John Beavan [1]
Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY 10964
Tel: (914) 359 2900; Fax: (914) 365 8150; e-mail: beavan@ldeo.columbia.edu
This project investigates the current state of deformation within, and the structure of the crust and upper mantle beneath, the Adirondacks. The current state of deformation is being assessed through a Global Positioning System (GPS) geodetic experiment that will reoccupy monuments from surveys conducted over the past 100 years. The structure of the crust and upper mantle is being investigated through the analysis of teleseismic data collected from a year-long deployment of 4 broadband seismometers in different parts of the Adirondacks. These seismic data will be used to determine the shear wave velocity structure beneath the stations from receiver function analysis, and upper mantle anisotropy from splitting of SKS arrivals, both of which will provide valuable clues about the existence (or lack thereof) of any anomalously hot region beneath the mountains. Additional information about the ambient stress field is obtained from recordings of local and regional earthquakes.
A fuller description of the background to the project is given in Appendix 1.
Two substantial sets of historical triangulation data are available. One is
the U. S. Coast and Geodetic Survey's (USCGS) measurements made in the early
1940s, principally in 1942. The network is shown in
Figure 1. We obtained
these data in 1994 from the National Geodetic Survey's (NGS) data base, with
the assistance of Richard Yorczyk of NGS.We note here that the data held in NGS's computer archive are not the original theodolite readings but are the averages of the set of directions made by the surveyor under the procedures in force at the time. Any reductions to center of offset instruments or targets have already been made in the archived data. The data are stored in the archive as "rounds" in the way they were observed in the field, which means that correlations in the data can be properly modelled in the analysis.
The second set of data are from the period 1872-1900. During this period surveys within the Adirondacks were made by two organisations. A relatively small number of observations were made by the U. S. Coast Survey (forerunner of USCGS and NGS). These data are held in the NGS database and were provided to us by Rick Yorczyk (in slightly non-standard format so as to preserve the correct century in the date, since only a two-digit year is represented in the standard format). A second and much more extensive set of data were collected under the direction of Verplanck Colvin, mainly in his role as director of the Adirondack Survey (AS, later renamed the New York Land Survey or NYLS) but also in part through his own resources. An introduction to Colvin's work in the Adirondacks is given by Rosevear [1994] and the (irregularly published) annual reports of the AS and NYLS also contain much information.
When the NYLS was terminated in 1900 and its responsibilities given to another part of the State government, a battle ensued over the ownership of the AS and NYLS records which unfortunately led to an incomplete archiving of these data. Colvin's original field books (which contain much valuable information in addition to survey records) are held in the archives of the New York State Department of Environment and Conservation (NYSDEC). Also held by NYSDEC is a cabinet full of paper records of processed angles, in which averaging of individual sights and reductions to center of offset targets have been computed (see Appendix 2 for some examples).
We did not have the resources to go through the original field books to extract data. We did, however, copy the complete set of processed angles, and entered most of these data into the computer. Some stations in the data we copied were not well connected to the network, and these were omitted from our analysis. Since only processed angles are available, we do not have information on whether the AS/NYLS observed in rounds or as individual angles, though this is in principle available from the field books. This means that correlations in the data are not necessarily properly modelled in the analysis (though with the relatively poor accuracies attained from the AS data, this is not a significant problem).
.
The combined late-1800s data set that we processed is shown in
Figure 2.
Unfortunately, this represents vastly less data than were actually collected by
the AS/NYLS, as described in the annual reports of the AS and NYLS. Several
different theodolites were used by Colvin, ranging from 4" models to a massive
20" model (made by Oerthling of Berlin) that weighed 300 lb and required a
special "system of harness and bars" to enable the packmen to transport it to
the summits. Unfortunately, we found none of the data from the most accurate
instrument - the 20" theodolite - in the processed angle records, and only a
relatively small amount from the other larger instruments.
It is undoubtedly possible to recover a much larger proportion of Colvin's work by extracting data from the original field books (which have been microfilmed by NYSDEC and are therefore available for study). This would, however, be a significant undertaking. We decided to first process what data were relatively easily available from the 1800s together with the 1942 triangulation and 1995 GPS. Depending on our results, future researchers can judge the merits of recovering and processing additional data from Colvin's field books.
