A Landslide Risk Rating System for Baguio, Philippines

Artessa Saldivar-Sali
a
, Herbert H. Einstein
b,

a
University of the Philippines, Philippines
b
Massachusetts Institute of Technology, 77 Massachusetts Avenue, Room 1-342, Cambridge, MA 02139, USA
Received 18 February 2005; received in revised form 9 November 2006; accepted 30 November 2006
Available online 29 December 2006
Abstract
This research formulates a Landslide Risk Rating System for the Greater Baguio area in the Philippines. It is hoped that the tool
will be made a part of the physical/urban planning process when used by engineers and planners and used to address risks posed by
landslides given the rapidly increasing concentration of population and the development of infrastructure and industry in the
Baguio area as well as in other parts of the Philippines.
This systemdescribes the hazard through socalled hazard contributing factors which inthis case are bedrock geology, slope gradients
and vegetation. This is then followed by characterizing risk through the risk contributory factors, which are population and land use.
The analysis performed in this study is based on the specific attributes of the study area, namely: an area with relatively extreme
topographic relief, underlain by variable bedrock geology, but without significant differences in amount of rainfall from one zone to
another during a major precipitation event.
Although this study and the proposed system are area-specific they have wider application. This is facilitated by emphasizing
what is specifically local and indicating how adaptation to differing local conditions might be done.
2007 Elsevier B.V. All rights reserved.
Keywords: Landslides; Risk rating; Geologic factors; Land-use planning
1. Introduction and background
With the rapidly increasing concentration of popula-
tion and development of infrastructure, the loss of life
and property damage caused by landslides in the
Philippines in general and the Greater Baguio area in
particular has grown significantly over the past decades.
Baguio City is located in Luzon Island, approximately
200 km north of Manila (Fig. 1). Like many other areas of
the Philippines, the Baguio area is subject to many factors
favoring the occurrence of landslides: the mostly moun-
tainous terrain; a humid climate with frequently occurring
typhoons and associated heavy rainfall; earthquakes; defo-
restation caused by human activity (mining, heavy deve-
lopment, agriculture).
Baguio City's geographical location is 1624 N
latitude and 12036 E longitude. Most of the developed
portion lies in the northern half of the city. Baguio City
extends 8.2 kmfromeast to west and 7.2 kmfromnorth to
south. It has a perimeter of 30.98 km. Baguio is best
known for its climate, unique in the otherwise tropical
country. On the average, the temperature is 8 C lower
than the temperature in the lowlands. Generally, the
maximum temperature experienced in the area is 26 C.
Engineering Geology 91 (2007) 8599
www.elsevier.com/locate/enggeo

Corresponding author.
E-mail address: einstein@mit.edu (H.H. Einstein).
0013-7952/$ - see front matter 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.enggeo.2006.11.006
The City owes this climate to its elevation of approxi-
mately 1500 m above sea level.
Baguio's climate is characterized by a pronounced
rainy season from November to April and a dry season
from May to October. Average annual rainfall in Baguio
is 3648 mm (19502003). An average number of five
tropical cyclones, with a maximum-recorded number of
nine tropical cyclones, affect Baguio every year. The
City receives the highest amount of rainfall in the coun-
try, twice the volume experienced in Manila. In 1911,
the world record of 1168.1 mm in 24 h (July 1415) and
2009.6 mm in 2 days, 15 h (July 1417) were recorded
in Baguio (Jennings, 1950).
The physical framework of Baguio City integrated
roadand park systems into one. Baguio was envi-
sioned to evolve into a compact garden city for 25,000
to 30,000 people with Burnham Park at the city center
(Fig. 2). Supporting this development plan was the
enactment of a charter approved on September 1, 1909
that provided administrative autonomy for the city. Soon
after the city's charter was enacted, Kennon Road
(Fig. 2) was opened to vehicular traffic. The existence of
an artery to Baguio City, the Cordillera Region's
distribution center, triggered the gold mining boom in
the surrounding areas in the early to mid 1930s. Baguio
City was the service and operations center for the
mining industry and, hence, a direct beneficiary of the
economic growth. The events of the Second World War
unfortunately left the city in total devastation. Fast
paced development, however, ensued following the war
years. Such development trends transformed the city
into what it is today, a premier urban center north of
Manila, performing a multiplicity of roles, as an
educational, trade, tourism and administrative center.
Baguio City has 129 barangays, the Philippines'
smallest unit of local government (http://www.baguio.
gov.ph, 2004).
In 1982, the population of Baguio City was
approximately 128,000. According to a 2000 population
census, Baguio is home to 252,386 people with a
population density of 5151/km
2
(http://www.baguio.
gov.ph, 2004). This is more than a doubling of the
population in only 18 years. The population in 2005 is
projected to reach 280,000. The city's population is
growing rapidly at an annual growth rate of 4.4% (more
than twice the national population growth rate) or an
actual increment of about 7900 individuals a year (http://
www.baguio.gov.ph, 2004). This puts immense pressure
on residential, commercial, institutional and infrastruc-
ture land use developments to expand proportionately.
