Local site effect characterization and determination of the shallow high impedance layerin downtown area of Managua

Abstract:

Managua city is the capital of Nicaragua
and has the most dense population of the country. It
is settled in the Pacific region of the country where the largest
seismic hazard exists. It is agreed by the research community,
regarding site response assessment, that dynamic properties of
the soils are one of the key factors, that increases the damage
an earthquake can cause in a given site. In Managua city, dynamic
properties of the soils has not yet been characterized to
detail.

In this study, fundamental periods and an
approximate thickness from the shallow material to
the high impedance layer were estimated. Fundamental period was
estimated based on microtremor measurements in 26 sites in the
old downtown are of Managua city and the Nakamura method, whereas
the thickness was estimated based on the modeling of
the empirical transfer function of microtremors.

Results led to two conclusions. First, the
fundamental resonance occurs in the period range of 0.18 – 0.22 s
in the old downtown area of Managua city. A second mode of
vibration seems to ocurrs in the range of ?? to ?? s. An increase
of the fundamental period from North to South was
observed.

Second, two main high impedance interfaces
seem to be detected by the H/V ratio, the first high impedance
interface show its larger amplitude around 0.22 s and seems to
occur to a depth of 13 m where as the second impedance interfaces
show its larger amplitude in ?? second and extends
up to ?? s. This results are very important as in the city most
of the constructions are one and two story houses
and medium rise buildings.

1.
INTRODUCTION

Nicaragua is an earthquake prone country.
This occurs as consequence of the continuous
subduction process of the Coco's plate beneath the
Caribbean plate. Additional to this seismic source there is the
existence of inland active shallow faulting. This faulting have
historically generated earthquakes with devastating consequences
for different cities in the country.

Managua is the capital city of Nicaragua,
the population as of 2012 were of 1028808 million people which
represents a sixth part of the total population of the country
(INIDE 2012). The city is located in the Pacific
region of Nicaragua. In this region there are two main tectonic
features. One is the volcanic chain and the other one is the
Nicaraguan depression (Frez and Gamez 2008; Figure 1). Both
geological features cross the country with an approximate
direction of N45W. The top layers of soil in the Nicaraguan
depression, are constituted mainly by volcanic soils. and shallow
faulting with strike-slip type faults is common in the depression
(INETER 2004, McBirney and Williams 1965).

Historically is known that Managua city
have been struck by earthquakes in 1931, 1968 and 1972 (Leeds
1974). The last one caused many collapses of buildings and
dwellings in the downtown area of the city in that epoch (Brown
1974). Moreover, paleo- seismological studies have been carried
out in the eastern part of the city, specifically in the eastern
edge of the zone known as the airport Graben, researchers found
that this faulting zone have generated earthquakes in the past,
being the most recent between the period of A.D. 1650-1810 (Cowan
et al. 2002).

One of the factors that affect the most the
damaging potential of seismic waves is the local
site conditions. This fact, have been learned from several
examples around the world being one of the most infamous the 1985
Michoacan earthquake in Mexico, where the epicenter of the
earthquake was located more than 300 km away of Mexico city and
yet this earthquake caused devastation in buildings and
dwellings. Such local response caused by the local site
conditions is known as site effect (Kagawa 1996).

A research aimed to investigate site effect
potential in the shallow soil deposits in Managua
city was carried out. This was done as part of the cooperation
program between the Institute of Geology and Geophysics (IGG)
which belongs to the National Autonomous University of Nicaragua
(UNAN-Managua) and the International Institute of
Seismology and Earthquake Engineering (IISEE) which belongs to
the Building Research Institute (BRI).

In this document preliminary results which
consists in determining site effect in the old
downtown area of Managua city are presented.

2. SEISMIC
ACTIVITY

Potential damaging earthquakes in Managua
city are generated due to three main seismic sources. These
sources can be divided into three functional types:

1. Interplate earthquakes along the
subduction zone. Seismic activity is generated due to the
interaction of the plates along the subduction zone. According to
the earthquake catalog compiled for the seismic hazard assessment
of Nicaragua, this is the most seismically active area (Resis II
2008). This seismicity occurs along the Pacific coast of
Nicaragua (Figure 3). A major earthquake with Mw=7.6, occurred in
1992 in this zone, this earthquake generated a Tsunami which
affected Masachapa town and some other nearby towns (Resis II
2008).