In several cases, all existing marks on a particular summit were so obscured by trees and/or a fire tower as to be unsuitable for GPS measurements. Since some of these sites were in crucial locations we planned to install new survey markers in clearings on these summits and to make geodetic ties between the new mark and the preexisting marks using a combination of conventional survey equipment (total station) for distance and angle determination, and GPS for geodetic azimuth determination.
The survey marks (both 1800s and 1940s) were in many cases damaged or missing. For completely missing marks a hole remains in the rock, which - provided it can be found and correctly identified - means the station can be recovered. Damaged marks are of two sorts. Defacing of the surface is not generally a severe problem since some text can usually be made out, and the center of the disk located quite accurately. More commonly, the disk has been pried off leaving a metal stem in the rock. Provided this can be confidently identified the station can be recovered by using the center of the stem (with allowance if necessary for bending of the stem).
There were several clues that helped to confidently identify missing AS/NYLS
marks. First, apart from the earliest marks installed by Colvin, they were of
a standard approximately 2" size. Colvin countersank the head of the disk into
the rock so that when the disk is missing a characteristic depression is
visible (
Figure 3). A second hole is often drilled a few inches away from the
disk (this was presumably used in some way in Colvin's tripod setups). In many
cases an equilateral triangle of small holes had been drilled around the survey
mark in which the feet of the survey tripod were placed. Finally, the targets
set up by Colvin consisted of large wooden towers (constructed in situ) that
were attached to iron ring bolts leaded into the rock. (A hole is drilled and
a sheath of lead placed in the hole, then the bolt is hammered in with the lead
deforming to hold it in place.) These bolts were generally in a near-perfect
square (or in a few cases triangle) surrounding the survey mark, with the
dimension of the square being recorded in Colvin's site descriptions. In a few
cases the actual rings are still in place, and in many other cases it is
possible to identify the iron stems of the ring bolts and the characteristic
sheath of extruded lead surrounding the stem. When the remains of the ring
bolts, the tripod holes and the characteristic shape of the survey mark hole
can all be identified, a hole or stem can be associated with Colvin's mark with
the greatest confidence.
The 1940s marks were in general easier to identify even when the disks were missing. This is because of the detailed NGS descriptions and the presence of reference marks at known distance and azimuth from the main mark.
The 1940s survey generally set two reference marks (RM) in addition to the main mark (both in cases where a new main mark was set, and where the AS/NYLS mark was reused). In most cases, angles were turned to these marks from some distant station and a distance was measured between the main mark and the RM, with this information being recorded in the NGS site description. In some cases RMs were set without one or both of these measurements being recorded. The distances to RMs were not necessarily made very accurately in the 1940s, and it is not always clear whether it is a horizontal or a slope distance that is recorded in the site description (though it is meant to be horizontal). In a few cases, precise distance measurements to RMs had been made by NGS in later (post-1960) surveys. In some cases, similar ties were made between the 1940s main mark and the AS mark.
These data are important for a number of reasons. In several instances one of the RMs at a particular site had the best sky visibility for GPS measurements. If we were confident that a good tie (both distance and azimuth) had already been made between the main mark and this RM then we could occupy the RM with GPS without needing to also make a geodetic tie back to the main mark. Similarly, if there were a 1940s mark and an AS mark on the same summit, we could utilise any preexisting tie between these marks. In some cases we repeated 1940s local ties in 1995 in order to give us a feeling for how accurate the 1940s ties had been and therefore how much reliance we should place upon them. (The worst difference we found - apart from actual blunders - was about 2 cm.)
In cases where we needed to occupy a RM that was not already tied to the main mark, or in cases where the AS and 1940s marks were not already tied, we performed our own local survey. The local network was measured using a Zeiss Elta-4 total station with an accuracy specification of 3 mm plus 3 ppm in distance, and probably better than 10 seconds of arc in direction. (For the short lines we were measuring the 3 ppm proportional error is irrelevant.) This procedure gives the size and shape of the local network but does not give the overall orientation of the network. (Through the measurement of vertical angles it also gives the height differences between the local marks, which were not measured in the 1940s.)
