Content uploaded by C. P. C. David
Author content
All content in this area was uploaded by C. P. C. David on Jan 02, 2020
Content may be subject to copyright.
Community-based monitoring for
flood early warning system
An example in central Bicol River basin,
Philippines
Catherine C. Abon and Carlos Primo C. David
National Institute of Geological Sciences, University of the Philippines, Diliman,
Quezon City, The Philippines, and
Guillermo Q. Tabios III
National Hydraulic Research Center, University of the Philippines, Diliman,
Quezon City, The Philippines
Abstract
Purpose – The purpose of this paper is to integrate the proactive role of communities and the use of
flood modeling in the implementation of a flood early warning system.
Design/methodology/approach – Manual rain gauges were installed in 20 houses of volunteers
living within the Bicol River basin to monitor rainfall. Rain information is sent twice daily via SMS
message to a receiving computer. The received data are used to run a basin model that was developed
in HEC-HMS, which converts precipitation excess to overland flow and channel run-off.
Findings – Different watershed models were developed for different rainfall events. Geomorphic
analysis using 3 s SRTM Digital Elevation Model (DEM) processed in a GIS platform was also done to
refine the overland flow. The derived hydrographs were used in the HEC-RAS hydraulic model which
has as main output threshold values for the rain-flood relationship.
Research limitations/implications – Although SRTM DEM that was used for the geomorphic
analysis was sufficient for the purpose of the study, higher resolution DEMs can further improve the
mapping of spatial extent of flood areas.
Practical implications – The results are used for the forecast of flood and the timely issuance of
flood bulletins.
Originality/value – This study is the first to incorporate the involvement of the community in
establishing a flood early warning system. The method can also be used as a prototype for other flood
models in other parts of the country.
Keywords Philippines, Floods, Rivers, Communities, Modelling, Precipitation, Early warning,
Geomorphology, Bicol River basin
Paper type Research paper
1. Introduction
Flooding is one of the most common natural hazards causing immense damage to
agricultural production, destruction of infrastructure, and loss of lives. Flooding in
most cases cannot be prevented (Chubey and Hathout, 2004); however, the associated
damage due to floods can be significantly minimized through a community
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0965-3562.htm
Disaster Prevention and Management
Vol. 21 No. 1, 2012
pp. 85-96
rEmerald Group Publishing Limited
0965-3562
DOI 10.1108/09653561211202728
This work has been funded by the Christian Aid and Manila Observatory. Rainfall and
hydrologic data from PAGASA-Bicol Region headed by Arturo Balang made the calibration of
the model possible. Grateful acknowledgement goes to Arlene Dayao and to the countless people
of Bicol who have patiently assisted the authors during field surveys. Thanks are also extended
to laboratory members Stephanie Frogoso, Michelle dela Cruz, Darwin Riguer, David Chuy and
Krystelle Banaag for assisting during the field surveys, and Jeremy Rimando and Jun Pellejera
for the laboratory assistance, the UP CSWCD students Jesus Dominic Dizon and groupmates,
supervised by Matt Wamil, and David Michael V. Padua who heads the TPC.
85
Community-
based flood early
warning
flood preparedness program. Aside from an early warning system and civil
protection measures, the program should include flood forecasting through an accurate
quantification of flood risk (Sanders, 2007; Pappenberger et al., 2007; Liu et al., 2004).
Hydrologic and hydraulic modeling and the incorporation of geomorphic analysis is
often employed in flood studies. Lee et al. (2008) emphasized that the rainfall-runoff
process in a watershed is primarily controlled by watershed geomorphic features and
rainfall characteristics. Topographic maps are typically used for characterizing
geomorphic features. Recent high-resolution digital elevation models (DEM) proved
even more useful for flood inundation studies (e.g. Jasrotia and Singh, 2006; Bates and
De Roo, 2000; Correla et al., 1998). The availability of public access DEMS and software
for hydrologic and hydraulic modeling provided a condition for combining geomorphic
data and rainfall data for flood forecasting especially when resources are limited.
The presence of operational models for an early warning system alone is not enough
to effectively minimize or prevent the damages from flooding. The past experiences of
the country from flooding showed that early warning systems are often neglected by
the people. One of the challenges in early warning systems therefore is implementing
and sustaining it. The idea of incorporating the active involvement of the people in the
community with an early warning system aims to increase the effectiveness of such
systems. Learning by actual participation and taking a part in the system enable
people to understand more the value of these systems not only for themselves but for
the whole community that will be affected, and make them become more responsible in
performing their tasks in implementing and sustaining the system.
The Bicol River basin (BRB) with a drainage area of 3,770 km
2
which is among the
largest river basins in the country, is one of the most if not the most frequently flooded
areas in the country. It is an elongate basin of about 130 km length and 25 km width
that is trending northwesterly. It is bounded along its length by the Ragay Hills to the
southwest and a chain of volcanoes to the northeast. The Bicol River is a meandering
river flowing in a gradual slope (0.03) to its only outlet in San Miguel Bay to the north.
