended with the setting of the sun. The mountain and the heavens had treated us well.
Computations and Conversions
Over the next week, the data files from all receivers were downloaded into the computer. The files were placed in GPSurvey, a Trimble software package. Vectors were generated, and the data was analyzed for correct measuring, antenna types and possible sources of error for future processing. Some of the data collected in the park area at ground stations was rejected for final network calculations due to poor satellite visibility problems.
The first computations were made between the outer circle of control stations in order to check out their reliability. The result of these computations produced high confidence in the published values for all the control stations. Although the points were established by different government agencies, the resulting error was no more than 1 to 3 cm of closure. The accuracy of these stations can be attributed to two reasons: All points were connected to the state-wide High Precision Network, and all were established using modern day dual-frequency receivers operated by highly qualified surveyors.
The next set of computations, between the outer ring of control points and the 1988 Mount Rainier measurement control stations, were analyzed. Two things had to be considered in these calculations. First, in 1988, the horizontal control was based on the original NAD 83 coordinates published for the NGS control stationsÃ'none of which used GPS data for calculations. Since that time, the state of Washington has established a High Precision Geodetic Network with published values based on a new adjustment that includes GPS vectors and, most recently, re-observed and re-adjusted for new NAD 83/99 values. The second issue is vertical computations. All of the control points used for the 1988 survey had second-order differential elevations. The elevations were based on the new NAVD 88 vertical datum. This datum is based on gravity models rather than tidal information, making it compatible with GPS measurements, which also use a gravity-centered system.
Values derived from GPS give a mountain's elevation above an imaginary ellipsoid model--a slightly flattened ball that approximates the surface of the earth. The ellipsoid used for the 1988 and 1999 surveys was GRS 80. Elevations prior to GPS were referenced to average sea level, or what we call the Geoid. To get elevations above sea level, surveyors must determine where sea level would be relative to the base of Mount Rainier.
Around the globe, average sea level can vary by hundreds of feet, depending on how strongly the Earth's gravity tugs at the water. In places where the rocks that make up the planet are very dense and heavy, gravity's pull is stronger and water tends to mound up. In spots where the pull is weaker, sea level is lower. This changes the separation between the ellipsoidal model and the Geoid. This is called Geoid undulation.
By measuring the force of gravity at millions of spots across the Earth, scientists have created a computer model called a Geoid, which estimates average sea level anywhere on the planet. That allows the conversion of GPS ellipsoidal elevations to feet above the Geoid, or average sea level, by the use of calculated separations between the two. Having known elevations in your network is a major contributor in obtaining good results and was taken into consideration during the 1988 survey, so differential levels were used for control points. This included a good elevation at McClure Rock on the mountain to model the difference between GPS ellipsoidal elevations and Geoid values. In a sense, with all the points having both elevations, it creates a local model for calculating the unknown height of the summit. For that, the Geoid model assists in the vertical calculation. In 1988, the Geoid model used was developed at Ohio State University. For the 1999 survey, GEOID96, as developed by the NGS, was used. We now have GEOID99, which will be used in future calculations. The second set of computations verified the differential elevations of all base stations to be used for the sub-net control points around the mountain, which included control points used in 1988. Values for stations Sunshine, Paradise and Stevens were all determined and accepted.
The final calculations were performed for Camp Muir, Summit 1, Summit 2 and Register Rock. The differential elevations were held fixed for all of the base stations. Finally, at Camp Muir, Summit 1, Summit 2 and Register Rock, GPS ellipsoid elevations were derived and the correlated orthometric elevations and Geoid separations were applied to produce NAVD 88 values. The sub-network indicated accuracies of 2 cm in the worst case. Two hundred thirty-nine baselines were processed in the two networks.
Elevations taken in 1956 by the USGS, were placed on the National Geodetic Datum of 1929 (NGVD 29). The vertical datum used for the 1988 and 1999 survey was the North American Datum of 1988 (NAVD 88). The NAVD 88 is based on gravity modeling, making it very compatible with GPS. The difference between NGVD 29 and NAVD 88 is over 1 meter.
Finally, we converted the NAVD 88 elevations to NAVD 29 elevations for comparison to previous values. This was done by using an NGS program called VERTCON. This program is constructed of a grid network that has both NGVD 29 and NAVD 88 elevations so that accurate differences can be interpolated for points within the grids. The resulting NAVD 88 and NGVD 29 elevations are shown in below for the 1988 and 1999 surveys.
Mt. Rainier Comparison of Elevations Between 1988 and 1999 NGVD 29
Station 1988 1999
Sunshine 2042.18 2045.31
Paradise 5389.60 HELD
Stevens 2180.62 HELD
Muir 10087.12 10086.9
Summit 1 ICE CAP 14411.1 14411.0
Summit 2 ICE CAP NONE 14410.7
After calculations, the 1999 survey elevations for all previously measured points checked within 2 cm of the 1988 survey. In 1988, almost 100 LSAW and Corps of Engineers volunteers were involved in a first-of-its-kind summit measurement using GPS. In 1999, there were at least 40 LSAW and Corps of Engineers volunteers. Some were from the 1988 survey and some were new, but the knowledge gained working together is something we all will remember.
One closing note: If the 1956 survey had used a mean of all the observations rather than that of a weighted mean, the summit elevation would have been 14,410'.8".
If there is anything to be learned from the long history and fascination of measuring Mount Rainier it is this: Learning new techniques in the surveying and mapping world is very important, along with teamwork and faith in what you are doing.