In the 1940s, the orientation of the local tie was provided by also observing the direction to a remote station that was part of the overall geodetic network. In 1995 this was not possible since targets were not set up on remote stations, and visibility due to tree growth at the local site was often a problem too. We adopted two different procedures to overcome this. (1) When an oriented 1940s local tie (i.e., including a remote azimuth) already existed and we needed to tie an additional mark to the network (a new mark or a previously untied AS mark) we simply turned an angle off the existing network. (2) When there was enough clear space on the summit we occupied a short line with two GPS receivers in order to define an azimuth. Note that one or both ends of this line can be temporary marks provided that the local total station survey is tied to the azimuth line before the temporary marks are moved.
In two cases we did not have the NGS tie data available, but we did have NGS's published coordinates of the two marks. In these cases we used the differences between the published coordinates to provide the tie data.
We arranged for ten Trimble 4000SSE GPS receivers for the project (Table 1).
| Receiver source | Qty | Receiver codes | Antenna | Firmware |
| LDEO | 2 | LD01, LD02, | geodetic | 6.10 |
| LDEO | 2 | LD04, LD05 | compact | 6.10 |
| IGNS | 2 | GNS01, GNS02 | compact | 6.10 |
| CT Male & Assoc. | 2 | CTMale01, CTMale02 | compact | 5.68 |
| CT Male & Assoc. | 1 | CTMale03 | compact | 6.01 |
| NASA/GSFC | 1 | GSFC01 | geodetic | 5.60 |
We kept three of these at fixed sites throughout the campaign, nominally collecting continuous data, and moved the other seven between sites (with one of these being used for local survey ties). We planned to have one 3-day burst of measurements, followed by a return to Troy and a few days rest, then a 4-day burst of measurements. An introductory session, which most of the volunteers attended, was held at RPI a few days before the first burst. At this session we described the reasons for doing the project, gave a brief introduction to GPS, practised tripod setups, sorted people into parties, and assigned the first three-day burst of measurements to each party.
The routes and parties were planned with the following constraints in mind.
We by no means succeeded in balancing all these constraints, but the first burst of measurements was completed reasonably according to plan. The weather was much kinder to us than it could have been, in that there was virtually no rain and the trails were unusually dry. But it was extremely hot and humid and this caused discomfort for everybody, and significant problems for a few. (Temperatures reached over 100deg.F in Plattsburg though it was somewhat cooler in the mountains.)
At two previously unreconnoitered sites it proved impossible to locate the survey mark. At one site, all the existing survey marks had such bad sky visibility that a new site was measured over a mark scratched in the rock (which we later replaced with a stainless pin in a drilled hole). Despite our efforts to check out batteries beforehand, some sessions were shortened due to batteries running down prematurely.
Other commitments, plus the discomforts of the first burst and the continuing high temperatures, ate into our pool of operators for the second burst so we had to scale back our plans to some extent. Even with the scaling back, a number of planned sites were not reached so that the second burst was less fully completed than the first.
The network as measured is shown in
Figure 4
and the station occupation
schedule is given in
Figure 5. The fixed sites were Mindy, Hallock Hill and
RPI. Of the 32 non-fixed stations, 29 were summits and 3 were NGS stations
that had previous occupations by GPS during the 1992 New England High Accuracy
Reference Network (HARN) campaign. (These are York, Indian and Ticonderoga;
the RPI station was also run during this campaign but its data were not
analyzed by NGS.) Of the 29 summits, 5 were measured twice and the remaining
24 only once.
A number of loose ends were left at the end of the campaign. These were of several types.
We attempted to address issues (1) through (4), and decided to leave (5) and (6) until a thorough analysis had been made of the data collected so far.
(1) was addressed by debriefing more carefully the personnel who visited that site, and in some cases by revisiting the site.
(2), (3) and (4) were addressed by visiting these sites with a combination of GPS receivers and total station. Some of this work was done during the survey, some between bursts 1 and 2, some just after burst 2, and some by RPI personnel in the 3 months following the main survey. The later work was done with LDEO's Zeiss total station and with two Ashtech Z-12 GPS receivers from RPI.
There remains some doubt on the identification of Colvin's monument at two sites (Hurricane and Snowy), and the angular tie at one site is poor (Owl's Head). However, these deficiencies are not sufficient to prevent adequate data analysis.
