Surveys of French Frigate Shoals and Laysan Island, February-March 2001




The dragon moray eel (photo courtesy of Hawaiian Reef

Fish, Planet Ocean Publications)

Of all the spectacular creatures that live in the coral reefs of Hawaii, none is more beautiful or exotic than the dragon moray eel.  Its bright coloration and ferocious jaws make it look fearsome, but it is actually shy and retiring.  Dragon morays can be found in the main Hawaiian Islands, but they are rare.  Around Maui, pairs of dragons are known at Molokini Crater and Pu’u Olai, where they are visited so frequently by divers that they have become tame.  Their primary habitat, where they can be found in the wild, is the Northwest Chain of the Hawaiian Islands.  I have enjoyed many marvelous sights while skin diving in the main Hawaiian Islands, but it was the idea of finding a dragon moray in the wild that led me to the Northwest Chain


The Northwest Chain is a string of 11 small island regions running west-northwesterly from Kauai to Midway Atoll and just beyond, to Kure Atoll.  They are geological ancestors of the main Hawaiian Islands, formed by volcanic activity just as the main islands were, but now eroded away to small pinnacles of lava, emergent sandbars, and coral atolls. 



National Geographical Society map of the Northwest Chain of the Hawaiian Islands

It is possible to visit Midway Atoll as a tourist, thanks to a concession granted by the US Fish and Wildlife Service (FWS) to the Phoenix Corporation.  Otherwise, however, the islands are off limits to visitors.  They are all wildlife refuges with extensive populations of endangered species, and access is strictly controlled by FWS (with the exception of Kure Atoll, which is administered by the State of Hawaii Department of Lands and Natural Resources).  All of the islands, with the exception of Midway Atoll, are officially part of the State of Hawaii.


My investigations led to the realization that the only people who regularly visit the Northwest Chain are taking part in scientific research expeditions.  So, for a period of three years or so, whenever I heard of such an expedition being scheduled for this part of the world, I would enquire with the scientific director to see if a geodesist (me) could be useful for the upcoming expedition.  Invariably I received polite replies in the negative, but I continued my enquiries, and eventually I enquired directly with the FWS.



The USFWS GIS Project in the Northwest Chain


In the Spring of 2001 Rod Low was senior GIS analyst with the FWS Remote Pacific Islands office in Honolulu, where his job was to prepare maps and other associated GIS data for the Northwest Chain.  (He has since moved on to the ESRI office in Honolulu.)  One of the problems that he experienced in creating these maps was that he didn’t know where the islands were.  He had Landsat and IKONOS satellite imagery for some of the islands, as well as old aerial photographs and some old CAD plots prepared in years past by FWS personnel on the islands.  He also had nautical charts.  But the charts were referenced to WWII-vintage astronomical datums created specifically for each island, and in some places the reported errors in these charts were on the order of magnitude of miles.  The published positions of some of these islands are so inacurrate that they can’t be used for navigation.  Rod understood that for his GIS work it would be necessary to bring precise, modern geodetic control to the Northwest Chain, so that the imagery could be properly registered, the astronomical datums converted to a modern geodetic datum, future surveys enabled, and the nautical charts corrected.  His next problem was that he didn’t have the funding to make this happen.


The University of Hawaii Pacific GPS Facility


If you want to bring precise geodetic control to remote Pacific islands, a logical place to look is to the University of Hawaii Pacific GPS Facility (PGF), whose director is Mike Bevis.  UHGPS specializes in long GPS baselines and CORS stations that are used as part of the facility’s ongoing study of crustal motion in the Pacific.  Using GAMIT baseline processing in multibaseline mode, and a stacking procedure for network adjustment that solves for velocity as well as coordinate parameters from one epoch to the next, Bevis and his cadre of adventurous, highly skilled GPS experts can achieve accuracies of 1 cm horizontally from baselines 2000 km in length.


Mike Bevis was sympathetic to Rod Low’s requirements, which to some extent converged with PGF’s own GPS projects.  As part of their crustal motion study, the facility constructs permanent GPS antenna poles on Pacific Islands for observing their long baselines.  The facility had no such poles in the Northwest Chain, and they were interested in constructing some of them there.  If this could be done, precise geodetic coordinates could be transferred to the Northwest Chain, addressing Rod Low’s requirements, and the same observations could be used in PGF’s crustal motion study.  In addition, the permanent antenna mounts could be used for local surveys in the islands, including RTK for LIDAR and photogrammetry. 


If FWS funding had been available for the GPS surveys, PGF probably could have been engaged to do everything the FWS required, but without funding there was a limit to what the facility could contribute.  They were prepared to make a substantial contribution, however, not just out of good will but also to help to upgrade the state’s geodetic infrastructure.  The facility offered to provide at cost the materials and tools for constructing the permanent poles, to provide training in the construction of the poles, to provide geodetic GPS receivers on loan, and to process the baselines resulting from a GPS survey.