This situation is compounded by inadequate and
ineffective control over land development. As a result,
the traditional role of the citythat of a mountain resort
is jeopardized. With a total land area of 49 km
2
, and a
Fig. 1. The Philippines and Baguio City.
86 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
population density of over 5000/km
2
, Baguio is
classified as a highly urbanized city. Attendant negative
results of recent and ongoing developments in the city
are rapid loss of open space, destruction of the natural
environment, and scarring of the landscape. Risks to the
safety of the residents as a result of hillside developments
have also increased significantly. Parenthetically, there
are far-reaching effects on the city's economy since
tourism, which is the City's basic industry, depends so
heavily on the natural environment for its viability.
One of the most important risks associated with the
uncontrolled development is the triggering of landslides
and conversely the effect of landslides on common
activities. This paper will propose a system for landslide
risk rating for the Baguio area. Before describing the
development of this system, it is necessary to discuss the
different types of land use in more detail. There are five
land-use categories distinguishing the Benguet Physical
Land Resources Evaluation conducted by the Bureau of
Soils of the Department of Agriculture; namely: built-up
areas, agricultural areas, grassland/shrubland areas,
woodland/forest areas, and miscellaneous land types
(Bureau of Soils and Water Management, 1980).
Built-up areas are used primarily for commercial and
residential purposes. Baguio City proper is the major
population center in Benguet province and in the entire
Fig. 2. Baguio City proper (C. Mapping and Reprography Dept. 1977 and 1995.).
87 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
Cordillera region. Two thousand six hundred hectares
(approximately half of Baguio City's land area) are
classified as built-up. Built-up areas are found mostly on
the Baguio Plateau (the portion shown in Fig. 2), along
broad alluvial valleys like La Trinidad, flood plains of
major streams, and near access roads and water sources.
In agricultural areas, rice is the primary crop that is
planted (in paddies or terraces). Other crops that are
planted on mountain and hill slopes, and on river terraces
are corn, vegetables, legumes, fruits, root crops, and
coffee, among others. Within the Greater Baguio area
(Fig. 3), 1557 ha are devoted to agriculture (Rillon, 1992).
Unforested grassland/shrubland areas are generally
found on steep to very steep slopes. These are usually
idle lands. Idle land means that it is not used for any
purpose except, very rarely as pasture. This type of land
comprises the largest percentage of the Greater Baguio
area at 57.3% or 56,427 ha (Rillon, 1992). Because these
are idle, future development will most likely take place in
these areas, except on the very steep slopes (N50%, see
Section 2.3).
Woodland/forest areas in Greater Baguio are found
mostly in the Northeast of Baguio City (see e.g. Fig. 3).
The total forest area is 33,718 ha or 34.2%. The ideal
figure for the entire Greater Baguio area is 60% forest
cover (Rillon, 1992). Ideal means that any future
development should maintain at least 60% forest in a
particular area. However, deforestation due to illegal
logging and slash-and-burn farming practices has made
this figure very difficult to achieve.
Miscellaneous land-use types include mine pit sites,
filling ponds, reservoirs, and riverwash/riverbeds. In
total, these areas occupy 750 ha, which is less than 1%
of Greater Baguio (Rillon, 1992).
Human activity plays a large role in increasing the
vulnerability of Baguio's slopes to landslides. Much of
the Baguio area is denuded due to heavy deforestation.
This was brought about by illegal logging and the
prevalence of slash-and-burn farming, which is a
traditional agricultural practice in the area. Building
construction, road blasting, and blasting associated
with mining activities disturb the static equilibrium of
Fig. 3. Landslide map of Greater Baguio, Philippines (after Mendoza, 1991). (Contour intervals: 100 m).
88 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
slopes. Undercutting of slopes to make room for
construction in adjacent areas likewise disturbs slope
equilibrium. These areas with high population density,
property, and economic activity that will be adversely
affected by geologic hazards must be given focus in
risk zonation.
The preceding comments explain why on the one
hand landslides occur quite frequently in the Baguio
area and why on the other hand they represent a
significant risk. Fig. 3 shows a map of the landslides in
the Baguio area. Most landslides are rotational, slump
types. A detailed description of each slide is provided
in the thesis underlying this paper (Saldivar-Sali,
2004).
2. Landslide Risk Rating System
2.1. Framework for the Landslide Risk Rating System
Landslide hazard and/or risk rating systems are an
essential tool of land use and urban planning. Areas,
which due to their topography, are subject to landslides,
i.e. mountainous regions have, therefore, seen early and
first applications of systems for landslide hazard/risk
mapping, for instance in the Alps (Antoine, 1978;
Bonnard and Noverraz, 1984; Einstein, 1988; Noverraz
and Bonnard, 1990; Einstein, 1997).