2. Intraslab earthquakes in the subducting
Cocos plate.

Earthquakes in this zone are also related
to the seismicity generated as a consequence of the subduction of
the Cocos plate beneath the Caribbean plate. However it also
includes seismicity from the Forearc zone of Nicaragua
(Güendel and Protti 1998). In this zone a fault system with
NE-SW strike predominates, located along the Pacific coast (Funk
et al. 2009; Segura et al. 1996). The biggest
historical earthquake in this zone occurred in 1956 and it had a
magnitude of Mw=7.3 (Resis II 2008).

Figure 1. Map showing the surroundings of
the study area (left). The right map shows the geomorphologic
units nearby the urban area of Managua city (relief data from
Farr 2007). The pink polygon in the right plot represents the
study area whereas the white lines represent the current urban
area of Managua city.

Figure 2. Left figure shows the geological
map of Nicaragua and geological faults around Managua city.
(INETER 2004). As can be seen from the figure the surface
materials in the city corresponds to volcanic
materials.

3. The third seismic source are the inland
earthquakes in and nearby Managua city. Historic experience
indicates that inland earthquakes are often highly devastating.
These inland earthquakes are mainly generated due to shallow
faulting systems (LaFemina et al. 2002). These
faults release only a small amount of the energy generated
through daily seismic activity. As consequence they
accumulate high amounts of stresses and are then able to generate
earthquakes of up to Mw =6.3 as experienced in 1972 (Brown et
al. 1974).

Figure 3. Earthquakes epicenters (filled
dots) in the period 1990-2007(INETER 2007) and
geological faults (lines) around the study area.

3. STUDY
AREA

The study area is the downtown area of
Managua city during the 1972 Managua earthquake. This area is
located in the Central-Northern part of the city between
the coordinates 0577500 E – 1344250 N and 0581000 E-
1341000 N , and it has a smooth slope towards the
North.

Figure 4. The map shows the study area
enclosed in a gray rectangle and the location of the 26 points
where microtremors data were recorded.

Managua city is settled within the
Nicaraguan depression, geologists have defined the lithology of
Managua soils and nearby Managua city (Figure 2 left and Figure
3). Some historical earthquakes have also occurred in this area
(Resis II 2008)

In a regional scale, Managua as all of the
Pacific region cities is located within a major geological
feature in the shape of a fringe along the Pacific coast of
Nicaragua named as the Nicaraguan depression. This is a tectonic
structure that lays in the Pacific region of the country with an
approximate direction of N45W and an approximate width of 75
km.

Two normal faults limits the depression
thus is classified as a Graben (INETER 2004). In a local scale,
Managua city is settled in a smaller Graben named as Graben of
Managua.

Geomorphologically, the Urban area of
Managua is surrounded by many volcanic edifices. It is limited in
the Northwest by Chiltepe volcanic complex, in the Southwest by
the Mateare scarp, in the Southeast by Masaya volcanic complex
and in the Northeast the Tipitapa plain (Hradecky et al. 2000).
To the South lays an scarp known as Las Nubes scarp, thus in
general, the city have a slope towards the North which becomes
smoother from the southern border of the city (Figure
1).

Lithologically, Managua city lays over
volcanic soils mostly pirocalstic which comprises scorias, ashes,
tuffs, pumice, lahars, ignimbrites, colluvials, paleosoils and
meteorized material. All these materials lays discordantly over
ignimbrites known as Grupo Las Sierras (Hradecky et al.
2000).

4.
METHOD

The main goal of this research was to
determine whether Managua city soils present local site effects
or not based on the fundamental period of vibration of the soils
as well as in the relative amplifications. The
method used in this study was the HVSR or Nakamura technique as
was popularized most by this author (Nakamura 1989). The main
assumption of this method is that we can use the vertical
component of a record as if this component does not have
amplification in it, this is, as if it was a component of a
statio n located in a rock site.