Mike Bevis told me with a grin that Rod Low was extremely skilled at persuading people to do things for the FWS for free.  What Rod needed was someone who would actually visit the islands, construct the poles, make the GPS observations, and assist in transforming the astronomical datums, all for free.  He sent me a positive reply to my email enquiry.


Tern Island


And so on March 3, 2001, I found myself aboard a twin-engined Piper airplane on a flight from Honolulu to Tern Island in French Frigate Shoals.  We flew between Kauai and Niihau before heading out over the open ocean for the 3-hour flight.  The copilot was equipped with a motorcycle helmet with plexiglass visor, in case a bird should crash through the windshield during the landing on Tern.  The only other passenger on the plane besides myself was Tony Palermo, the Acting Refuge Manager for French Frigate Shoals, who was returning to Tern after a 5-week leave in Honolulu. 


I was acting as a volunteer for FWS.  I paid for my own airfare between my home in California and Honolulu, as well as for my room, board and transportation while I was in Honolulu.  FWS paid for all of my expenses during my trip to French Frigate Shoals and Laysan Island.   Before taking the flight I had spent the previous week in Honolulu consulting with Rod and Mike.


The two seats between Tony and me and the cockpit had been removed and were replaced by my pile of surveying gear.  My company (at that time I worked for Trimble Navigation) had loaned the expedition a 4700 receiver with L1/L2 microcentered antenna with ground plane, an automatic level, and a level rod.  PGF provided one of their Z-12 Ashtech receivers with a Dome Margolin antenna, an 80-lb. battery pack to power the receiver’s long baseline observations, the aluminum antenna pole with associated hardware, an electric drill and other tools for constructing the pole, and two Wild tripods with tribrachs and adaptors.   My personal gear consisted of a briefcase with laptop computer, a carry-on canvas bag, and a full-sized suitcase.  Inside the suitcase were my snorkeling gear and my favorite fish ID book.



My survey gear, minus the aluminum antenna poles

I also had with me a thick sheaf of documents provided by Dave Doyle of the National Geodetic Survey (NGS) recording a 1961 astronomical survey of the Northwest Chain by the US Coast & Geodetic Survey aboard the naval ship USS Duval County.  The public NGS database has no information about control monuments in the islands, but Dave had in his file cabinet the monument recovery notes, sketches of the control monuments, and unpublished astronomical coordinates.  He had been trying unsuccessfully for years to have this data inserted into the NGS database, and he was pleased that something was finally being done with it.  He provided me with a copy of all of the documentation.


I also had with me Tern Island tide gauge data provided by Jerard “Ziggy” Jardin of the University of Hawaii’s Sea Level Center.  A tide staff and tide gauge on the island have been maintained by the Center since 1974, and transferring a mean sea level value from the gauges to any astronomical USC&GS monuments I could recover would make possible an initial estimate of geoid separation and slope on Tern Island.  I also had a sheaf of satellite and photographic imagery of all of the islands I would visit, provided by Rod Low.


I would have to move fast while I was on Tern.  My initial schedule of five days at French Frigate Shoals had been reduced to three because of a delay in our flight.  During these three days I would have to construct the antenna pole, observe both it and any USC&GS monuments with GPS, make fast-static imagery registration GPS observations on Tern and its outlying islands, and carry sea level elevations to the GPS monuments.


When our plane reached Tern Island, about noon, we made one pass over the runway to scope out the danger of striking any of the cloud of birds in flight over the island.  The birds were as numerous as bees around a beehive.  But we landed without incident.


I wasn’t able to begin surveying until about 2 p.m. (both French Frigate Shoals and Laysan Island operate on Honolulu time, although technically Laysan is in the next time zone west).  During those two hours I walked the runway to inspect my new surroundings.  The birds stay off the runway for the most part, but the air and all of the land surrounding the runway were filled with them.  They do not fear humans, and paid no attention to me as I walked past.  The noise they made was high-decibel and incessant, although I soon got used to it.  There were little white terns, plovers and noddies, red-tailed tropic birds, Laysan and black-footed albatrosses, masked and red-footed boobies, and the impressive black frigate birds with their scarlet throat sacks.  The frigate birds are known as Ewa (“thief”) in the Hawaiian language because of their practice of stealing food from other birds, and they have been  known to strike unsuspecting humans in the head, sometimes drawing blood, if the humans should venture too near their nesting sites.  Motionless gray albatross chicks covered the ground in little depressions.  The adult albratrosses were dancing and clacking their beaks everywhere.  The noise these birds generate is eery and disorienting, and often sounds almost human.  Sometimes they sound like a crowd in a baseball stadium, with handclapping right behind your back, but it’s only the bill-clacking of the albratrosses.  One of the bio-tech volunteers on Tern told me that more than once she had turned around to see who was calling her name.  All along the south beach endangered monk seals had hauled out on the sand for a sound sleep.  The clouds, towering high over the sea and the exposed little island, dominated the landscape.  The northeasterly trade winds blew constantly across the island, keeping the tropical heat of the island comfortable.