The study area is a 2.9-km segment of the Bicol River and the Pawili River, which
drains the southern side of Mt Isarog. It is part of the central BRB with a total
watershed area of 504 km
2
(Figure 1).
The climate in the region is characterized as having no dry season but a very
pronounced period of maximum rainfall from November to February. The average
annual rainfall varies from 2,000 to 3,600 mm in the basin (Philippine Atmospheric,
Geophysical and Astronomical Services Administration PAGASA), 2007). Occurrence
of typhoons is relatively frequent in the Bicol Region with an average of three to four
typhoons per year passing through the area. In November 2006, extensive flooding in
the BRB caused by Typhoon Reming claimed 1,200 lives and left 120,000 homeless.
An estimated 3.3 billion pesos of worth of farmlands and property were lost
(National Disaster Coordinating Council (NDCC), 2004, unpublished report). In 2009,
Typhoon Dante also caused extensive flooding in the central and northern portion of
the BRB.
2. Methodology
2.1 Installation of home-based stations
The distribution of the rain gauges within the basin was first designed and
pre-determined using 1:50,000 topographic maps. Two volunteers were then selected
in each marked location. The volunteers were taught of the importance of their
participation to monitor rainfall in the overall early warning system. They were also
86
DPM
21,1
provided with a protocol for sending SMS or text messages in a definite format the
rainfall level as they read it in their rain gauges. These messages will be sent in a
central computer and stored accordingly in an easily accessible database. Rainfall
recording and reporting will be done twice daily, one in the morning around 06:00-08:00
and in the afternoon at 16:00-18:00. More frequent recording and sending of rainfall
data is done during days of heavy rainfall. Volunteers are also encouraged to make any
observations related to monitoring rainfall. The rainfall data are used in simulating the
basin and channel models to be able to evaluate flood potential.
The installation of rain gauges in different households was undertaken for at least a
month. Different reactions were met during the installations. Most of the people easily
understood the importance of their involvement in the early warning system so they
promptly agreed to have the rain gauges installed in their houses. Some of them are
reluctant though because of their monitoring might interfere with their daily activities.
A total of 20 rain gauges were installed in different houses within the central BRB
and grouped in two covering a total of nine cities/municipalities (Figure 2). Locating
two stations close to each other is deliberately done to have redundancy of data,
wherein reported rainfall data can be checked by comparing it with the reports from its
adjacent station.
The manual rain gauges installed employed the use of a funnel collector fastened on
the side of a house’s roof and attached to a graduated container by a plastic hose.
The design implemented allows rainfall reading easier as this can be done inside the
volunteer’s home (Figure 3). This was a critical design aspect, as conventional rain
Mt Isarog
N
km
0510
Legend
Central BRB
Bicol River
Pawili River
Tributaries
Ragay Hills
Figure 1.
The study area (central
BRB) and the index map
(inset) showing the
location of the study area
in the Philippines
87
Community-
based flood early
warning
gauges placed even a few meters away from homes often precludes the measurement of
rainfall at the height of a rain event.
2.2 Flood model development
The flood model development includes the integration of geomorphological analysis
and basin and channel modeling. Each of the processes included are discussed in more
detail below.
Mt Isarog N
km
2024
Lake baao Mt Iriga Legend
Volunteers
Central BRB
Bicol River
Pawili River
Tributaries
Ragay Gulf
Figure 2.
The location of the 20
installed home-based
stations in the central BRB
Figure 3.
Manual rain gauge
installation for home-
based stations by the
students and local
volunteers. Photo
contributed by volunteer
students of UP CSWCD
headed by Jesus Dominic
Dizon
88
DPM
21,1
2.2.1 Historical flood reconstruction from documentary sources and field interviews.
Major rainfall events (those associated with tropical cyclones) were culled out of the
daily precipitation record of PAGASA for the period 1981-2008. However, most
of the corresponding flood extents brought by these tropical cyclones were not
recorded. Written documents about the precise flood extents and timing were scarce
hence, field interviews with local residents especially those living along the rivers
were done. From these interviews, the information obtained included the
characteristics of the different typhoon-induced floods, such as frequency, duration,
spatial extent, flood height, and flood timing with respect to the intensity and duration
of rainfall. Historic flood marks in the trees along the river and in the houses and posts
in the urban areas were also recorded. This information will be used to calibrate the
flood models to be generated.
2.2.2 The geomorphological analysis: ILWIS 3.3. The geomorphological analysis
was done using the DEM extracted from the public access data of the USGS SRTM.
A number of studies have confirmed the sufficiency of SRTM DEM for hydrologic
model applications and flood risk analysis (Schumann et al., 2008; Demirkesen et al.,
2007; Ludwig and Schneider, 2006). SRTM DEM has also been used to the extent of
extracting river cross-section for hydrodynamic modeling (e.g. Pramanik et al., 2009).