Appendix 3 gives a summary of what mark was occupied at each site, what the condition of the marks is, and what future work could be done to improve the site itself or the measurements at the site.
Each day of data was then reanalyzed in a network solution holding RPI1 fixed at the value determined in the previous step, and solving for all other stations including Algonquin and Westford [3]. Tropospheric parameters were estimated every two hours, and ambiguity resolution was attempted in two steps - first for L5 (wide-lane) ambiguities then for L1 (narrow-lane). Between 80% and 97% of ambiguities were successfully resolved, depending on the day. The coordinate and covariance files were then combined, together with local survey data when several marks were present on the same summit. This combination was accomplished by variation of coordinates using program ADJCOORD [Crook, 1992] and the coordinate results are given in Table 2 [4].
These coordinate estimates are with respect to the WGS-84 ellipsoid. The table gives a posteriori 95% confidence limits on these estimates, and also includes an estimate of geoid-ellipsoid separation calculated from NGS's GEOID93 model. Station height above sea level can be approximated by subtracting the geoid height from the ellipsoidal height, though this does not give an exact result since there are small distortions between GEOID93 and NGS's North American Vertical Datum 1988 (NAVD88) [Ed McKay, pers. comm., 1996].
Since we occupied three stations whose North American Datum 1983 (NAD83) horizontal coordinates and NAVD88 orthometric heights (heights above sea level) had been calculated by NGS from 1992 GPS data, it is possible to adjust our data, using ADJCOORD, to fit these coordinates exactly and thus to transform our coordinates into a reference frame consistent with NAD83 and NAVD88. The GEOID93 model is used to provide differential geoid height corrections between the known stations and the other stations in the network in order to provide orthometric heights for all stations consistent with the NAVD88 datum. The results of this adjustment are given in Table 3. Since NGS gives the orthometric heights of the three 1992 GPS stations only to the nearest 0.1 m and errors in the geoid model may also be at this level, we give the orthometric heights in Table 3 to the nearest 0.1 m (optimistically) and also to the nearest foot (probably a more realistic error estimate). We find that the heights of the highest peaks we measured differ by up to several feet from their established values, but that our height for Mt. Marcy is the same at 5344 ft.
We then did a static adjustment of the 1800s triangulation combined with the 1995 GPS, again with the aim of detecting outliers that might be made apparent by the addition of the far more accurate GPS data. Here we were looking for significant outliers that might bias the results of subsequent deformation analysis, under the assumption that any deformation between the 1880s and 1995 must be small. A similar procedure was done for the 1940s triangulation.
The results for 1880-1995 are given in Figure 6a, with 95% confidence ellipses on the observed displacements. Note that few of the estimates are larger than their error estimates and that the displacements - more than 2 m near the edges of the network - are unrealistically large. The large errors to the north and south are perhaps explainable by the poor connection of these stations to the network ( Fig. 2) so the analysis is repeated in Figure 6b with these stations removed (note the factor of two scale change). The displacements still approach 1 m (or 1 cm/yr over 100 years) and are generally not significant at the 95% confidence level.
Similar results for 1942-1995 are given in Figures 7a and 7b. Though the displacements are substantially smaller - 40 cm maximum - none are really significant even when the largest-displacement stations are omitted from the analysis.
It is possible to go through the exercise of calculating the uniform strain field that best fits these data, and determining whether this strain field is statistically significant. For the 1880-1995 period, the (engineering) shear strain is 0.12+/-0.11 urad/yr with the axis of maximum relative contraction directed NW-SE. For the 1942-1995 period, the shear strain is 0.06+/-0.05 urad/yr with the axis of maximum relative extension directed NW-SE (95% confidence estimates). While the first of these is fortuitously aligned in the same direction as stress indicators from the area [e.g., Zoback and Zoback, 1988, 1991], the second is in the orthogonal direction. In any case, both estimates are too large to be realistic, and neither is significant at the 95% confidence level.