After the plane’s departure and a slapdash lunch Tony and I toured the island as part of my orientation.  Rod and Mike had provisionally agreed on using the roof of the (now empty) generator building as a location for the permanent antenna pole, and so Tony and I took a close look at it.  (The generator building dates back to the 1960’s when Tern Island was a Coast Guard LORAN station.  The Refuge now uses a solar-powered electrical system.)  From the interior of the building we could see that the roof was a thick concrete slab, suitable for drilling in the bolts for the monument’s base plate.  We leaned a 20-foot ladder against the building, climbed to the top, and Tony approved the site for building the pole.  I went right to work, hauling up the ladder all the gear I would need, including the goshdarn battery pack.  I wanted the longest possible occupations with the Ashtech to allow for accurate computation of the long baselines to it from remote other islands.


Once I went to work, however, things started to go awry.  Everywhere I tried to drill holes in the roof, I found myself in layers of tarpaper so deep that the drill bit didn’t even reach the concrete slab.  The only way to construct the antenna pole on the roof would have been to hack out a huge hole in the tarpaper, and even then there was no guarantee that the 1-meter pole would project high enough above the roof to make GPS observations.


So I climbed down to find another site.  I quickly found the ideal spot, a huge, very hard concrete slab over an old disused rainwater catchment cistern.  It was at least 3 feet high, 25 feet long, and 10 feet wide, and was situated in an easily accessible spot between the barracks building, where all the residents live, and the generator building.   There is a clear view of the sky from this slab with the exception of an old 2.5” diameter flagpole about 30 feet away to the south, which might cause a minor multipath signal.  Otherwise, the 15-degree elevation mask on GPS antennas clears all obstacles.  I found Tony in the barracks, and he inspected and approved the cistern slab as a site for the antenna pole. 


The afternoon sun was starting to drop towards the horizon, and I realized that I might not be able to complete building the antenna pole before nightfall.  If I waited until the pole was constructed before making GPS observations, I might not be able to make them until the following day.  I decided to pound a PK nail into the southeast corner of the cistern slab and observe it with the Ashtech receiver while I built the pole.  In this way, even if I couldn’t complete the pole before nightfall I could still begin making GPS observations, and I could transfer precise geodetic coordinates from the nail to the pole after the pole was built.  I almost killed myself trying to pound a PK nail into the extremely hard concrete slab, and all I had to show for my trouble was 3 PK nails bent into pretzel shapes.  I finally came to my senses, climbed back to the generator building roof and back down again with the drill, and drilled a small start hole in the cistern slab.  I filled the hole with silicone and pounded a PK nail down into it flush with the slab’s surface.


I next pulled out a tripod, tribrach, and adaptor for mounting the Ashtech antenna over the PK nail.  My intention was to rotate the tribrach inside pencil marks on the tripod plate as a rough check on the tribrach’s optimcal plummet and fisheye level bubble.  But when I went to look into the tribrach’s optical plummet, it wasn’t there.  I stared at this tribrach in disbelief.  I knew in an abstract way that some tribrachs were indeed manufactured without optical plummets, but I had never used one myself, and now that was all I had.  Both of the PGF tribrachs were plummet-less.


Then I realized, far too late to do me any good, that Eric Kendrick of PGF had probably tried to warn me about this situation, but I had misunderstood him.  While I had been visiting the lab, Eric had produced for me an optical collimator, of the type used to adjust tribrach level bubbles and to calibrate optical plummets.  I had thought at the time that this was the purpose for which Eric had brought out the collimator, but now I realized that he had probably brought it out so that I could simply center the tribrach over a survey mark.  I should have examined the tribrachs more closely at UH.  I should have brought the collimator with me.  Worse yet, I had violated the Surveyor’s Cardinal Rule: always wear your plumb bob!  (However, even if I had brought my plumb bob it still would have been difficult to use it to center the tribrachs, because the tripods lacked plumb bob attachments for their fastening knobs.)


I had to sit down and think.  I was enveloped in a cloud of gloom and despair.  The birds flew and danced and ululated all around me.  Quite possibly, I thought, my entire surveying mission was an abject failure, fouled up beyond repair.  It was getting late, and I still had nothing to show for my efforts.


I finally determined two methods for centering the tribrachs without optical plummets, both of which I eventually employed.  The first was to simply place the tribrach down directly over the survey mark, whenever the tribrach could be seated firmly and securely.  This would allow me to position the adaptor’s bolt within a couple of millimeters of the true center of the tribrach.  The second method involved the use of the automatic level: I could use the vertical crosshair in the level to plumb the adaptor bolt over the survey mark, but I would need at least two intersecting lines of sight to position the bolt horizontally, and so this method involved a tedious series of multiple setups as well as a cooperative and fast-learning assistant.


I couldn’t use the first method on the PK nail, because it would have placed the antenna so low that a pump housing on the cistern slab would have obstructed the horizon, and moreover the cistern slab surface was so slick that it didn’t afford a secure seat for the tribrach.  So I walked briskly to the barracks to find an assistant to help me employ the second method.  I felt like a complete moron, and I could only hope that my 4 fellow residents on the island wouldn’t notice my disorganization.. 