Similarly, if ground conditions, rainfall and topogra-
phy combine to produce landslides, corresponding
hazard/risk mapping systems have been developed
such as in Southern Italy (Carrara et al., 1977; Carrara,
1984) or in the San Francisco Bay area (Nilsen et al.,
1979; Brabb, 1984). A particularly interesting location
is Hong Kong where the above-mentioned natural
conditions combine with high population/building
density. The approaches used there (Roberds and Ho,
1997; Wong et al., 1997; ERM, 1998; Ho et al., 2000;
Wong, 2005a,b) are, therefore, exemplary. A somewhat
specialized area of slope instability hazard and risk are
rockfalls affecting roads and railroads. This domain is
interesting since it often involves assessments by people
who are not geotechnical or geologic specialists (New
York State, DOT, 1990; Pierson et al., 1990; Oregon
DOT, 2002). Given the wide range of landslide or slope
Fig. 4. Geologic Map of Greater Baguio (after Lands Geology Division, 1995). (Contour intervals: 100 m).
89 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
instability hazard and risk assessment approaches, it is
not surprising that generalizations have been attempted
or that reviews of such systems have been made (Varnes,
1978, 1984; TRB, 1996; Leroi, 1996; Cruden, 1997; Dai
and Lee, 2002; Leroi et al., 2005; Fell et al., 2005;
Cascini et al., 2005; Picarelli et al., 2005).
What will be proposed here is somewhat different,
although in principle following the standard approaches
consisting of the determination of a hazard and followed
by combining the hazard with predicted consequences to
estimate risk. The difference is that the proposed system
is based on relatively simple characteristics, which can
be easily assessed in the field or from generally available
records. This makes it easily applicable in developing
countries. However, and this will be discussed toward
the end, the system also allows one to extend it by
including higher level geotechnical information (and
more detailed information on the consequences for that
matter). This would eventually lead to a full fledged
system similar to those in the previously mentioned
literature and do so in a systematic way.
In recognition of the situation discussed in Section 1,
a Landslide Risk Rating System for the Baguio area is
formulated. The aim for this tool is to contribute to the
mitigation of the problem as part of the physical
planning process when used by engineers and urban
planners.
This Landslide Risk Rating System integrates
different hazard contributory factors (bedrock geology,
slope gradients, vegetation) and factors contributing to
risk (population, land use). In this study, Hazard refers
to the probability that a particular danger will occur
within a given period of time, while Risk refers to
the hazard multiplied by the potential worth of loss
(Einstein, 1997).
The analysis performed in this study is based on the
specific attributes of the study area, namely: an area with
relatively extreme topographic relief, underlain by
variable bedrock geology, but without significant
differences in amount of rainfall from one zone to
another during a major precipitation event.
Although strong precipitation is identified as the main
trigger of these landslides, it is not included as a variable
in this particular risk rating system because heavy rains
associated with typhoons are almost always uniform
throughout the entire Baguio area. For this reason it is
assumed that all slopes, regardless of location, are subject
to the same hazard level due to rainfall.
Three factors are, therefore, identified as being
variable across geographical locations within Baguio
and are the inputs of this risk rating system, viz. bedrock
geology, slope gradient and vegetation. A 54 km
2
-area
was delineated on the 1:15,000 scale Engineering
Geomorphology Map of Baguio City (after Mendoza,
1991, Mines and Geosciences Bureau). All of the
landslides shown on that map fall within the study area
(Fig. 3).
2.2. Bedrock geology
Bedrock geology is the basic contributory variable
used in this Landslide Risk Rating System. The rationale
for using geology as the base factor is that the
engineering properties and behavior of a site vary
according to the subsurface materials. More importantly,
the engineering properties of the residual soils involved
in the landslides can be related to the parent rock from
which these soils originated. The result of this is that the
relative differences of the residual soils can be expressed
by the relative differences between the underlying bed-
rock units.
This is adequate at this point of development and
applicable to the risk rating system but it is also an
aspect that can be refined. For instance, actual soil
properties (index properties such as Atterberg limits or
engineering properties) can be described. Also, the
bedrock geology could be refined through inclusion of
structural aspects (if visible).