University NAVSTAR Consortium (UNAVCO) archive, Boulder, Colorado
Rensselaer Polytechnic Institute, Troy, NY
Institute of Geological and Nuclear Sciences, Lower Hutt, New Zealand
There are items that should be submitted to the National Geodetic Survey, but this has not yet been done because we have not had the resources to transform the data to NGS's "Blue Book" format. First, station recovery notes should be submitted for all stations that we visited and for new marks that we installed. Second, the 1995 GPS data and local tie (total station) data should be submitted. Third, the processed results of the 1995 survey should be submitted. The archives described above contain all the information required for Blue Book format, but a considerable time investment will be necessary to convert the data.
With these errors, and with estimated uplift rates in the 1-3 mm/yr range we could not expect a future survey to detect uplift for at least 20-50 years, though we might expect to detect associated horizontal deformation substantially sooner than that. However, the accuracy of the 1995 data would be significantly enhanced if additional data were collected at the same points over the next few years. Then the baseline for future studies would consist of a combination of the 1995 and next few years' data. There is the possibility of doing this on a piecemeal basis with one or two receivers and without mounting a major campaign, provided that a base station operates in or near the Adirondacks while the summit measurements are underway.
The 1995 measurements covered the eastern slopes of the Adirondack dome ( Fig. 4) but failed to cover any of the less accessible western slopes. Future measurements should redress this problem, since it is probable that the largest horizontal deformations occur around the slopes of the dome rather than in its centre. Since we have demonstrated that the 1800s and 1940s triangulation surveys are not precise enough to be used for Adirondack deformation studies, it is not necessary for future sites in the western Adirondacks to occupy preexisting survey marks. This may allow sites with easier accessibility to be chosen for such measurements.
We estimate that a resurvey of the 1995 network in about 20 years would have a chance of detecting Adirondack deformation. This time scale would be significantly shortened if the 1995 measurements were enhanced by a program of piecemeal measurements over the next few years.
We are grateful to Des Darby at the Institute of Geological and Nuclear Sciences (IGNS), New Zealand, Steve Cohen at NASA Goddard Space Flight Center, and Clark Donkin at CT Male Associates for the loan of GPS receivers for the 1995 measurement campaign. We are indebted to the RPI and LDEO personnel who participated in the measurements, and in particular to the volunteers who little suspected what they were getting themselves into! We single out Marcy Howe for special mention, and also thank: Robbie Abad, Scot Ballard, Daniel Bentz, David Borton, Jennifer Butler, Arthur Chen, Nathan Davies, Ilya Dricker, Kirk Gendron, Hamilton Gillett, Shannon Jock, Dennis Keane, Neal Knitel, Ted Koczynski, John McMurray, Patrick Radibeau, Garfield Raymond, Frank Revetta, Steve, Deborah, Geoffrey and Gretchen Roecker, Kevin Rudman, Chris Schildge, Jay Sklar, Peter Spotts, Howard Stoner, John Sykes, Steve Tice, Mary Roden-Tice, Peter, Elizabeth, Hanya and Michal Zwick.
We are grateful to the landowners and caretakers who allowed access to sites on their land and provided assistance in the field: Steve Bashaw, Len Cornier, Doug Crary (Camp Otter Brook), Joe Cummings, Ed Davis (Hamilton Lake Seminar Center), Anne Garrand, Jim Hunley (International Paper), Doug Knight, Dave Kress (Big Tupper Ski Center), Harold Lawson, Tim Sprague (International Paper), Earl Svendsen, Nick Westbrook (Fort Ticonderoga), Joe Whalen.
We thank IGNS, which provided facilities for the analysis of the GPS data, and particularly Charlotte Heinz who did the GPS analysis and the initial deformation analysis at IGNS during part of her Second Practical Term from Fachhochschule München.
This project has been funded in part by a U.S. Geological Survey Earthquake Hazards Reduction Program grant to LDEO, grant no. 1434-94-G-2473.
Brown, L., C. Andos, S. Klemperer, J. Oliver, S. Kaufman, B. Czuchra, T. Walsh, and Y. Isachsen, Adirondack-Appalachian crustal structure: The COCORP Northeast Traverse, Geol. Soc. Am. Bull., 94, 1173-1184, 1983.
Connerney, J. E. P., and A. F. Kuckes, Gradient analysis of geomagnetic fluctuations in the Adirondacks, J. Geophys. Res., 85, 2615-2624, 1980.
