TSA/Colorado Bend Project

Project Date: May 11-139-12, 2000

Reported by: Dale Barnard

Report Date: 5/232/2000

Person-hours: 165

Personnel: (14 folks)

Angel Stuewe Butch Fralia Charlie Savvas

Dale Barnard Heidi Dues Jack Johnson

James Lopez James Overfelt Keith Heuss

Melonie Alspaugh Paul Atkinson Rosanne Larson

Terry Holsinger Will Harris

May 11-13, 2000 was the last project weekend of the season. The field work will continue the second Saturday in October. We had another small turnout, again most likely due to the weather. On Friday, a big storm blew through briefly, but the spectacular lightning show continued to the south of us for many hours.
Special thanks to Melonie Alspaugh who, for the second time now, stayed late on Sunday to help the park out by leading a Gorman Falls tour. The park has occasionally been short-staffed and asked for help from the cavers.
The best news of the month is that the US Government turned off the GPS satellite selective-availability (scrambling). Now, we are seeing GPS readings within about five meters of accuracy. This beats the 50-100 meters that we have experienced in the past. See Butch Fralia’s extensive report at the end of this report for more information.

Team 1: Will Harris, Jack Johnson, Angel
Time: 4 hours *3 people = 12 hours

This team went to the Red Gate area where they relocated and took GPS locations for several caves and karst features. Most of these locations were already GPSed, but now that Selected Availability (SA) is turned off, they needed to be more precisely located. They noted that SAB251 had bad air 10-12 feet into the cave. They also went into SAB304 to try to enlarge a small crack to a room on the other side. Bad air stopped them.

NAD27 UTM Post-SA
SAB207 XX X XXXXX XXXXX
SAB232 XX X XXXXX XXXXX
SAB233 XX X XXXXX XXXXX
SAB250 XX X XXXXX XXXXX
SAB251 XX X XXXXX XXXXX
SAB252 XX X XXXXX XXXXX
SAB253 XX X XXXXX XXXXX
SAB265 XX X XXXXX XXXXX
SAB266 XX X XXXXX XXXXX
SAB304 XX X XXXXX XXXXX
SABK031 XX X XXXXX XXXXX
SABK032 XX X XXXXX XXXXX
SABK033 XX X XXXXX XXXXX
SABK034 XX X XXXXX XXXXX
The Red Gate XX X XXXXX XXXXX

Team 2: Heidi Dues, Melanie Alspaugh, Charlie Savvas, Paul Atkinson
Time: 5 hours *4 people = 20 hours

This team went to SAB189 with the intent to survey it. With help from Terry Holsinger, they located the cave after walking in circles around some cedar trees a few times. Charlie and Melonie made several trips back to camp so as to bring one item at a time back to the survey project. One trip was for the forgotten survey book. A second trip was for the pencils.
Then, they were ready to dive into the pulsating swarm of Harvestment who excitedly greeted them with jumps onto their hair, backs, and arms. The women enthusiastically saluted the harvestmen with several screams before scrambling down about sixty degrees into the cool limestone entrance.
Paul led the survey by tying ribbons on survey points while Heidi relayed the angles of inclination and orientation to the struggling cave artist, Melonie, who was still at the surface. Charlie, as Melonie’s mentor, instructed Melonie in the fine art of cave drawing.
They slowly proceeded into the next level of the cave where the second welcoming committee awaited. Jumping with joy, the crickets prodded and examined their head lamps and shirts. Once assured that their gear was in order, the crickets moved back to the ceiling to hide behind rocks and flowstone.

Team 3: James Overfelt, James Lopez, Will Harris
Time: 5 hours *3 people = 15 hours

James and James first went to Polish Pit. They descended the first 90-foot drop, but did not have enough for the next 50. They then headed up the road to a recent find, which they called Nylon Stretch. They tagged it and did a quick, crude survey down 4.4 meters to the bottom.
Later, they met up with Will and decided to check out Lone Bat II. They noted nice formations in the cave. Then, it was time to head to the swimming hole. They wrote a simple description of these caves. Another lazy day for James and James came to an end.

Team 4: Dale Barnard, Rosanne Larson
Time: 6 hours * 2 people = 12 hours

This team completed flagging the backpacking trail. It now reaches from the parking area for the Connecting Trail to Lively Pasture near the corner karst. In the fall, with approval, the trail construction can begin.
Along the way, they set a tag on SAB306, which is next to SAB283, and GPSed both caves.