There are seven significant geologic units present in
Greater Baguio (Fig. 4). The oldest is the Cretaceous
Pugo Formation (KPpf), a stratified sequence of basaltic
and andesitic volcanic rocks. This is followed by the Late
Oligocene Zigzag Formation (PNzf), a series of con-
glomerates, sandstone, and shale with limestone lenses in
some areas. The Late Oligocene Central Cordillera
Diorite Complex (Pcdc), is an intrusive consisting of
Table 1
Percentage of landslide-affected area by geologic units
Baguio formation
(NQbf )
Zigzag formation
(PNzf/Mzf )
Kennon limestone
(Nkl )
Mirador limestone
(Nml )
Klondyke formation
(Nkf )
Pugo formation
(KPpf )
Area, (km
2
) 24.4 10.8 0.6 14.3 2.5 1.4
Landslide Area, (km
2
) 1.4 0.7 0.0 0.6 0.02 0.2
Landslides (%) 5.7 6.5 0.0 4.2 0.8 14.3
90 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
diorites and granodiorites. The biohermal Early Miocene
Kennon Limestone (Nkl) is distinguishable by its karst
topography. The Middle Miocene Klondyke Formation
(Nkf) is a thick sequence of conglomerates, tuffaceous
sandstone, volcanic and tuff breccia, some siltstone, and
mudstone, followed by the Late Miocene Mirador
Limestone (Nml), and the youngest is the Pliocene
Baguio Formation (NQbf) composed mainly of
pyroclastics.
To determine the relative influence of bedrock
geology on landslide occurrences, the area of each
geologic unit lying within the study area was measured
(Fig. 3, see also Table 4). Within each geologic unit, the
area affected by landslides was also measured (Fig. 3).
The percentage (by area) of landslides within each
geologic unit was computed, with the results shown in
Table 1.
Based on the results presented in Table 1, it is
proposed that the geologic units be rated as shown in
Table 2. Under this proposed classification scheme, the
geologic units would be rated as shown in Table 3.
2.3. Slope gradient
Baguio City proper is located on a relatively flat upland
plateau with small elevation differences. Within the city
limits the topographic conditions are mild to moderate
with rounded hills and mountains. Elevations range from
1300 to 1600 m above sea level. Changes in elevation and
topographic conditions, however, become extreme outside
the City proper. The surrounding areas are heavily
dissected by young, vigorously eroding river systems,
giving the area large elevation differences. For instance: in
the Northwest Plateau 1500masl to 300 masl over 7 km; in
the Northeast 1700 masl to 600 masl over 8 km, in the
Southeast 1500 masl to 500 masl over 5 km; and in the
Southwest 1400 masl to 700 masl over 8 km.
Five slope categories are recognized (by the
Department of Environment and Natural Resources)
according to gradients: Class I (08%): level to gently
sloping; Class II (918%): gently sloping to undulating,
Class III (1930%): undulating to moderately steep,
Class IV (3150%): moderately steep to steep, Class V
(N50%): very steep. These slope categories super-
imposed on the geological units are shown in the map
of Fig. 5.
The five slope categories are also adopted for the
Landslide Risk Rating System in order to maintain
congruence with the existing land-use regulations. The
engineer or geologist can then determine the gradient of
the location being evaluated and assign a class to the
slope.
Usually steep slopes are likely to be less stable than
flatter slopes underlain by the same type of rock. Runoff
velocity also increases with slope and so, consequently,
does the erosion rate. However, it was not possible to
assume independence of the two factors, geology and
slope, in their contribution to landslide occurrence in the
Greater Baguio area.
The first step in the procedure employed to determine
the dependence of topography on geology through
the contribution of the two factors (slope inclination
and bedrock geology) to landslide occurrence is
shown in Table 4 as percentage (by area) of each
slope class found in the geologic units.
Table 3
Geologic unit hazard ratings
Baguio
formation
Zigzag
formation
Kennon
limestone
Mirador
limestone
Klondyke
formation
Pugo
formation
% landslides 5.7 6.5 0.0 4.2 0.8 14.3
Class II III I II I V
Table 2
Classification of landslide percentage areas per geologic unit
% landslides Classification
03% Class I
3.016% Class II
6.019% Class III
9.0112% Class IV
12.0115% Class V
N15% Class VI
91 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
The percentage (by area) of landslides occurring in
each slope class found in the geologic units was
determined. These results are shown in Table 5.
The normalized difference between the percentage of
landslides in each slope class within a given geologic
unit (Table 5) and the percentage of each geologic
unit occupied by that particular slope class (Table 4)
were determined. These normalized differences, ND,
Fig. 5. Slope categories superimposed on geologic units in study area. (The map section corresponds to the marked area in Fig. 4).
were calculated as follows:
ND
% Landslides in Each Slope Class Within Given Geologic Unit% Geologic Unit Occupied by Slope Class 100
% Geologic Unit Occupied by Slope Class
1
and are shown in Table 6.
Hence, a normalized difference of zero (0) indicates
that the expected number of landslides occurred in the
particular slope class, given its areal proportion in the
geologic unit. The normalized difference of 100
indicates that no landslides occurred in the geologic
unit (within the study area). The negative values of
normalized difference indicate that fewer landslides than
were expected occurred in the particular slope class,
given its areal proportion in the geologic unit. Positive
92 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
values of normalized difference indicate that more
landslides than were expected occurred in the particular
slope class, given its areal proportion in the geologic
unit.