Connerney, J. E. P., A. Nekut, and A. F. Kuckes, Deep crustal electrical conductivity in the Adirondacks, J. Geophys. Res., 85, 2602-2614, 1980.
Crook, C.N., ADJCOORD: A Fortran program for survey adjustment and deformation modelling, N. Z. Geological Survey, EDS Report, DSIR Geol. and Geophys., Lower Hutt, New Zealand, 22 p., 1992.
Dawers, N.H. and L. Seeber, Intraplate faults revealed in crystalline bedrock in the 1983 Goodnow and 1985 Ardsley epicentral areas, New York, Tectonophysics, 186, 115-131, 1991.
Dreger, D.S., and D.V. Helmberger, Broadband modeling of local earthquakes, Bull. Seism. Soc. Am., 80, 1162-1179, 1990.
Geraghty, E.P., and Y.W. Isachsen, Investigation of the McGregor-Saratoga-Ballston Lake fault system - east central New York, U.S. Nuclear Regulatory Commission/CR-1866, 44 pp., 1980.
Helmberger, D.V., Generalized ray theory for shear dislocations, Bull. Seism. Soc. Am., 64, 45-64, 1974.
Isachsen, Y.W., Possible evidence for contemporary doming of the Adirondack mountains, New York, and suggested implications for regional tectonics and seismicity, Tectonophysics, 29, 169-181, 1975.
Isachsen, Y.W., Contemporary doming of the Adirondack mountains: Further evidence from releveling, Tectonophysics, 71, 95-96, 1981.
Klemperer, S. L., L.D. Brown, J.E. Oliver, C.J. Ando, B.L. Czuchra, and S. Kaufman, Some results of COCORP seismic reflection profiling in the Grenville-age Adirondack mountains, New York State, Can. J. Earth Sci., 22, 141-152, 1985.
Kosarev, G.L., N. V. Petersen, L. P. Vinnik, and S.W. Roecker, Receiver Functions for the Tien Shan Analog Broadband Network: Contrasts in the Evolution of Structures Across the Talas-Fergana Fault, J. Geophys. Res., 98, 4437-4448, 1993.
Makeyeva, L.I., L. P. Vinnik, and S.W. Roecker, Shear-Wave Splitting and Small-Scale Convection in the Continental Upper Mantle, Nature, 358, 144-147, 1992.
Nabelek, J. and G. Suarez, The 1983 Goodnow earthquake in the central Adirondacks, NY: a broad band teleseismic analysis, in: K. Jacob (editor), Proceedings from the Symposium on Seismic Hazards, Ground Motions, Soil Liquefaction, and Engineering Practice in Eastern North America, Natl. Center for Earthquake Eng. Res. Tech. Rep., NCEER-87-0025, 300-317, 1987.
Owens, T. J., Crustal structure of the Adirondacks determined from broadband teleseismic waveform modeling, J. Geophys. Res, 92, 6391-6401, 1987.
Rocken, C. and C. Meertens, UNAVCO Receiver Tests, UNAVCO Memo dated Nov 8, 1992, available from UNAVCO/UCAR, Boulder, Co.
Rosevear, F.B., Colvin in the Adirondacks: A Chronology and Index, North Country Books, Utica, New York, 164 p., 1992
Rothacher, M., G. Beutler, W. Gurtner, E. Brockmann and L. Mervart, Bernese GPS Software Version 3.4 Documentation, Astronomical Institute, University of Bern, 1993.
Sbar, M.L., and L.R. Sykes, Seismicity and lithospheric stress in New York and adjacent areas, J. Geophys. Res., 82, 5771-5789, 1977.
Seeber, L. and J. Armbruster, A study of earthquake hazards in new York State and adjacent areas, U.S. Nucl. Regul. Comm. Rep., NUREG/CR-4750, 98 pp, 1986.
Seeber, L., and J.G. Armbruster, Low displacement seismogenic faults and nonstationary seismicity in the eastern United States, Annals N.Y. Acad. Sci., 558, 21-39, 1989.
Tice, S. J., A paleoseismic investigation of the McGregor fault, east-central New York, M. S. thesis, SUNY Albany, 129 pp., 1993.