But there was no help to be found.  The barracks were deserted, with everyone either behind closed doors or in the field.  Dinner would be prepared at sunset, and sunset was now close at hand.


I fell back on my last resort.  I returned to the cistern slab and set up a tripod and the Ashtech antenna on a random, unmarked, imaginary point.  I seated the tripod legs as firmly as possible on the slab and marked the concrete where the legs were set, so that I could tell if they moved.  I tightened all the screws on the tripod legs.  The generator building shielded this site from the brisk trade winds, so the wind probably wouldn’t blow over the tripod, and I could only hope that a red-footed booby or some other bird wouldn’t crash into it.  I turned on the Ashtech receiver and began my GPS observations.  I called this setup CP Innocence, because it could never be recovered, and I called the PK nail CP Useless, because in fact it never got used.  My only solace was that at least I was doing GPS, which was the whole point of my being where I was.  In the few minutes before dinner I made repeated trips to the roof of the generator building to bring back down all of the gear I had hauled up there, including the goshdarn battery pack.



Observing GPS on CP Innocence, with the barracks building in the background

The poor beginnings of my survey had put me in a foul mood, but this quickly cleared away when I joined the genial company at the dinner table.  The living is easy on Tern Island.  Besides Tony and myself, the three other residents were bio-tech volunteers Allison Veit, Laura Kennedy, and Sarah Ward.    Laura had prepared a tasty vegetarian pasta meal.  Tony and I, having just arrived that day, were scheduled to wash dishes.  The dinner table was set with candles, and as the sky darkened outside, the candlelight inside the dining room played on smiling faces and stories of wildlife in the Pacific, Alaska and various other refuge sites.  Allison had been a volunteer off and on for six years on Tern, and she could do everything from driving the tractor to running the boat hoist.  Laura and Sarah were so young that they could have called me “Pops” and gotten away with it, but they were too polite to do this.  They were recent arrivals on Tern, and there were two things Laura especially wanted to do while she was there: to see a moray eel (any variety) while snorkeling and to see the sunset’s green flash.


The next morning I started to work early, and things began to improve.  CP Innocence was undisturbed.  I used the 4700 to observe 4 image registration points at the four corners of the seawall that bracketed the island’s runway, and I constructed the permanent GPS pole.  After building the pole, I mounted the 4700’s antenna on it and began transferring coordinates from CP Innocence to the pole, which I named PGF Pole Tern 1.  My plan was to observe three static GPS sessions of at least an hour each in this configuration: one at the end of the afternoon, one an hour before I retired for the night, and a third one before breakfast the following morning.  The next day I would swap antennas, placing the Ashtech antenna on Tern 1 and the 4700’s antenna on CP Innocence, and I would then leave the Ashtech on Tern 1 for the duration of my stay at French Frigate Shoals.



The 4700’s first session on PGF Pole Tern 1

I finally mounted the 4700’s antenna on Tern 1 just before six, and began the first session.  Sunset arrived at seven, and if I moved fast I could squeeze in an hour of snorkeling before nightfall.  I did move fast, and so did Laura when I came running into the barracks to grab my snorkeling gear.  Sarah was in the field, and missed this opportunity.  Laura and I took two of the Refuge’s bicycles and rode into the face of the tradewinds down the length of the runway to the eastern end of the island, where a nice little beach (known locally as East Beach) has accreted to the northeast corner of the sea wall. 


Swimming north and east from the beach, Laura and I crossed a sandy passageway about thirty feet deep to the north fringing reef, then followed the reef  to the east, then down to the south to a maze of coral mesas, arches, and caves in about twenty feet of water, and then back west to East Beach.  The coral around Tern isn’t as various or colorful as it is in the more southerly latitudes of the main Hawaiian Islands, but it is lush and plentiful.  The water was crystal clear, and even in the late sunlight the underwater world seemed bright.  Along the way we were checked out by a monk seal who swam parallel with us, then down and under us and then away; a green sea turtle; schools of enormous unicorn fish; crowds of butterfly fish and tangs; a shy rockmover wrasse; two of the largest yellow-tailed wrasses I have ever seen; and in general a multitudinous, widely varied swarm of the usual suspects.  I didn’t see a dragon moray, bit I saw plenty of reef life.  About halfway out, I felt Laura grabbing frantically tugging at my arm.  She had spotted a white-mouth moray!


We exited the sea in the day’s last light and cycled back to the barracks, this time coasting along with the tradewinds at our backs, as the orange disk of the sun sank down on the clear horizon over the west end of the runway.  Laysan albatrosses soared past our ears no more than three feet from our heads.  As we approached the barracks the sun


IKONOS near-infrared image of Tern Island, with East Beach at upper right (northeast)

finally sank below the horizon, leaving behind a huge, brilliant, two-second green flash.  Laura was joyous.