Based on the results presented in Table 6, a modifier
is applied to the classification of geologic units
presented in Table 3. The modified geologic formation
ratings are likewise shown in Table 6. For normalized
differences between 50 and +90, the original geologic
class is retained, reflecting the fact that those specific
topographies do not contribute to a significant increase
or decrease of the occurrence of landslides within the
geologic unit. For normalized differences of 50.01 and
below, the modified geologic class is lowered by one
increment from the original geologic class. This reflects
the fact that those particular topographies are associated
with a significant decrease of landslides within the
geologic unit. For normalized differences of +90.01 and
above, the modified geologic class is increased by one
increment. This reflects the fact that those particular
topographies contribute to a significant increase of
landslides within the geologic unit. One exception is
gradient Class Vin the Pugo Formation. In this case, the
modified geologic class is increased by two increments
because of the extreme positive value of normalized
difference.
2.4. Vegetation
The most common vegetational cover in Greater
Baguio is the broadleaf (or short leaf) pine tree that is
able to grow in the area because of the cool weather
(James Montgomery Consulting Engineers, 1974).
Other types of vegetation include bushes/scrub, grass,
or crop land/agricultural land.
Within each geologic unit, the percentage of land-
slides occurring in areas covered by each vegetation
type was determined. The results are shown in Table 7.
As shown in Table 8, hazard classifications are
assigned to the different types of vegetation based on the
results presented in Table 7. Areas covered with Class I
vegetation are the least susceptible to landslides, while
those in Class IV are the most susceptible.
2.5. Faulting/tectonics
The effects of faults on landslides in Baguio were not
taken into consideration in this study for several critical
Table 5
Percentage by area of landslides in each slope class
Geologic unit Total area of
landslides (km
2
)
Class I area of
landslides (%)
Class II area of
landslides (%)
Class III area of
landslides (%)
Class IV area of
landslides (%)
Class V area of
landslides (%)
Baguio formation
(NQbf )
1.45 62.88 7.63 1.55 11.64 16.30
Kennon limestone
(Nkl)
0.00 0.00 0.00 0.00 0.00 0.00
Mirador limestone
(Nml)
0.64 51.75 19.30 28.95 0.00 0.00
Klondyke formation
(Mkf )
0.02 100.00 0.00 0.00 0.00 0.00
Zigzag formation
(PNzf )
0.69 27.24 4.88 28.45 15.85 23.58
Pugo formation
(KPpf )
0.17 0.00 0.00 60.29 0.00 39.71
Table 4
Percentage (by area) of slope classes per geologic unit within study area
Slope classification Class I (08%) Class II (918%) Class III (1930%) Class IV (3150%) Class V (N50%)
Geologic unit Total area (km
2
) Area (%) Area (%) Area (%) Area (%) Area (%)
Baguio formation (NQbf ) 24.36 30.86 52.64 5.70 4.99 5.82
Kennon limestone (Nkl) 0.65 6.67 0.00 93.33 0.00 0.00
Mirador limestone (Nml) 14.36 15.19 5.70 79.11 0.00 0.00
Klondyke formation (Mkf ) 2.49 100.00 0.00 0.00 0.00 0.00
Zigzag formation (PNzf ) 10.77 20.55 3.65 40.18 22.83 12.79
Pugo formation (KPpf ) 1.40 0.00 0.00 95.24 0.00 4.76
Total 54.03
93 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
reasons (DENR-MGB, Task Force, 1990). Tectonically,
the whole area is very active and therefore practically all
the geologic units are highly faulted. Secondly,
considering the very high density of faulting, it is
doubtful that all the faults in Baguio have been mapped.
To use the mapped faults in risk rating would give bias
to the well-mapped areas. Therefore, like precipitation,
faults are treated as a common denominator for the
whole of Greater Baguio.
2.6. Possible refinements
As mentioned earlier, the proposed system is
intentionally simple, i.e. relying on easily identifiable
characteristics and easily distinguishable factors. Future
refinements can involve the inclusion of factors which are
not considered now such as faulting or river erosion.
Already mentioned was the possibility to use more spe-
cific ground characteristics. At other locations rainfall
Table 7
Percentage of slides per vegetation type, by geologic unit
Percentage of landslides
occurring in:
Baguio formation
NQbf
Kennon
limestone Nkl
Klondyke
formation Mkf
Pugo formation
KPpf
Zigzag formation
PNzf
Mirador
limestone Nml
Broadleaf 47.35

100

10.53
Broadleaf Mix 2.33

Scrub

3.51
Grass 5.05

Crop/Agricultural Land 10.87

3.51
None 34.41

100

82.46
*Vegetation data not available.
