Tecronophysics.

130 (1986) 23-31 23

Elsevier Science Publishers B.V,, Amsterdam - Printed in The Netherlands

REVIEW OF THE HUAYTAPALLANA PROJECT IN PERU

ADAM CHRZANOWSKI r and WALTER WELSCH ’

’ D~F~~trne~tof Surveying E~~~ne~rin~, U~ic~ersi~~of New Brun~~~~k* Fredertctan.

N.B. WB SA3 (Canada)
of the Federal Armed Fwees, Munich, Werner- Heisenberg- Weg 39,
’ Depurtment nf Geodesy ~fli~~rs~~.v
D-809014Neubiberg (F. R. of Germanv)

(Accepted April 7, 1986)

ABSTRACT

Chrzanowski. A. and Welsch, W., 1986. Review of the Huaytapallana project in Peru. In: H.G.
Henneberg (Editor). Recent CrustaI Movements. 1985. ~e~?~~~~~.~sj~s, 130: 23-31.

In 1969. two consecutive earthquakes activated a reverse fault at an elevation of 4600 m in the
Huaytapallana mountain range near Huancayo in centrai Peru. In 1975, a small geodetic network was
established and measured across the fault covering an area of about 1.5 km’. The network has been
remeasured in 1976. 1977, 1978. and 1982. using standard geodetic instruments, as a joint effort by two
Canadian and one German universities in cooperation with the Peruvian Institute of Geophysics. ln
1978. the network was expanded to cover 6 km*. sigh-altitude sickness, logistic problems with old
vehicles, civil unrest and riots plagued the survey expeditions. The results show a cyclic rigid body
motion of the southwest side versus the northeast side of the fault in a general east-west direction of
about k 3 mm/year. The direction of the motion agrees with the direction of compressive forces
expected in this subduction region. The next survey campaign is planned for 1987.

INTRODUCTION

The Huaytapallana project is named after the Huaytapaliana mountain range in

the high Andes of central Peru where, since 1975, a small geodetic network has been
repeatedly observed as a joint effort by the University of New Brunswick (UN@
and the University of Alberta (U of A) in Canada and the University of the Federat
Armed Forces (UFAF) in Munich, West Germany, in cooperation with the Pe-
ruvian Institute of Geophysics (IGP).
The network spans across a reverse sinistral fault at an elevation of 4600 m about
20 km northeast of the town of Huancayo (Fig. 1) where IGP has a research station.
Initially, the network consisted of only 11 points (point i-II in Fig. 2) covering an
area of 1.5 km2. In 1978 the network was expanded (points 20-27) to cover about 6
km”. A total of five survey campaigns have been performed to date, with the first

e 1986 Elsevier Science Publishers B.Y.

Fig. 1. Location of the epicentre of the 1969 earthquakes (Deza, 1971).

three survey expeditions organized by UNB and U of A in 1975, 1976, and 197’7.
The last two surveys were organized by UFAF in 1978 and 1982. The next
campaign is planned by the authors for 1987.
Due to financial limitations and other reasons explained in the next section, only
standard geodetic instruments and observation techniques have been used in the
angle and distance measurements aiming at detecting horizontal strains of 10es5 and
relative displacements of 5 mm at the 0.95 confidence level.
The area of the project is seismically very active and is of a particular interest to
geophysicists because the Huaytapallana fault is one of only two reverse faults in
the high Andes known to have been broken by an earthquake (G. Suarez, written
pers. commun., 1983). The Huaytapalla~a fault was activated by two consecutive
earthquakes on 24 July and 1 Cktober, 1969 (Deza, 1971), and the other fault is
associated with the 1944 and 1979 San Juan earthquakes in central Argentina. Both
faults occur where seismic oceanic ridges are being subducted: the Nazca Ridge in
Peru and the Juan Fernandez Ridge in central Argentina.
Deza (1971), who gave a detailed account of the seismic events of 1969, reported
that the July earthquake (M = 56) produced an uplift of the ground of about 40 cm
to the northeast of the fault. The-ground displacements increased to about 1.6 m in
the vertical and 0.7 m in the horizontal directions (left-lateral strike-slip motion)
Fig. 2. Location of the survey points