My own mood was greatly enhanced by the suceess of my GPS observations.  By dinner time I had already observed more than an hour of static GPS with the 4700 on the new pole, and I had accumulated more than 24 hours observing CP Innocence with the Ashtech.  Even if the tripod on CP Innocence were to be knocked over, I should nevertheless be able to transfer precise coordinates from this unmarked point to the permanent antenna pole.  My strategy with CP Innocence had been vindicated, and by breakfast the following morning I would have even better observations on these two setups.


Back at the barracks, just before dinner, I turned off the 4700.  Later that evening, an hour before retiring, I took a flashlight and, stepping gingerly through the crowd of gray albatross chicks roosting in their little depressions in the ground everywhere, and trying to avoid stepping in noddy burrows, I returned to the cistern slab and turned on the 4700 receiver for another 1-hour session.  There was a bright three-quarter moon, and I took the opportunity to watch the birds swarming and screaming in the sky, the Laysan albatrosses continuing their odd dances and bill-clacking, and the bold, curious little white terns hovering just over my head.


The following day, March 5, was a sunny, clear day for my trip to East Island, Little Gin Island, and Trig Island for image registration.  Before breakfast I fired up the 4700 for its last session on Tern 1.  After breakfast we had to launch the boat as soon as possible, so I didn’t have time to complete the antenna swap, but later that evening I would complete the antenna swap so that the Ashtech occupied Tern 1 and the 4700 occupied CP Innocence.  Tony gunned the red whaler away from the small dock for our trip around French Frigate Shoals.  We carried 4700 with us and left the Ashtech receiver on CP Innocence for our base station.




The NOAA nautical chart for French Frigate Shoals

Our first destination was East Island, about 8 miles away.  East Island is now entirely unoccupied, except for one month a year when researchers for the National Marine Fisheries Service set up camp to tag green sea turtles.  In 1963 the Coast Guard used the island as a temporary location for a LORAN antenna until construction work on Tern Island was complete.  All that is left on the island from the LORAN camp is one large pole and some mysterious iron fittings.  The remainder of the island is occupied by turtles, seals, and the usual multitude of birds.  There isn’t a tree anywhere.



Approaching East Island in the red whaler

We landed on East Island near the pole.  Tony anchored the boat just off the beach while I jumped off the bow and waded ashore carrying the 4700.  Referring to the IKONOS image of East Island provided by Rod Low, I selected two image registration points.  One of them was in the middle of the island in line with the pole.  The shadow of the pole is visible in the satellite imagery.  I paced off the distances from the pole to the high water marks on both sides of the island, and recorded a sketch of the position in my field notes.


While I was hauling gear up the beach I got slammed in the head out of nowhere.  I looked around to see a big black frigate bird soaring away.  I touched my forehead and found a small smear of blood.  Looking up the beach, I saw that I had parked my gear next to a bush that had a frigate bird’s nest in it.  I moved my gear.


My next registration spot was at the tip of vegetation on the northwest end of East Island.


Our beach landing on Little Gin Island was more challenging, and was in fact the most exciting part of my trip to the outer islands of French Frigate Shoals.  Heavy swells were hitting the beach on both sides, and at first it looked like we wouldn’t be able to land.  But some careful maneuvering by Tony, through the reef on the crest of the swells, got us close to the beach near the southwest tip of the island, and I jumped off the bow into the water.  The swells were so large that the troughs of the swells were at my waist, but the crests of the swells were over my head.  I held my breath and raised my arms and, while underwater, I felt the receiver get dumped into my hands.  I managed to walk up the slope of the beach, holding my breath, until I was on dry land.    I found the usual crowd of birds on the crest of the tiny island, but nothing else except sand.


The sea conditions looked marginal.  A squall was moving in from the northeast.  I decided to observe my registration point as close to the boat as possible, on the south tip of the island, so that if necessary I could race all the gear back to the boat before sea conditions forced the red whaler to leave its anchorage.


My last image registration point on the outer islands was on Trig Island, only a few miles to the east of Tern.  Our passage there from Little Gin was rough.  The squall hit the Shoals, and the trade winds got cold.  The seas were rough, and both Tony and I were drenched.  The red whaler smashed through the waves with bone-jarring impacts.  But when we finally reached Trig the sun came out, and the sea conditions on the north shore of Trig were benign.  Anchoring the red whaler off the beach was much easier than our Little Gin anchorage, and we only had to wade thigh-deep to haul ourselves and the gear from the boat up on to the island.  We took off our wet shirts and basked in the warmth of the sun during this fast static observation.



The Ashtech is moved to its last occupation, PGF Pole Tern 1

Tony and I arrived back at Tern Island with an hour and a half to spare before sunset, allowing plenty of time for me to complete the antenna swap.  I placed the 4700 on CP Innocence and moved the Ashtech to Tern 1, with the intention of observing three more sessions between the two monuments during the night, and then I collected Laura and Sarah for another snorkel expedition.  We returned to East Beach, but this time the current was so strong that we could only venture a short way from our entry point.  Sarah didn’t get a chance to see the reef life that Laura and I had seen the previous night, but Laura knew where to return when better conditions prevailed.