Microtremors has been acknowledged as being
able to characterized the fundamental vibration period of a site
with reliability by the scientific community whereas in the case
of the relative amplification this method is not totally accurate
and still some debate is going on (Bonnefoy-Claudet et al. 2006).
The use of microtremors for site effect assessment was initiated
in Japan (Nogoshi and Igarashi 1971) and most popularized by
Nakamura (1989).

The capability of microtremors for
determining fundamental periods of the soils accurately, comes
from the fact that a high contrast boundary between soil and
bedrock is clearly detected in the frequency spectrum with a
peak, this has been studied numerically and analytically by some
researchers (Uebayashi et al. 2012; Bonnefoy-Claudet et
al. 2006).

Eventhough it has been pointed out the weak
accuracy of the method to determine relative
amplifications, in this study, it was decided to show the
amplitudes of the HVSR, so we can have a rough idea of what kind
of amplification values we could expect in the soils of
Managua.

In order to evaluate the fundamental period
of vibration and relative amplitudes the analysis of
26 microtremor records based on the H/V ratio was performed. The
duration of the records was fixed to approximately 30 min and
data acquisition was carried out at selected sites in the urban
area of the city.

5. DATA
ACQUISITION

The data acquisition had as target to
determine short period and long period microtremors.
Thus we recorded 30 min long records at 26 points distributed in
the city (Figure 4).

A Mcseis MT Neo instrument was used as
datalogger (OYO corp.) with a sampling frequency of 100 sps and
three long period seismometers (Tokyo Sokushin) with a bandwith
of 0.2Hz to 100Hz (Figure 6). The time window for analysis was of
87 s (Figure 5 right). The length was defined by keeping only,
the windows with less cultural noise, this was done by using and
sta/lta algorithm. Moreover, In order to select the less noisy
part we checked the deviation of the average HVSR with respect to
+/- one standard deviation.

Figure 5. Right figure shows the
microtremor record with a part highlighted indicating
the portion to analyze and the left figure shows the
highlighted part of the right figure in a zoom in.

Figure 6. Frequency response of the long
period sensors.

Figure 7. Flow chart of the steps followed
in the analysis of the microtemors records.

5.1 Data processing

The data acquired was first converted from
binary to ascii. As mentioned the analysis was based
on the HVSR technique, this procedure was applied to each record
to determine the FFT and the H/V ratio. The analysis was
performed based on the flowchart shown in Figure 7.

The procedure could be summarized as
follows: in the time domain, we tapered the record
by applying a sinusiodal function increasing from 0 to the
maximum amplitude in the first 10 percent of the time window
selected and then decreasing from the maximum amplitude to 0 in
the last 10 percent of the selected window. If the time window
was not a power of two, zero padding was applied until this was a
power of 2.

Then we applied the Fast Fourier Transform
and smoothed the record in the frequency domain by applying the
Parzen window. Then we obtained the H over V spectral ratio by
taking the square root of the sum of the squared horizontal
components spectra and divided by the vertical component
spectrum.

Result was taken as good or bad based on
the deviation of +/- one standard deviation from the average HVSR
curve.

6. RESULTS AND
ANALYSIS

A total of 31 ambient vibrations records
were collected in old downtown area of Managua
city.

Figure 8. H/V ratio curves grouped in three
types according to its frequency content. The information shown
in the microzonation maps was obtained from the H/V ratio
curves. The curves were grouped according to its
characteristics in three families (see Figure 8), this is,
according to the frequency content of each record. Moreover
zonations are proposed based on the periods and on the relative
amplitudes of HVSR.

The zonation maps were plotted taking into
account the spatial distribution of points and based on the range
shown in the legend we tried to plot the "best" involving
contours to our criteria.

The proposed zonations, based on the
Fundamental periods of HVSR and relative amplitudes
of HVSR are detailed below:

6.1 Fundamental periods of
HVSR

The proposed zonation is based on
fundamental periods of HVSR, this periods were divided in three
categories 0.10 – 0.18 s; 0.18 – 0.22 s; 0.22 – 0.37 s (Figure
9).