after the October shock which produced the ground rupture over a total length of
about 3.5 km with the strike veering between N115”E and N130”E. The total
estimated length of the fault is between 12 and 20 km (G. Suarez, written pers.
commun., 1983). Philip and Megard (1977) give a detailed structural analysis of the
superficial deformations.
The aseismic behaviour of reverse faults in subducted areas is not well known.
Therefore, it was hoped that the Huaytapallana survey project would contribute to a
better understanding of the mechanism of the earth’s crustal movement in that
region of the high Andes.
Technical details of the surveys and preliminary evaluation of the results have
already been presented in several papers by various authors, for example.
Chrzanowski et al. (1978), Nyland et al. (1979) Welsch (1979). Welsch et al. (1983).
and Chrzanowski and Secord (1983). Several graduate students, such as Dennler
(1979) and Margrave (1980) have been involved in the project and have used the
material in their theses. The results have also been used by the international ad hoc
committee (Chrzanowski et al., 1981) of the Federation Internationale des Gtometres
(FIG) to compare different approaches to the analysis of deformation surveys. So,
26

the Huaytapallana project, though small and of a moderate accuracy of measure-

ments, has become well known as a unique project of its kind in South America.
Therefore, on the occasion of the International Symposium on Recent Crustal
Movements in South America, which was held in Maracaibo in 1985, the authors
were invited to give a review of the project with emphasis on difficulties and
problems encountered by the survey expeditions. The high-altitude sickness, per-
sistent fog, logistic problems with ancient vehicles, gastric disorders, civil unrests
and riots, are listed in the diaries of the leaders of expeditions as the major
problems. The technical difficulties of rather simple measurements seem to be
secondary.

CHRONOLOGICAL ACCOUNT OF EVENTS

In 1973, Dr. Fdo Nyland, of the Institute of Earth and Planetary Physics of U of
A in Canada, developed contact with a seismology group at IGP in Lima, Peru.
Hillary Hennessey, Daniel Huaco and Ernest0 Deza were among the IGP group
which expressed an interest in developing a joint project, together with their
Canadian colleagues, to initiate a program of geodetic monitoring of active faults,
particularly the Huaytapallana fault, and a normal fault in the Cordillera Blanca
near the small settlement of Pitec about 20 km from the town of Huaraz. The area
in and around Huaraz had suffered catastrophic damage and over 50,000 casualties
in a major earthquake and subsequent landslide in 1970.
In 1974, Edo Nyland, provided with a small research grant from the Develop-
ment Research CounciI of Canada and from U of A, arrived in Lima, Peru, to make
a reconnaissance for the monitoring surveys. The area of Pitec could not be visited
and the reconnaissance was limited to the Huaytapallana area. This preliminary
investigation indicated that a small geodetic network could be established across the
fault. IGP offered accommodation for the survey crew in their research station near
Huancayo, at an elevation of about 3000 m. from where the crew could reach the
project area. Edgar Camargo, a seismologist from the IGP station, helped in the
reconnaissance and became a member of all subsequent survey expeditions.
Edo Nyland, upon his return to Canada, contacted Adam Chrzanowski, of the
Department of Surveying Engineering at UN& to help in designing and establishing
the-monitoring network.
As mentioned earlier, the network was to be designed to detect strains of 10e5. or
smaller, over distances of a few hundred m&es using light-weight, sturdy and
comparatively inexpensive instruments so that IGP could afford to take over the
monitoring surveys after the UNB team had supervised the initial two or three
campaigns.
In designing the network, priority had to be given to the stable monumemation
of the survey stations in rock outcrops with in~~ibility bfzEween Ehe statiuns in the
rugged terrain. Therefore, the preliminary “ideal” design of the network (Chrzanow-
 0 0.5 km

Fig. 3. Huaytapallana monitoring network.

ski et al., 1978) had to be replaced with the “practical” configuration of the
network, as shown in Fig. 3, with angles and distances designed to be measured with
standard deviations of 1.5” and 5 mm, respectively, and with over 100 degrees of
freedom.
In August 19’75, a small survey team consisting of Peter Polak from UNB, and
Peter Chrzanowski established the network with brass markers cemented in the
rocks. Edgar Camargo of IGP helped in the surveys. A Wild T2 theodolite and an
HP3800 infrared distance meter were used in the first and two subsequent survey
campaigns. The advantage to working at that high altitude in comparison with sea
level is that there is about a twice smaller influence of errors in the temperature
measurements on the accuracy of distance reductions.
A year later, a larger UNB team, consisting of Adam Balut and Robin Steeves,
with help from Peter Chrzanowski and Edgar Camargo, resurveyed the network
using the same survey instruments.
In 1977, Walter Welsch of UFAF in Munich was a visiting scientist at UNB and
led the third survey expedition with Gary Margrave from U of A, Peter Chrzanow-
ski and Edgar Camargo. Walter and Gary made a reconnaissance and established a
previously planned microgeodetic network in the Pitec area in cooperation with the
Peruvian Institute of Mining and Geology.
The area of the planned survey appeared to be more difficult than that in
Huaytapallana. The survey stations had to be established on a steep slope at an
altitude of 4000 m with elevation differences of up to 500 m. The first survey of the
Pitec network (Fig. 4) was made a year later by a survey team from UNB (Mark
Dennler and Dave Lehman).
At the same time, another team from Germany fK. Blachnitzky and W. Kltiker),
with Edgar Camargo, conducted the 1978 survey at Huaytapallana and expanded
n=307
u=24
df=293
9