Table 6
Modified geologic classes
Geologic class Gradient class Normalized difference Modified geologic class
Pugo formation KPpf (Class V) V 733.91 VII
IV 0 V
III 36.70 V
II 0 V
I 0 V
Zigzag formation PNzf (Class III) V 84.43 III
IV 30.58 III
III 29.20 III
II 33.59 III
I 32.57 III
Baguio formation NQbf (Class II) V 180.26 III
IV 133.49 III
III 72.75 I
II 85.51 I
I 103.74 III
Mirador limestone Nml (Class II) V 0 II
IV 0 II
III 63.41 I
II 238.82 III
I 240.69 III
Klondyke formation Mkf (Class I) V 0 I
IV 0 I
III 0 I
II 0 I
I 0 I
Kennon limestone Nkl (Class I) V 100 0
IV 100 0
III 100 0
II 100 0
I 100 0
94 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
may not be uniform and would have to be distinguished.
All this is simply an indication as to how the system may
be expanded if desired and possible.
3. Hazard rating
The resulting hazard ratings based on bedrock
geology and slope gradient (as combined in the
Modified Geologic Class) and vegetation type are
shown in Table 9. This is simply one possible rating
system that may be derived from the results presented in
the previous sections. A range of 2 is assigned to each
hazard rating in order to allow the user to evaluate
specific sites, taking into account their unique char-
acteristics, and use his or her judgment to differentiate
whether the Hazard is relatively high or relatively low
for its rating class.
Hazard ratings range from a high of 99 for the
unstable type of slope to a low of 35 for the most stable
type of slope. The hazard rating numbers in Table 9 are
an attempt at expressing the relative hazard. The most
hazardous combination that was found was given the
highest hazard rating and the least hazardous combina-
tion that was found was given the lowest hazard rating.
Note that no combination was given a hazard rating of
100 in recognition of the fact that there may be other
combinations of contributory factors that are even more
hazardous. Similarly, Hazard Ratings from 0 to 34 may
be assigned to areas whose combination of contributory
factors proves to be less hazardous than any found in
this study.
4. Risk rating
4.1. Principle
To define the potential consequences of landslides, a
typology of damage related to landslide risk is
established, namely:
Loss of life
Loss of property
Hence, hazard can be related to risk by including the
potential consequences of a hazardous event (a threat or
danger, the occurrence of which is uncertain) (Einstein,
1988). The Hazard Rating (HR) can be converted to a
Risk Rating (RR), by using multipliers based on land
use and on population.
RR HRTLUMTPM 2
where:
RR Risk rating
HR Hazard rating
Table 9
Landslide hazard rating of the Greater Baguio Area
Modified geologic class Vegetation type Hazard rating
VII IV 9799
III 9596
II 9394
I 9192
VI IV 8990
III 8788
II 8586
I 8384
V IV 8182
III 7980
II 7778
I 7576
IV IV 7374
III 7172
II 6970
I 6768
III IV 6566
III 6364
II 6162
I 5960
II IV 5758
III 5556
II 5354
I 5152
I IV 4950
III 4748
II 4546
I 4344
0 IV 4142
III 3940
II 3738
I 3536
Table 8
Classification of different types of vegetation
Type Class
Broadleaf mix or bushes/scrub I
grass or crop land/agricultural land II
Broadleaf III
None IV
95 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
LUM Land-use multiplier
PM Population multiplier
4.2. Land-use multipliers
For Land Use, a multiplier of 1.0 is applied to the
most critical land-use situations. Multipliers less than 1
reduce the Risk Rating when land use in the area is less
critical (Table 10). In this study, multipliers were
selected ordinally (higher position in sequence
higher risk) to reflect the relative reductions to the
Risk Rating depending on how critical the land use is in
the area.
The most critical Land Use situations in terms of loss
of life and property are in the Built-up areas. These areas
are used primarily for commercial and residential
purposes. For the purposes of this Landslide Risk
Rating System, roads and highways are also included in
the Built-up classification because of the significant
impact of their closure when affected by landslides. It is
clear that Built-up areas contain capital-intensive
equipment and machinery, civil works, and critical
economic activities. The effects of a landslide would
thus have the most import, and the Risk Rating must not
be reduced.
Grasslands are unforested areas covered with grasses
and shrubs and are generally found on steep to very
steep slopes. They are usually idle (see Section 1) land
very rarely used for pasture purposes. However, because
the land is idle, future development will most likely take
place there. The Land Use multiplier is thus set at 0.95 in
recognition of the fact that grassland areas will most
likely be tapped for future development (except in the
case of very steep slopes).
A reduction of 10% is applied to the Risk Rating in
agricultural areas. Agriculture is a significant component
of the region's economy and, thus, is assigned a Land
Use multiplier of 0.9.
Miscellaneous areas include mine pit sites, tailing
ponds, reservoirs, and riverwash/riverbeds. Since mine
pit sites and tailing ponds are components of the
operations of mines, their destruction by landslides
would have some economic impact. Landslides near
reservoirs and riverbeds might lead to siltation and/or
damming, and have adverse ecological impact. These
are given a Land Use multiplier of 0.85.