These ranges were chosen by taking bins of
equal value and according to the spat ial
distribution of the fundamental periods of HVSR. This
criterion was chosen in order to avoid a too complex zonation
map.

Zone 1: The soils in this zone have
fundamental periods between 0.10 s to 0.18 s. Taking into account
the fundamental periods observed in this zone and the Lithology
information collected by ENACAL (2004) we could infer that in
this part of the city the shallow soils have a smaller thickness
compared to the other parts of the area. In this zone, it is
important to avoid the coincidence of the natural period of
vibration of buildings with 1 to 2 stories height
with that of the soil.

Zone 2: This zone have fundamental
periods between 0.18 s – 0.22 s. Taking into account the
fundamental periods observed in this zone and the Lithology data
collected by ENACAL (2004) we could infer that in this part of
the city the thickness of the shallow soils increase outwards the
diameter of the Tiscapa Lagoon. In this zone, it is important to
avoid the coincidence of the natural period of vibration of
buildings with 2 to 3 stories height with that of the
soil.

Zone 3: This zone have fundamental
periods between 0.22 s – 0.37 s. Taking into account the
fundamental periods observed in this zone and the lithology data
collected by ENACAL (2004) we could infer that in this part of
the city the thickness of the soils of fine grain it is a bit
larger than the other 2 zones. It is important to avoid the
coincidence of the natural period of vibration of buildings with
2 to 4 stories height with that of the soil.

Figure 9. Zonation based on the fundamental
periods.

6.2 Amplitudes of HVSR

The relative amplitudes of HVSR observed in
the old urban area of Managua city range from a minimum of 1.7
times to a maximum of 5.2 times (Figure 10). Predominating
the values in the range of 3.9 to 5.3. In this study
relative amplification is shown for reference purposes as the
HVSR cannot determine accurately the soil amplification.
The zonation presented here, can be explained as
follows:

Zone 1: This zone presents HVSR
amplitudes in the range of 1.7 to 3.0 times. This amplitude
values can be explained as the change in the thickness of the
lithology towards the Tiscapa Lagoon, since the soil layers with
big impedance contrast reduce its thickness, thus
the amplitudes reduces.

Figure 10. Zonation based on the relative
amplification.

Zone 2: This zone presents HVSR
amplitudes in the range of 3.0 to 3.95 times. This amplitude
values can be explained similarly as the last zone, this is, the
high impedane layers increase their thickness outwards the
diameter of the Tiscapa lagoon, thus the relative amplitudes
increases its values.

Zone 3: This zone presents HVSR
amplitudes in the range of 3.95 to 5.17 times. This amplitude
values follow the same behavior as Zone 2. This is, as the high
impedance contrast layers increases in thickness so does the
amplitudes values.

6.3 Soils thickness

Soil model was obtained here based on Arai
and Tokimatsu (2005). According to these authors the soil
empirical transfer function can be modeled by means of fitting
the theoretical phase velocity of Rayleigh
waves.

By following the above mentioned procedure
we were able to model a soil structure which show two main
impedance contrasts, one around 0.25 s and the other one from 0.6
to 2 s. These result indicates that although the fundamental
period of vibration of Managua soils is of 0.25 s,
there is another vibration period which influences the dynamic
behavior of the soils and is in the range of 0.6 to 2 s (Figure
11).

Figure 11. Soil modeling using H/V
empirical transfer function.

The soil model we obtained by modeling is
shown in the table below:

Table 1. Soil structure obtained
from H/V empirical transfer function.

The Nicaraguan national building code (RNC
07) classifies the soils in four groups according to shear waves
velocities as follows:

TYPE I: Outcrop with Vs > 750
m/s,

TYPE II: Stiff soil with 360 < Vs
<=750 m/s,

TYPE III: Moderately soft soil, with 180
<= Vs <= 360 m/s

TYPE IV: Very soft soil, with Vs < 180
m/s.