0 0.5 IKIll

Fig. 4. Pita monitoring network.

the network by stations 20 to 27 (Fig. 2). The AGA 14A electro-optical instrument
replaced the HP3800 of the previous expeditions in distance measurements.
In 1979, the Pitec survey was repeated by Mark Dennler with Julio Leal and
Pablo Romero from UNB. Details of the Pitec expedition are given in Dennler
(1979).
The Huaytapallana surveys were repeated once more in 1982 by the UFAF group
(H. Heister and R. Baumer with E. Camargo from IGP). The 1982 expedition used
an HP3820 total station for distance and angle measurements. A three-dimensional
approach was initiated in the I982 surveys by including more accurate than before
measurements of vertical angles and precise measurements (with a specially design-
ed device) of the heights of the instruments. Due to the difficult topography, the
geometrical levelling which was tried in the first survey in 1975 had to be aban-
doned as uneconomical.

EVALUATION OF RESULTS AND CONCLUSIONS

A detailed evaluation of the observation data for the first four epochs of
measurements has been given by Dennler (1979) and summarized in Chrzanowski et
al. (1983). The 1982 campaign has been evaluated by Welsch et al. (1983). The
estimated standard deviations are listed in Table 1.
The overall achieved accuracy of angles and directions is worse than the designed
1.5” (for angles measured in three sets with relevelling of the instrument between
the sets). The generally poor visibility, fog, rain, snow, sudden changes in the air
temperature and fluctuations of the atmospheric refractionexplain the poorer than
expected results. However, the distance measurements, except the 1978 campaign
(AGA 14A instrument), gave better than expected accuracy. Thanks to the large
redundancy (degrees of freedom), the overall a posteriori accuracy of the monitoring
 29

TABLE 1
Estimated standard deviations of observations.

1975 1976 1977 1978 1982

Distances (mm) 4.0 3.5 2.1 5.5 2.9
Angles (“) 2.6 2.2 _ _ 1.8
Directions (“) _ _ 2.8 2.5 _
Zenith angles ( “) (for distance reductions only) 4.0
Degrees of freedom 114 145 128 124 134

network has become better than expected from the pre-analysis. Strain components
and relative rigid body motions of 3 x 10m6 and 3 mm, respectively, could be
detected at the 0.95 confidence level.
As was mentioned earlier, the survey data has been subjected to various ap-
proaches to the geometrical analysis of deformation surveys within the activity of
the FIG committee. A comparison of some of the results of the first four epochs
(from 1975 to 1978) has been given by Chrzanowski and Secord (1983). The analysis
of the results is not easy because the deformations, if any, have been marginally
small (at the border line of the statistical significance).
The UNB group has used their generalized method (Chrzanowski et al., 1983:
Chen, 1983) in the geometrical analysis. The method is based on a trend analysis of
a datum independent display of displacements and selection of such deformation
model which fits the observation data in the best (statistically) way. Several models
have been tested including the no-deformation, homogeneous strain, and a few
other models. Finally, the model in which points 4 and 11 (Fig. 3) move as one rigid
block (Block A in Fig. 5) with respect to all the remaining survey stations (Block B)
treated as a stable block has been accepted as the best model. The detected
displacements of Block A has been determined with confidence levels larger than
0.95 (probability > 95%). This means that either points 4 and 11 undergo superfi-
cial movements in the same (statistically) direction and with the same magnitude,
which would be very coincidental, or there is a tectonic discontinuity branching off
from the main fault somewhere between stations II and 1, and the main fault is
locked south of that point. The second alternative is supported by G. Suarez
(written pers. commun., 1983) who gives evidence that the whole region under
investigation is being deformed by a horizontal compressive stress oriented roughly
east-west, in the direction of plate motion. The graphical display (Fig. 5) of the
movements of Block A clearly shows that the general direction of the detected
relative displacements, though of a cyclic nature (compressive and tensile), is also
east-west. Hopefully, the planned resurvey of the extended network will add more
light to the interpretation of the results.
The authors realize that the small aperture geodetic networks are not best suited
30