The largest reduction is applied to forested areas.
This is because these areas are largely undeveloped and,
ideally, will remain undeveloped in order to facilitate
reforestation of the greater Baguio area. A Land Use
multiplier of 0.80 is applied.
Clearly, these Land Use multipliers can be changed if
desired.
4.3. Population multipliers
For Population, the multipliers range from 0.751.0,
with high population in the area in question being
assigned a multiplier of 1.0 (Table 11). This reflects the
fact that the full Risk Rating must be applied in areas
with the highest population.
4.4. Possible refinements
Similar to the refinements of the factors affecting the
hazard (see Section 2.6) it is also possible to describe
risk, i.e. the consequences in more detail. For instance,
the land use can be further subdivided. A further step
would be to express the consequences on infrastructure,
material property and life in form of utilities as has been
discussed in Einstein (1988).
4.5. Use of the Landslide Risk Rating System
Practically, the risk rating system is to be used as
follows:
determine combined slope class/geologic unit (e.g.
like map, Fig. 5)
determine modified geologic class for each of these
combinations
for each area as obtained above, subdivide each for
vegetation and associate hazard rating (Table 9)
Table 10
Land-use multipliers
Land use Land-use multiplier
Built-up 1
Grasslands 0.95
Agriculture 0.9
Miscellaneous 0.85
Forest 0.8
Table 11
Population multipliers
Population per barangay Multiplier
01000 0.75
10002000 0.8
20003000 0.85
30004000 0.9
40005000 0.95
N5000 1
96 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
for each area with its particular hazard rating, mul-
tiply with appropriate land use/population multiplier.
To show how the application of the proposed hazard
and risk system would look, we have applied it in the
District V area of Baguio City. Table 12 lists all the
entries (geology, etc.), the resulting hazard rating and
then with the population multiplier the resulting risk
rating. The ratings apply to the barangays, i.e. villages
named in Table 12. Hence, the land units to which the
ratings apply are those villages. All this information is
also shown on the map of Fig. 6.
The risk rating system described above has just been
developed but has not yet been practically applied
except for this example application. One of the major
benefits of using a tool such as this is that it can serve as
a basis for improving building codes. Instead of having a
generalized building code, a code adapted to local
conditions can be created. In addition, building regula-
tions may be modified to take into account variations in
Table 12
Hazard and risk ratings for selected villages in the Greater Baguio area
Barangay Geologic
class
Gradient
class
Modified geologic
class
Vegetation
type
Hazard
rating
Land-use
multiplier
Population
multiplier
Risk
rating
Engineer's Hill NQbf I III IV 65 66 1.00 0.85 55 56
Military cut-off NQbf II I III 47 48 1.00 0.8 38 38
Puliwes NQbf II I IV 49 50 1.00 0.85 42 43
Camp 7 NQbf II I I 43 44 0.95 1 41 42
San Vicente NQbf II I IV 49 50 0.90 0.95 42 43
Camp Sioco NQbf III I IV 49 50 1.00 0.8 39 40
Balsigan NQbf II I IV 49 50 0.95 0.85 40 40
Phil-Am NQbf II I III 47 48 0.95 0.75 33 34
SLU-SVP PNzf III III IV 65 66 0.90 0.85 50 50
Sto. Rosario NQbf III I III 47 48 1.00 0.85 40 41
Bakakeng Central NQbf IV III IV 65 66 1.00 0.95 62 63
Sto. Tomas proper PNzf III III IV 65 66 0.95 0.9 56 56
Fig. 6. District Vof Baguio City hazard and risk rating associated with villages named in map. (See also Table 12).
97 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599
geotechnical conditions across a given planning area.
Significant savings could, in principle, be achieved by
introducing a certain amount of flexibility into building
codes. Another use, quite clearly, is the prioritization of
mitigating measures. Such measures might include:
diversion of the flow of stream and river channels away
from roads to prevent undercutting of slopes in road
cuts, structural stabilization by installation of piles at the
toe of an undercut slope to increase shear strength of the
slope, stringent control of blasting during road con-
struction and mining operations in order to minimize
vibrations, reforestation of denuded areas with fast-
growing trees, improvement of surface drainage sys-
tems to reduce downward percolation of rainwater into
slopes, slope-stabilization of existing and proposed
slopes, removal of water already present in the ground
using subsurface drains, rock bolting in slopes where the
planes of weakness daylight, review of the strategy
for building road networks, and mapping of existing
slide areas in geologic exploration reports to minimize
adverse effects associated with building on old slide
materials.
5. Conclusions
In this study, a Landslide Risk Rating System was
developed for Greater Baguio as a tool for engineers and
planners to delineate land-use and building constraints.