From the previous results, it appears that
the engineering basement in Managua city is located to an
approximate depth range of 13 to 37 m.

7.
CONCLUSIONS

Based on the analysis carried out in this
study we can summarize the most important findings
in the following conclusions:

Two types of zonations are proposed in this
study, one is based on fundamental periods of HVSR
and the other based on amplitudes of HVSR. The fundamental period
zonation was proposed by dividing the vertical to horizontal
spectra in three categories 0.10 to 0.18 s; 0.18 –
0.22 s; 0.22 – 0.37 s.

Regarding relative amplifications two zones
were identified, the first zone have amplifications
that ranges between 1.7 to 3.0 times; the second zone show values
in the range between 3.0 to 4.0 times; and the third zone show
values from 4.0 to 5.2 times. Managua soils show moderate soft
soil in the shallow layes (depth < 10 m) and then the rigidity
increases in such a way that can be considered as stiff soil
according to the Nicaraguan building code soil classification.
Despite the fact that the fundamental periods changes as we moved
towards the southern part of the study area, it is clear that
from the fundamental periods of HVSR that the soils of the old
downtown area of Managua city shows a similar lithological
composition.. The difference of fundamental periods
observed in the city can be interpreted as a consequence of
the thickness of the shallow material rather than as a change in
the material.

As exposed by the fundamental period
results, Managua shallow soil response coincide s with the
vibration mode of one, two and three stories houses. Thus it is
important for the authorities to take into account
the information found here to better design the dwellings and
buildings in the future.

Finally, two main high impedance interfaces
seem to be detected by the H/V ratio, the first high impedance
show its larger amplitude around 0.22 s and seems to occurs to
a depth of 13 m whereas the second impedance
interface show its larger amplitude in ?? seconds
and seems to occur to a depth of ?? m. This results are very
important as in the city most of the constructions
are one story houses and medium rise buildings. Moreover to date,
the highest building in Nicaragua, formerly known as the BAMER
building have 16 stories which is in the range of
the second mode detected here. Thus is very important to assess
in further studies whether this building is affected by this
second mode or not.

The results seems to be in agreement with
the geology and geotechnical information of the study area. From
the results, it appears that two interfaces of high impedance
was detected in the spectra of the records. Based on
these results it appears that the engineering basement in the old
downtown area of Managua is located to an approximate depth
interval of ?? to ?? m.

8.
ACKNOWLEDGEMENTS

The first author would like to thank to his
former students Julio César Muñoz, Yoel
Morales and Stanly Pérez for their help in field
measurements.

The instruments used for data acquisition
were donated by the Japan International Cooperation Agency (JICA)
and facilitated by the Instituto de Geologia y Geofisica (IGG)
from National Autonomous University of Nicaragua
(UNAN).

9.
REFERENCES

Arai H., Tokimatsu K., 2005. S-wave
velocity profiling by joint inversion of Microtremor dispersion
curve and Horizontal to Vertical (H/V) Spectrum. Bull. Seismol.
Soc. Am., Vol. 95., No. 5, pp. 1766-1778.

Bonnefoy-Claudet, S., Cornou, C., Bard,
Pierre-Yves, Cotton, F., Moczo, P., Kristek, J., Fäh, D.
2006. H/V ratio: a tool for site effects evaluation. Results from
1-D noise simulations. Geophys. J. Int. Vol 167., pp.
827-837.

Brown, R. D., Ward, P. L. and Plafker G., 1974. Geologic
and Seismologic Aspects of the Managua, Nicaragua, Earthquakes of
December 23, 1972. Bull. Seismol. Soc. Am., Vol. 64., No. 4, pp.
1031.

Cailleau, B., LaFemina, P. C. and Dixon, T.
H. (2007) Stress accumulation between volcanoes: an explanation
for intra-arc erthquakes in Nicaragua?. Geophys. J. Int.,
Vol. 169. No. 3, pp. 1132-1138.

Cowan H., Prentice C., Pantosti D., de
Martini p., Strauch W. and workshop participants,
2002. Late Holocene earthquakes on the Aeropuerto fault,
Managua, Nicaragua.