The following day, March 6, was the day the Townsend Cromwell, a 165-foot NOAA research vessel, would carry me away from French Frigate Shoals to Laysan Island.  There was still more work for me to do in the time I had left.  The first task was to recover all the old USC&GS monuments I could find dating from the 1961 astronomical survey.  The documents that Dave Doyle gave me described a total of four of them.  The FWS personnel on Tern could not recall seeing any of them.  Two of them (FRIG 1961 and AZIMUTH MARK 1961) were reported destroyed as early as 1963.  I could find no trace of AZIMUTH MARK 1963 on the SE end of the island.  But I did find FRIG RM2 set in the top of a concrete pillar about 4 feet high, situated about a quarter mile east of the barracks and south of the runway.  The pillar showed cracks in its sides and missing chunks of concrete, but it was rock solid.  All of the FWS personnel on Tern were familiar with the pillar, but none of them had looked on top of the pillar to find the guano-encrusted USC&GS brass disk set there.


I had wanted to find at least three USC&GS monuments, for the best possible transformation of the astronomical datum, but this one (RM2) would have to do.  I lovingly wiped the guano off the disk and admired the familiar stamping on it.  I tried to place the tripod over the pillar, but the legs were too short for a setup, so I employed my first method of placing the tribrach directly down on the station mark, where a T3 astronomical theodolite had originally been mounted.



Red-footed boobies nesting between FRIG RM 2 1961 and the old fuel shed

While the 4700 observed the brass disk I enlisted Sarah to help me carry elevations from the tide gauge on the west end of the island down to PGF Pole Tern 1 and FRIG RM 2.  We recovered three epoxied bolts recorded as bench marks for this station, and we ran a level loop that closed with an error of 6 mm.  The UH Sea Level Center provided a value for mean sea level at the tide gauge for the period from 1974 to the present, and this value can provide an initial estimate of the geoid separation at Tern Island.  Differencing the astronomical coordinates at FRIG RM2 against the geodetic coordinates will provide an estimate of the slope of the geoid, in the form of deflections of the vertical in latitude and longitude.


The Tern Island Survey

At 11:50 a.m., following a 2 ½ hour session, I turned off the Trimble receiver on FRIG RM2 1961.  At 1:56 p.m. I turned off the Ashtech receiver on PGF Pole Tern1.  My GPS observations at French Frigate Shoals were at an end.  In spite of the dubious beginnings with CP Innocence, this imaginary point proved itself quite useful in the final analysis.  I was able to observe useful GPS observations with the Ashtech receiver for virtually the entire duration of my stay, and this would make the processing of the long baselines from other Pacific islands, including the main Hawaiian Islands, more likely to produce accurate results.


After lunch I packed up all of my gear, and at 2:30 p.m. the Cromwell appeared on the horizon to the southwest and anchored 2 miles away from Tern Island.  LaPerouse Pinnacle stood guard as the ship’s boat was launched to pick me up on Tern, and by dinnertime I had boarded the Cromwell for the two days’ sail to Laysan Island, the next chapter in my survey in the Northwest Chain.


Aboard the Cromwell


The Cromwell was to be my home until my return to Honolulu on March 18.  Stationed in Honolulu, she is attached to the National Marine Fisheries Service (NOAA) Laboratory there.  When I boarded her off Tern Island she was being used to carry supplies and personnel to NMFS and FWS camps on Laysan Island and Lisianski Island, as well as to offload personnel and material on Tern..



NOAA photograph of the Townsend Cromwell

In the wet lab aboard the Cromwell were the materials for constructing another PGF antenna pole, this time on Laysan.  Included in these materials were 750


Boarding the Townsend Cromwell for Laysan from Tern, LaPerouse Pinnacle watching

pounds of dry Quickcrete.  Because the sand is so deep on Laysan, the PGF engineers had determined that only a large concrete underground monument would provide sufficient stability for an antenna pole.  It would take at least two days to build the monument, allowing time for the concrete to cure.  My schedule called for three days on Laysan (shortened from the original schedule of five days), during which time I would build the monument, occupy any USC&GS monuments, and observe image registration points.  This probably wouldn’t allow for enough time to occupy the new monument long enough to compute a baseline to it, but one of the old monuments could be used for a long occupation, and precise geodetic coordinates could be transferred from the old monument to the new pole after it was constructed.  There were no tide gauges on the island, and so the value of running levels from the sea surface to ground monuments was minimal priority.


The food was good and the living was easy aboard the Cromwell during our two-day passage to Laysan Island, but I was eager to get ashore on Laysan as soon as we arrived.




Arrival at Laysan Island, a grove of coconut palms on the left

Unfortunately, however, seas as high as 30 feet prevented making a beach landing when we arrived.  The entire beach was closed off by surf.  The Cromwell altered its agenda by leaving Laysan for Lisianski Island, where 3 monk seal researchers were to be dropped off with their field camp.  The 1961 NGS survey documents indicate existing survey monuments on Lisianski, but because the entire offloading operation was over within 4 hours there was no point in my attempting a survey there.