1975-1976
N

9 L___-______,

d=3.3 mm (Pr.99%)
1

d=O.3 nun (~r.4%)

;
T7
B

-------1

J1
_~
IA
I
L-----

~~
d=6.2m (Pr.99%)
d=7.1 mm (Pr.99%)

0 amm
I ’ ’ 1 ,

d= 4.6 mm (Pr.99%)

Fig. 5. Detected displacements and their probabilities (Pr %) of Block A (points 4 and 11) with respect to
remaining points (Block B) of the monitoring network.

for investigations of tectonic movements because the whole area near the fault is
non-homogeneous and any deduced strain is nothing else but closing or opening
local discontinuities. In addition, any small local instabilities of survey markers may
lead to wrong interpretations. On the other hand, small aperture networks may
“see” the microtectonic movements better than the networks with long distances. In
any case, a compromise must be reached between whaf is economically possible and
what is most suitable. The use of the new space techniques, particularly the G&&al
Positioning System (CM), will perhaps give more alternatives for the compromise
solutions within the next few years, as soon as the GPS receivers become small and
light enough to be carried at the high elevations.
 31

ACKNOWLEDGEMENTS

The reviewed project has been modestly financed by all three universities
involved, by the Natural Sciences and Engineering Research Council of Canada, by
the Deutsche Forschungsgemeinschaft, and by the Instituto Geofisico de1 Peru.

REFERENCES

Chen. Y.Q.. 1983. Analysis of deformation surveys-a generalized method, Dep. Surv. Eng. Tech. Rep.,
94-Univ. New Brunswick, Fredericton.
Chnanowski. A. and Secord, J., 1983. Report of the ‘ad hoc’ committee on the analysis of deformation
surveys. FIG 17th Int. Congr., Vama. Pap. No. 605.2.
Chrzanowski, A., Nyland, E., Dennler, M. and Szostak. A., 1978. Microgeodetic networks in monitoring
tectonic movements. In: Proceedings of the International Symposium on Deformation Measurements
by Geodetic Methods. Wittwer, Stuttgart.
Chrzanowski, A. with contributions by members of the FIG ad hoc committee. 1981. A comparison of
different approaches into the analysis of deformation measurements. FIG 16th Int. Congr. Montreux.
Pap. No. 602.3.
Chrzanowski. A., Chen, Y.Q. and Secord, J., 1983. On the strain analysis of tectonic movements using
fault crossing geodetic surveys. Tectonophysics, 97: 297-315.
Dennler. M., 1979. Evaluation of microgeodetic networks for monitoring tectonic movements in Peru. M.
Eng. thesis, Dep. Surv. Eng.. Univ. New Brunswick, Fredericton.
Deza, E., 1971. The Pa~ahuanca earthquakes, Huancayo, Peru: July-October 1969. In: Recent Crustal
Movements. Bull. R. Sot. N.Z., 9: 77-83.
Margrave. G.F., 1980. Microgeodesy and South American tectonics. Ph.D. diss., Univ. Alberta, Edmon-
ton.
Nyland. E., Chrzanowski, A., Deza, E., Margrave, G., Dennler, M. and Szostak. A., 1979. Measurement
and analysis of ground movement using microgeodetic networks on active faults. Geofis. Int. (Mex.).
18( 1): 53-71.
Philip. H. and Megard, F., 1977. Structural analysis of the superficial deformation of the 1969
Pariahuanca earthquakes (central Peru). Tectonophysics, 38: 259-278.
Welsch. W.. 1979. A geodetic micro-network for monitoring tectonic movements in a Peruvian earth-
quake fault zone. In: A. Vogel (Editor), Terrestrial and Space Techniques in Earthquake Prediction
Research, Proc. Int. Workshop on Monitoring Crustal Dynamics in Earthquake Zones, Braun-
schweig/Wiesbaden, pp. 531-542.
Welsch. W., Heister. H. and Baumer. R., 1983. Geodetic monitoring of an active fauit in the Peruvian
Andes. The Huaytapallana fault 1975-1982. Zentralbl. Geol. Palaontol., 1(3/4): 363-374.