Farr, T. G., et al. (2007), The Shuttle
Radar Topography Mission, Rev. Geophys., 45, RG2004,
doi:10.1029/2005RG000183.

Frez, J., Gamez, E., 2008. Aspectos de la
sismotectónica de Nicaragua y su alrededor. GEOS, vol. 28,
No. 3.

Funk, J., Mann, P., McIntosh, K. and
Stephens, J., 2009. Cenozoic tectonics of the Nicaraguan
depression, Nicaragua, and Median trough, El Salvador, based on
seismic- reflection profiling and remote-sensing data. Bull.
Geol. Soc. Am., Vol. 121, No. 11, pp.
1491-1521.

Guendel, F. and Protti, M., 1998.
Sismicidad y sismotectónica de América Central.
Física de la Tierra. No. 10, pp. 19-51.

Hradecky P., Havlicek P., Navarro M.,
Novák Z., Staník E., Sebesta J., 2000. Estudio para
el reconocimiento de la amenaza geologica en el area de Managua,
Nicaragua.

INETER, 2007. Boletínes
sismológicos. http://www.ineter.gob.ni. accessed in
January 2014.

INETER, 2004. Mitigación de
Geo-riesgos. INIDE 2012. Censo poblacional.

Uebayashi H., Kawabe H. and Kawabe K.,
2012. Reproduction of microseism H/V spectral features using a
three-dimensional complex topographical model of the
sediment-bedrock interface in the Osaka sedimentary basin.
Geophysical Journal International, Vol. 189, pp.
1060-1074.

Kagawa, T., (1996), Estimation of velocity
structures beneath Mexico city using microtremor array data,
Eleventh World Conf. on Earthq. Eng. No 1179.

Kinemetrics Inc. Q330 Operator
manual.

Kuang J. 1971. Geología de la Costa
del Pacífico de Nicaragua. Informe # 3 (Geology of the
Pacific Coast of Nicaragua. Report #3). Managua,
Nicaragua.

Kudo, K., Sawada, Y., Horike, M., 2004.
Current studies in Japan on H/V and phase velocity dispersion of
microtremors for site characterization. Thirteen WCEE, Vancouver,
B.C., Canada.

Lafemina P. C., Dixon T. H. and Strauch W.,
2002. Bookshelf faulting in Nicaragua. Geological Society of
America Bulletin. Vol. 30.

Leeds, D., 1974. Catalog of Nicaraguan
earthquakes. Bulletin of the Seismological Society
of America. Vol. 64, No. 4, pp. 1135-1158.

McBirney A. and Williams N.S. 1965.
Volcanic History of Nicaragua. University of
California Press.

Nakamura, Y., 1989. A method for dynamic
characteristic estimation of subsurface using microtremor on the
ground surface. Q Rep. RTRI, 30(1), pp. 25-33.

Nogoshi, M. & Igarashi, T., 1971. On
the amplitude characteristics of microtremor (part
2), J. Seismol. Soc. Japan, 24, 26–40. (In Japanese
with English abstract)

Resis II, 2008. Evaluación de la Amenaza
Sísmica en Centro-America. CEPREDENAC- NORSAR.

Saweda, Y., Taga, M., Watanabe, M., Nakamoto, T.,
Nagumo, H., Kudo, K., Horike, M., Sakajiri, N., Masaki, K.,
Sasatani, T, 2004. Applicability of microtremor H/V method for
KIK-NET strong motion observation sites and Nobi plain. Thirtheen
WCEE, Vancouver, B.C, Canada.

Autor:

Edwin Nadir Castrillo

National Autonomous University of Nicaragua
(UNAN-Managua).

International Institute of Seismology and Earthquake
Engineering (IISEE), Building Research Institute (BRI),
Japan.

Lund University, Department of Geology,
Sweden.

Eto Kiminobub

Tokyo Soil Research, inc.

Toshiaki Yokoic

International Institute of Seismology and Earthquake
Engineering (IISEE), Building Research Institute (BRI),
Japan.

Peter Ulriksend

Lund University, Department of Geology,
Sweden.