Wallace, T.C., and D.V. Helmberger, Determining source parameters of moderate size earthquakes from regional waveforms, Phys. Earth Planet. Inter., 30, 185-196, 1982.
Wallace, T.C., The inversion of long-period regional distance body waves for crustal structure, Geophys. Res. Let., 13, 749-752, 1986.
Wallace, T.C., Determination of source parmaters for small earthquakes from a single, very broad band seismic station, Seism. Res. Let., 61, 26, 1990.
Zoback, M. D. and M. L. Zoback, Tectonic stress field of the continental United States, in: Geophysical framework of the continental United States, L. Pakiser and W. Mooney, Eds., Geol. Soc. Am. Memoir, 1988.
Zoback, M.D. and M.L. Zoback, Tectonic stress field of North America and relative plate motions, in Slemmons, D.B., et al., eds., Neotectonics of North America, Boulder, Co, Geol. Soc. Am., Decade Map Vol. 1, 1991
The Adirondack mountains of upstate New York ( Figure A1) lie in a roughly circular region about 200 km in diameter. Elevations of several peaks exceed 1200 meters, and the highest, Mt. Marcy, is about 1600 meters high. The Adirondacks are composed largely of metamorphic rocks similar in age (middle Proterozoic) and type to the extensive Grenville Province exposed to the northwest in Canada. They were at a depth of about 30 km for most of their history, but were uplifted beginning some time perhaps at late as the Tertiary, although the details of the uplift history are controversial. As uplift continued, ongoing erosion removed all of the overlying sediment and cut deep gorges into the basement, exposing a window on the basement lying below much of the eastern U.S. If uplift commenced as late as 10-20 million years ago, an average uplift rate of about 3 mm/yr would be required to bring the lower crust to the surface. Such rapid uplift would constitute a major source of strain accumulation on the faults that bound the mountains if this rate continues to this day. However, the current rate of uplift of the Adirondacks, if uplift is occurring at all, is a hotly debated topic.
Prior investigations of contemporary uplift in the Adirondacks provide intriguing, if inconclusive, evidence of active deformation. Some indirect observations, such as a gravity study showing the sharpness of fault scarps at depth along the boundaries of the dome (Tice, 1993; Roecker et al., manuscript in prep., 1993) suggest that recent uplift has outpaced erosion. Releveling lines run through and along the edge of the Adirondacks in the mid 1970s (e.g., Isachsen, 1975; 1981) of lines previously surveyed in 1931 and 1955 suggest that the mountains may be uplifting at rates of as much as 3.7 mm/year, or about 10 times the erosion rate ( Figures A1 and A2). However, a recent reevaluation of these results (Isachsen, personal communication, 1993), mostly correcting improper reduction techniques in the original analysis, suggest that the real rates may be more on the order of 1 mm/year. The results of smaller scale geodetic surveys by the State of New York along parts of the range front in the Glens Falls area (Tice, personal communication, 1992) suggest large lateral displacements of several mm/year, but it is difficult to assess the accuracy of these reports. Finally, the 100-year-old building housing the Saratoga Springs Police Department, which straddles one of the faults bounding the Adirondacks to the east, the Ballston-McGregor fault, recently had to have cracks in its foundation repaired (Tice, personal communication, 1992).
On the other hand, there have been no major (M>6) earthquakes along these faults in recent history, and trenching near the Ballston-McGregor fault failed to find conclusive evidence of large earthquakes in the past 20,000 years (Tice, 1993; note that this doesn't mean there weren't any, but that no positive evidence was found). Also, the few earthquakes in the Adirondacks for which we have focal mechanisms (e.g., Sbar and Sykes, 1977; Seeber and Armbruster, 1986; Nabelek and Suarez, 1987) show reverse faulting with P axes oriented WSW-ENE, a direction of maximum stress consistent with other shallow indicators (Zoback and Zoback, 1988). A stress field with a horizontal maximum stress axis is not at all consistent with uplift-related doming, which should produce vertical maximum compression. Clearly, strain accumulation in the Adirondack regions is, at best, poorly understood, and a reliable set of geodetic data is required to address properly the question of active deformation. The creation of this data set through a GPS campaign is one of the primary goals of this project.