Factors that are taken into account in the formulation of
the rating system are bedrock geology, slope angles, and
vegetation as well as land use and population density.
The actual procedure consists of associating these
factors in sequence with particular areas eventually
resulting, e.g. in a map showing risk ratings for different
areas. This has been shown in an example application to
the District Varea of Baguio City.
One of the major benefits of practically using a tool
such as this is that it can serve as a basis for improving
building codes. Instead of being generalized, building
codes may be more effectively adapted to local condi-
tions. In addition, building regulations may be modified to
take into account variations in geotechnical conditions
across a given planning area. Significant savings could, in
principle, be achieved by introducing a certain amount of
flexibility into building codes. Another use, quite clearly,
is the prioritization of mitigating measures.
References
Antoine, P., 1978. Glissements de terrains et amnagement de la
montagne. Bulletin de la Socit Vaudoise des Sciences Naturelles
74 (353) (Pasc. 1).
Bonnard, C., Noverraz, F., 1984. Instability risk maps from detection
to administration of landslide risk. Proc. 4th Int'l. Symp. on
Landslides. Canad. Geotech. Soc, Toronto.
Brabb, E.E., 1984. Proc. innovative approach to landslide risk
mapping. Proc. 4th Int'l. Symp. on Landslides. Canad. Geotech.
Soc., Toronto.
Bureau of Soils and Water Management, 1980. Present Land Use Map,
Province of Benguet (Philippines). Department of Agriculture.
Carrara, A., 1984. Landslide hazard mapping: aims and methods.
Mouvements de Terrains. Association Francaise de Gographie
Physique. Colloque de CAEN, pp. 142151.
Carrara, A., Publiese-Carratelli, E., Merenda, L., 1977. Computer based
data bank and statistical analysis of slope instability phenomena.
Zeitschrift fur Geomorphologie N.F. 21 (2), 187222.
Cascini, L., Bonnard, Ch., Corominas, J., Jibson, R., Montero-Olarte,
J., 2005. Landslide hazard and risk zoning for urban planning and
development. Proc. Landslide Risk Management, Vancouver.
Balkema.
Cruden, D.M., 1997. Estimating the risks from landslides using
historical data. In: Cruden, Fell (Eds.), Proc. Landslide Risk
Assessment. Balkema.
Dai, F.C., Lee, C.F., 2002. Terrain-based mapping of landslide
susceptibility using a geographic information system. Canadian
Geotechnical Journal 38.
DENR-MGB Task Force, 1990. Preliminary Earthquake and Geological
Hazard Mapping of Baguio City. Department of Environment and
Natural ResourcesBureau of Mines and Geosciences, Manila.
Einstein, H.H., 1988. Special lecture landslide risk assessment.
Proc. 5th, Int'l. Symposium on Landslides. Balkema.
Einstein, H.H., 1997. Landslide risk systematic approaches to
assessment and management. In: Cruden, Fell (Eds.), Proceedings
Landslide Risk Assessment, Honolulu. Balkema.
ERM, 1998. Landslides and Boulder Falls from Natural Terrain: Risk
Guidelines Report to Geotechnical Engineering Office of Hong Kong.
Fell, R., Ho, K.K.S., Lacasse, S., Leroi, E., 2005. A framework for
landslide risk assessment. Proc. Landslide Risk Management,
Vancouver.
Ho, K., Leroi, E., Roberds, W., 2000. Quantitative risk assessment
applications. Myths and Future Directions. Proc. Geo Eng.
Technomic Publishing Co.
James Montgomery Consulting Engineers, Inc., 1974. Report on the
Water Resources Evaluation Program, Baguio City, Benguet
Province. Prepared for Local Water Utilities Administration.
Jennings, A.H., 1950. World's greatest observed point rainfalls.
Monthly Weather Review 78, 45.
Lands Geology Division, 1995. Geological Map of Baguio City
Quadrangle. Bureau of Mines and Geosciences. Sheet 3169 III.
Leroi, E., 1996. Landslide hazard-risk maps of different scales
objectives, tools and developments. Proc. 7th Int'l. Symp. on
Landslides, Trondheim.
Leroi, E., Bonnard, Ch., Fell, R., McInnes, R., 2005. Risk assessment
and management. Proc. Landslide Risk Management, Vancouver.
Balkema.
Mapping and Reprography Department, 1977. Topographic Map of
Baguio City. National Mapping and Resource Information
Authority. Sheet 7175 IV.
Mapping and Reprography Department, 1995. Topographic Map of
Baguio City. National Mapping and Resource Information
Authority.
Mendoza, C.V., 1991. Engineering Geormorphology Map of Baguio.
Bureau of Mines and Geosciences.
New York State, DOT, 1990. Rockfall Hazard Rating Procedure.
98 A. Saldivar-Sali, H.H. Einstein / Engineering Geology 91 (2007) 8599