Laysan Island


We returned to Laysan, but the seas continued to prevent a beach landing.  For two days the Cromwell circled Laysan, waiting for a chance to offload.  The only way that the Cromwell could have returned to Honolulu and kept on schedule would have been to leave Laysan after the first day.  But we waited one more day, and finally on the morning of March 13 conditions allowed for a beach landing.  Instead of 3 days on Laysan, I would have at the most 8 hours.  The Cromwell would be leaving Laysan for Honolulu that same evening.  I was aboard the first boat to land on the beach, and I was aboard the last boat to leave the beach to return to the Cromwell and Honolulu.



Laysan Island, looking to the southwest

All visitors to Laysan Island are required to wear brand-new clothing to prevent any accidental contamination of the indigenous vegetation with alien seeds.  The pelican cases for the receivers had to be frozen for at least 24 hours to kill alien life forms.  These precautions seemed excessive to me, but I was glad to follow them in order to go ashore on this special island


The FWS and NMFS camps are close at hand to the small cove on the northwest corner of the island that is used for boat transfers.  My guide was Eric Lund, a FWS volunteer who would be returning to Honolulu aboard the Cromwell after 5 months on the island.  While we surveyed the island the rest of the landing party offloaded the 750 pounts of Quickcrete.  The first step in our survey was to find the two USC&GS monuments known to exist on the island, LAYS 1961 and LAYS RM2 1961, both conveniently located in camp.


I would have preferred to use RM2 for my long-baseline occupation, but a wooden antenna mast for a defunct radar had been lashed to the concrete pillar, and neither GPS antenna could be positioned directly over the brass disk because of it.  Instead, I used LAYS 1961, a brass disk in a concrete post projecting about a foot above the sand, as the long-term occupation.  I plumbed the antenna over the station mark with Eric’s assistance using the auto-level method.


Looking southeast to LAYS 1961 and the FWS camp, with Laysan’s lake in the background

Somehow I would have to transfer coordinates to LAYS RM2, but I didn’t have any time to waste.  I made two arbitrary black marks on the westerly top of RM2’s pillar, called them RM2-A and RM2-B, and carefully measured the short distances between them and the punch mark on the brass disk.  The coordinates of LAYS RM2 1961 could be accurately computed using these observations.  I then occupied each of these eccentric stations for about a half-hour each, while Eric and I chose image registration points, planned our expedition around the island, and ate a lunch of power bars.


The FWS camp would show up well on the IKONOS image of Laysan, and so the pillar could be used as a registration point.  Eric and I chose and old shipwreck on the south end of the island, and an unusual formation of rock on the east side of the island, as the other two registration points, and we began our hike down the western edge of the island to reach them.


It took about an hour.  Wildlife regulations required us to make wide detours around the monk seals on the beach, and if a misstep should collapse a bird burrow we had to dig it out.  Albatross chicks, frigate bird nests, red-tailed tropic birds, and boobies were everywhere, and it was impossible to walk in a straight line.  Laysan’s westerly beach is sandy and lovely, and just off the beach the coral reefs and brilliant turquoise water promised superb snorkeling.  I enjoyed the beauty of the scene even more than I felt sorry for myself for not being able to go snorkeling there.  Finally we reached the shipwreck at the tip of the island, and observed the north end of the old steel ship’s prow sticking up through the beach sand. 


The LAYS RM2 1961 pillar, with the radar antenna pole lashed to the east side

As soon as this session was complete we resumed our hike, proceeding north along the island’s easterly shore, until we reached the rock pile at the edge of the vegetation line.  I deployed the receiver’s antenna in the middle of this rock pile for our second registration point.


We were moving as quickly as we could.  I wanted to observe another USC&GS monument known to FWS personnel to be on the northwest end of the island, and to reach it Eric and I moved inland, to the east shore of the unique brackish lake in Laysan’s interior.  By moving along the fringe of the lake we could make the best time.  A grove of coconut palms is situated at the north end of the lake, and by hiking through this grove and up to a low ridge we could reach the monument.  I think this monument is probably AZ. MK., an unstamped disk, but I wouldn’t know for sure until I saw it.  The 1961 documentation also records LAYSAN MAGNETIC somewhere in the northeasterly area of Laysan, but because there is no other record of this monument and because there is no recorded distance to it from LAYS and because the FWS personnel knew nothing about it, I didn’t even look for it.  There wasn’t enough time.


Throughout this trek I had been carrying a hand-held radio, with which I monitored communications between the Cromwell and the shore parties.  Sometimes our reception was interrupted by line-of-sight obstacles, but we always moved back into range after a few minutes.  Eric and I were on the easterly shore of the lake, opposite the FWS camp, and moving quickly towards the coconut grove, when I heard the captain of the Cromwell on the radio summon a search party to find us.  His instructions were to send one of the Cromwell’s boats to circumnavigate the island, to see if we could be found.  Obviously he had been trying to communicate with me while I was out of radio range.