NAD27 UTM Post-SA
SAB306 XX X XXXXX XXXXX
SAB283 XX X XXXXX XXXXX

Team 5: Butch Fralia, Keith Heuss
Time: 6 hours * 2 people = 12 hours

Butch and Keith have spent years gathering cave location data at Colorado Bend State Park. In the beginning, there was overland survey. Referencing from an altitude benchmark on the park with a location assumed from the accuracy of a topographic map a monster survey began. High tech cave survey compasses with an advertised accuracy of +/- 1 degree, cave survey inclinometers with the same accuracy and a two hundred foot tape was the survey system of choice. If not the system of choice, at least it was the system available. Several attempts to gain use of a precision electronic transit system fell through. Plodding along reducing seemingly tons of data and a lot of math, approximate locations for many caves were derived. The problem with this method was that it was SLOW! Caves were being located faster than they could be surveyed in.
And then there was GPS! In the infancy of GPS, Texas A&M loaned the project a pair of survey grade units. Theoretical accuracy was +/- one meter with post processing. It was slow to use because there were fewer satellites aloft in those days. It was often necessary to wait until there were enough satellites above the horizon to get a 3-D lock. And then there was the Aggie survey lab from whence came the units. After a long grueling weekend of caver labor, they tended to lose the data. The small amount of location data returned was in the wrong datum. The locations, plotted on a topographic map were 250 meters from the approximate locations we knew them to be in. In those days, the dark ages of GPS, cavers knew little of datum’s. In fact, few outside the scientific community knew there was such a thing as a datum. The bottom line was that while the future was bright, GPS wasn’t ready for prime time!
And then there was hand held GPS. More satellites were launched bringing the constellation up to twenty-four. This allowed enough satellites to be in view at all times to obtain a 3-D lock. Manufacturers like Magellan took the risk to develop hand held GPS units for commercial applications. The world watched in awe as the price dropped below $2000.00 then finally $1000.00. Once burned on GPS, Butch and Keith wouldn’t be caught again, especially at that price. That was a lot of aluminum beer cans to sell. But finally it came to pass when amid glowing reports of GPS, the price dropped below $200.00. Yes, that actually means the price was $199.99 for the Magellan 2000 four-channel serial receiver unit. Butch bought one and miracle of miracles, it actually worked. If you knew the location of a cave, you could actually navigate to within a couple of hundred feet of it! Keith followed suit, as did many cavers and these low cost Magellan units began to appear on the caving scene.
Caves were located and location data collected. This was where cavers started to learn about geodetic datums. USGS Topographic maps used the NAD27 CONUS datum. A GPS set to display another datum really confused the issue. They learned to ask when collecting location data taken by others; “What datum is your GPS set on?” Until that moment, most cavers didn’t know what a datum was. While lacking in accuracy for precise map making, the early Magellan units were quicker to use than overland survey. Experience proved that if you knew the location of a cave (usually taken with the GPS) you could return to within 75 feet (23 meters) of it. GPS receivers were rated accurate to ten meters internal accuracy. The remaining error was cause by selective availability (SA). SA is an error induced on the satellite transmission by the DOD (Department of Defense) to keep civilians from knowing precisely where they are in the interest of national defense. The error created random inaccuracies of up to 100 meters although 35 meters was more typical. If you owned a GPS, you knew approximately where you were and exactly what time it was. The satellites transmit time signals calibrated with the National Bureau of Standards universal clock at Fort Collins Colorado. In fact, if you had two GPS receivers that was the only thing they could ever agree on. This was a transposition of the old adage that a man with two watches never knew what time it was. A man with two GPS receivers knew what time it was but never where he was.
GPS became better when 12 channel parallel receivers became available. They could process signals up to twelve simultaneous satellite signals at once. A peculiarity of SA is that the transmission error wasn’t coordinated between the satellites. Each satellite transmitted it’s own random error often off-setting the error of another satellite. The more satellites being processed at one time resulted in less error. Time consuming averaging tended to minimize the error but not eliminate it.