A complementary approach to understanding the state of current deformation in the Adirondacks is to discern the dynamics of mountain building in this region. The question of why the Adirondacks are there at all may be examined by determining the structure of the crust and upper mantle beneath the mountains and which of competing hypotheses best corresponds to that structure. Based on the mapping of a zone of layered reflectors, Brown et al. (1983) suggest that the crystalline rock at the surface may have overthrust sedimentary rocks during collisional orogeny. Others (e.g., Isachsen, 1980) have suggested that the Adirondacks are underlain by a hot spot that causes uplift by local thermal expansion and by a low density body pushing up towards the surface. The circular and dome-like expression of the uplift, along with the pervasive expansion joints in the Adirondacks, provide gross structural support for this idea, as does the presence of hot springs along the boundary (e.g., at Saratoga Springs and Ballston Spa). The presence of a zone of high conductivity in the lower crust (Connerney et al., 1980; Connerney and Kuckes, 1980) provides tantalizing geophysical evidence of a thermal anomaly. At the same time, Owens (1987), in a study of teleseismic waveforms, claims that the high velocity zones in his model are inconsistent with an interpretation of partial melting, and that the zone of high conductivity may be caused by mineral hydration in the lower crust. However, the mechanism for emplacement of the required water is, as Brown et al. (1983) point out, problematical. Owens (1987) used data from one station (RSNY) located at the northern edge of the Adirondacks ( Figure A3), and, based on correlation with the results of COCORP lines (Klemperer et al., 1985; Brown et al., 1983), argued that his results could be generalized to the entire Adirondack region. We note, however, that this correlation is somewhat circumstantial in that it correlates velocity and reflectivity results. Moreover, the structure determined beneath RSNY appears strongly dependent on azimuth, suggesting a significant amount of lateral heterogeneity beneath the station. Therefore we feel that is not obvious that these results are necessarily representative of the entire region. In this project, we plan to deploy additional broad band stations in other locations in the Adirondacks, to see whether or not Owens (1987) results are indeed relevant to the entire region. In particular, a station located in the area where the COCORP lines and the electromagnetic studies of Connerney et al. (1980) were conducted (Figure A3) would provide a useful calibration for correlating reflectivity with shear wave velocity.
Part of the motivation for conducting the seismic study in the Adirondacks derives from the recent experience of one on the PI's (Roecker) in the analysis of data from the Tien Shan range in central Asia. At first blush, the Tien Shan would appear to have little relation to the Adirondacks, but in fact it may be an example of the same mountain building dynamics. Using data recorded at intermediate band seismic stations throughout the Tien Shan, Roecker and his colleagues (Makeyeva et al., 1992; Kosarev et al., 1993; Roecker et al., 1993), determined that the central part of the Tien Shan was underlain by a region of active upwelling, rather than by a lithospheric root as had been previously surmised. Specifically, they determined that the shear velocities in the crust and upper mantle in the region above the plume were anomalously low, and that the direction of flow in the mantle, determined from splitting of SKS, was locally contrary to the prevailing direction. Two of the techniques that they used, receiver function analysis (Kosarev et al., 1993) and SKS splitting (Makeyeva et al., 1992) are appropriate for addressing the question of possible upwelling beneath the Adirondacks and will be applicable to the data set we propose to collect.
This appendix gives examples of the two most common styles of triangulation data reduction sheets found among the Colvin archives, and from which we entered data into computer files to proceed with our analysis.
[1] Now at: Institute of Geological and Nuclear Sciences, P. O. Box 30-368, Lower Hutt, New Zealand.
[2] The coordinates were not fixed to their exact ITRF93 values due to an oversight. However, the error for subsequent deformation analysis is insignificant.
[3] Other processing details: Antennas were of two types (Table 1) but official IGS elevation-dependent antenna phase patterns were not used since our analysis was completed prior to the publication of those patterns. IGS final combination orbits in the ITRF93 reference frame were used, and values of polar motion and UT1-UTC were interpolated from the International Earth Rotation Service (IERS) Bulletin B listings.
[4]The solved coordinates for Algonquin and Westford are given in Table 2, and can be compared with the official IGS values if it is desired to see how close our coordinates are to ITRF values (they are within a few cm).