I called the Cromwell and canceled the search party.  Captain Callahan sounded relieved to hear from me, but he had bad news.  The offload of goods and personnel from the Cromwell to the beach was complete, and so was the onload of goods from the beach to the Cromwell.  Everyone still supposed to board the Cromwell was on the beach, ready to go, except for my guide and me.  I was instructed to return to camp, dismantle my GPS equipment, and report to the beach immediately.


Eric and I walked briskly around (and, in places, through) the lake, and up a slight slope to the FWS camp.  There I turned off the Ashtech receiver, after logging a session of 5.5 hours.  It will take some good luck, low ionospheric energy levels, and expert baseline processing to produce good baselines from LAYS 1961 to the CORS sites.


Everyone was eager to help me pack up my gear and carry it down to the beach.  We waded into the surf, loaded the gear into the boat, and as soon as I climbed aboard the coxwain gunned the motor and ferried us back to the Cromwell.  I had been on Laysan for a total of 7 hours.



The Cromwell along Kauai’s NaPali coast on her return to Honolulu

The Cromwell sailed back to Honolulu and docked at Sand Island at first light on March 18.  My surveying campaign in the Northwest Chain of the Hawaiian Islands had come to an end.




Both GPS surveys were post-processed by WAVE using the precise IGS ephemeris and holding fixed ITRF97 coordinates on the CORS stations Kokee Park and Kokole Point on Kauai and Upolu Point on the Big Island.  The long baseline coordinates on Tern Island were adjusted with formal a posteriori errors of .0045m, .0114m, and .0185m in latitude, longitude, and ellipsoid height respectively.  The long baseline coordinates on Laysan were adjusted with formal errors of .0184m, .0427m, and .1233m in latitude, longitude, and ellipsoid height respectively.  The coordinate shifts from astronomical to geodetic were 365.5m in an azimuth of 157 degrees on Tern, and 231.7m in an azimuth of 70 degrees on Laysan.  The observed geoid height on Tern Island (using the mean sea level datum from the tide gauge) was 17.0655m, versus an interpolated geoid height of 15.1111m taken from the global 15‘ x 15’ geoid model WW15MGH.  Because no tide station exists on Laysan Island it wasn’t possible to observe an apparent geoid height using mean sea level.  The numerical results of the surveys are summarized below:



FRIG RM2 1961

LAYS RM2 1961

LAYS 1961

Geodetic f                 

N 23°52'10.475108"

N 25°46'27.362395"

N 25°46'28.026712"

Geodetic l

W 166°16'58.852292"

W 171°44'20.136409"

W 171°44'20.064546"

Astronomical f  (1961)

N 23°52'21.440000"

N 25°46'24.890000"

N 25°46'25.550000"

Astronomical l  (1961)      

W 166°17'03.730000"

W 171°44'27.990000"

W 171°44'27.910000"





h      (sea level)




D f




D l




N  (observed)




x  (observed)




h  (observed)




LaPlace Azimuth Corr.




N  (from geoid model)




x  (from geoid model)




h  (from geoid model)




LaPlace Corr (from model)





There are significant differences between the observed deflections of the vertical for both surveys and the deflections of the vertical estimated from the geoid model.  The reason for the difference is graphically illustrated by a wire frame 3D image of the WW15MGH geoid as it is modeled from the main Hawaiian Islands to Laysan Island.


A 3D representation of the geoid (WW15MGH) along the Hawaiian Ridge

The geoidal slopes along the Hawaiian Ridge are so steep and deep that a relatively small change in horizontal position will result in dramatically large changes in geoid slope and geoid height.


Another (profile) view of the geoid along the Hawaiian Ridge clearly reflects the remnant islands that constitute the Northwest Chain.


The profile of the geoid surface running along the Hawaiian Ridge.

It will obviously require a great many well-distributed observations of geoid height to define the geoid accurately along the Hawaiian Ridge.




Although I began this project without much fanfare, by the time I actually made the trip a great deal of new interest in the NW Chain had been generated by the Coral Reef Ecosystem Initiative announced by President Clinton in his last month in office.  In August and September NGS mounted two expeditions to the NW Chain to bring in GPS to all of the islands.  Information about these expeditions can be found on the NGS website,


Also, since this report was written, UHPGF has produced a Pacific GAMIT solution for the primary monuments on Tern and Laysan Islands.  These results are not included in this report.


Further information


Information about working as a volunteer with the USFWS can be found on this agency’s web page at   Any questions about this article can be directed to me at and by visiting  Rod Low can be contacted at the ESRI offices in Honolulu.  Mike Bevis can be contacted at  Dave Doyle can be contacted at  The Townsend Cromwell has its own web page.




I owe a debt of gratitude to everyone I met and worked with during my expedition to the Northwest Chain of the Hawaiian Islands.  In particular I want to thank Rod Low, Mike Bevis, and Dave Doyle.  I also owe a debt of gratitude to my old company, Trimble Navigation, for supporting me in this volunteer effort, and to Cindy Rehkemper of FWS in Honolulu for handling with good humor all the logistics of my trip.