Better GPS and an external system called differential (DGPS) GPS offered more precise locations. One weekend, Keith was able to rent a DGPS system accurate to one meter to use with a 12 channel Garmin 12XL. They determined they could repeat measurements within two meters. Unfortunately the DGPS service was provided in the pager bands of an FM radio station located in Austin so it didn’t work in all areas of the park. To further complicate the problem, the provider went bankrupt so the system quit working in any area of the park. Butch purchased a DGPS system from another vendor with similar results. It didn’t work in all areas of the park. To further duplicate the first DGPS experience, the nearest provider in Waco, dropped the service. A post processing GPS system purchased earlier this year worked reasonably well with accuracies of 1-5 meters. This system consisted of a 12 channel Delorme Earthmate GPS receiver coupled to a 3Com Palm Pilot III hand held computer. The data was processed at a later time after being transferred to an Internet capable computer. If there was a problem with a data set, it wasn’t known if the data was usable until long after the fact. And so it went, trial and error with the ever-present dream that one day, SA would be turned off making accurate location data from a hand held GPS without augmentation a reality.
And then it came to pass in recent times that the dream came true. The US Government, namely the DOD, permanently removed SA from its satellite transmissions for a number of economic and political reasons. The child of Al Gore, invented in his spare time between inventing the internet and dealing with affairs of state became a potentially useful tool. At last a dream come true but how accurate is it? A few quick test at home demonstrated an interesting phenomena. When standing still the numbers on the GPS display tend to be still and not change unless the GPS is moved. The estimated positional error displayed on most GPS units typically 60-90 feet suddenly dropped to 10-15 feet. To good to be true! How accurate is it?
All the before mentioned trials and tribulations of locating caves at CBSP resulted in a couple accurately located points on the park. One of these happens to be a corner of the fire pit at caver camp. Unable to wait any longer, Butch rose early Saturday morning, grabbed his Garmin 12XL and a printed location list and hurried to the fire pit. As he neared, GPS in hand, Dale Barnard and another caver came forth to compare readings. Two Garmin 12XL receivers and a Magellan 315 receiver read the location to three meters of the “benchmarked” location! To the further amazement of the cavers, all three receivers read exactly the same. Now, the man with two GPS receivers knows where he’s at as well as what time it is!
So how accurate is it? Probably accurate to +/- five meters for any of the 12 channel civilian receivers available off the shelf today!
With this happy thought in mind, Butch and Keith set out to start relocating all the caves on the park, especially those with previous location data. Like the children from the night before Christmas, they had visions of sugarplums and accurate location maps dancing in their heads. They spent most of the day in Lively pasture until Butch had to return to Fort Worth at 3:00 pm.
They were able to take location data on fourteen caves. Five caves and one other location used both the Garmin hand held units and the post processing unit for later comparison. All the location data is presented later in this report.
During the course of the day, a significant observation was made. Without SA, once a location was recorded, they could return to the same location and be within a few feet of the original location. They almost wasted their time returning to the same cave over and over just to prove they could! Gas Pipe Cave (SAB267) has always been hard to locate. It’s hidden in some bushes and even using GPS, you still had to hunt. Keith and Butch returned to the cave several times during the course of looking for other caves in the area and went precisely to it each time. Having SA turned off is going to be wonderful!
A small Karst feature was found in the location for Cave No 6 (SAB224). The tag for the cave was never located. The tag needs to be replaced at some time. They spent nearly an hour searching for MM Hole (SAB191) and never found it. It’s a small hole, and from Keith’s memory; it’s located somewhere in the area. A lot of the area is covered with cactus, making the search unpleasant. It’s possible they were watching out for their feet more than watching out for the cave.
One of the caves located is Varmint Trap Cave (SAB178). This was the first cave tagged on the park project. It’s also the first cave number assigned in San Saba County that wasn’t identified in the 1972 TSS Publication, “The Caves of San Saba County.”
The dream has almost come true now. The accuracy, while not survey quality, is more than good enough for caver and TPWD work.

Butch Fralia
Travel Time: 6 hours
Work time: 6 hours
Report (post processing time): 3 hours
Total: 15 hours

Keith Heuss
Travel Time: 4 hours
Work Time: 6 hours
Total: 10 hours

All locations shown in datum NAD27 CONUS.
Readings beside hand held GPS w/o SA were taken 5/13/2000.
Readings beside Post processing GPS were taken 5/13/2000.

SAB178 – Varmint Trap Cave

Hand held GPS With SA: XX X XXXXX XXXXX
Hand held GPS w/o SA: XX X XXXXX XXXXX
Post Processing GPS: XX X XXXXX XXXXX

SAB179 – Sore Toe Cave

Hand held GPS With SA: XX X XXXXX XXXXX
Hand held GPS w/o SA: XX X XXXXX XXXXX
Post Processing GPS: XX X XXXXX XXXXX

SAB180 – Skunk Hole

Hand held GPS With SA: XX X XXXXX XXXXX
Hand held GPS w/o SA: XX X XXXXX XXXXX

SAB182 - Horseshoe Chimney

Hand held GPS With SA: XX X XXXXX XXXXX
Hand held GPS w/o SA: XX X XXXXX XXXXX

SAB183 – Gorman Creek Crevice