Sanders (2007) also concluded that SRTM DEMs are valid for flood models but
emphasized the importance of performing ground surveys of study sites when such
DEMs of limited resolution are utilized. Hence, additional elevation points were
gathered using a handheld GPS during the field surveys and integrated with
the SRTM DEM. Other parameters, such as floodplain characteristics, channel
modifications, and previous flood records were also gathered during the field survey
conducted. Channel width and channel bank height were also estimated given its
importance in controlling flood processes. The field data points and DEM were
integrated in a GIS platform to generate the base map for the study.
The DEM was processed using a hydrologic processing package in ILWIS 3.3
developed by ITC International Institute for Geoinformation Science and Earth
Observation. The processing includes DEM visualization, flow determination, and
network and catchment extraction processing. The digital delineation of sub-catchments
was counterchecked using higher resolution topographic maps and necessary
adjustments were undertaken. The sub-catchment map was then overlaid on a soil
map to further delineate sub-basins with relatively uniform soil type. A total of
22 sub-catchments were generated in the area. The physical parameters of each
sub-catchment such as area, overland flow length, wetness index, and stream power
index were all extracted and represented in tabulated forms.
2.2.3 The basin model: Hydrologic Engineering Center’s Hydrologic Modeling
System (HEC-HMS). The (HEC-HMS), version 3.2 was used to develop the hydrologic
model. The model is developed by the US Army Corps of Engineers, designed
to simulate the precipitation-runoff processes of dendritic watershed systems
(HEC, 2008a). This method was employed to generate peak flows and hydrographs for
selected typhoon events. Historical rainfall events recorded in the PAGASA were
used to simulate historical events while data from the newly installed rain gauges
were used for the recent typhoon events. The inputs include information from the
basin model, the meteorological model, time-series data and paired data, and control
specifications. The Green and Ampt method for calculating water loss due to
infiltration is used, a simplified version of Richard’s equation for unsteady water
flow in soil.
89
Community-
based flood early
warning
Translation of excess precipitation to runoff is done using the Clark unit
hydrograph transformation (Chow et al., 1988) which only requires two parameters: the
time of concentration (T
c
) and storage coefficient (R):
R¼0:6Tcð1Þ
T
c
represents the maximum travel time in the sub-basin and is used in the development
of translation hydrograph. The translation hydrograph is routed using linear reservoir,
which accounts for storage attenuation affects across the sub-basin.
Derived maps in ILWIS were loaded to the HEC-HMS as background maps. The
incorporation of hydrologic modeling in GIS redefines the model in spatial context
by giving new possibilities in understanding the fundamental physical processes
underlying the hydrological cycle and the solution of mathematical equations
representing those processes (Castrogiovanni et al., 2005; Liu and De Smedt, 2005;
Correla et al., 1998; Coroza et al., 1997). The maps also aid the visualization of the
hydrologic components in the modeled basin. The hydrologic components include
sub-basin, reach and junction. Each sub-basin was assigned with Green and Ampt
parameters based on soil classes in the area. The soil classes were obtained from the
Bureau of soils, digitized and superposed on the sub-basin map. The soil class map is
necessary in assigning the Green and Ampt parameter values such as suction and
conductivity. Likewise, the digitized land-use map was also overlaid on the sub-basin
map in order to estimate the percentage of impervious area in each sub-basin. Other
required parameters such as sub-basin area and initial discharge are derived from the
DEM processing and field historical data, respectively.
2.2.4 The river channel model: HEC River Analysis System (HEC-RAS). The flood
hydrographs obtained from the basin model were used as input data in the hydraulic
model. The river channel model was developed using the HEC-RAS, version 4.0
(HEC, 2008a). This model was designed to perform one-dimensional hydraulic
calculations based on the St Venant equations when flow conditions are unidirectional,
one phase and gradually varying and can estimate the water surface (from water
height), velocity, and depth profiles, computed from one cross-section to the next by
solving the energy equation with an iterative procedure. Previous flood studies have
validated the usefulness of this model (Shahrokhnia and Javan, 2005; Alho et al.,
2007; Machado and Sajad, 2007; Lastra et al., 2008; Knebl et al., 2005; Horritt and
Bates, 2002).
An important input in the channel model is the river geometry represented by a
series of cross-sections perpendicular to the flow direction. The cross-sections were
obtained by field surveys and higher resolution topographic maps availed from the
NAMRIA. The field survey was conducted using GPS and survey techniques,
obtaining a total of 112 cross-sections through the 35-km modeled river segment of the
Bicol River. The channel geometry was considered only for the Bicol River because
it is the largest stream in the basin. The georeferenced map of the study area was
used as background map in completing the geometry data of the model. Each
cross-section was georeferenced to the basin in order to develop floodplain maps. Other
parameters include left and right bank locations, downstream flow lengths roughness
coefficients (Manning’s n), and contraction and expansion coefficients. These
parameters were obtained using GIS tools and field surveys. The roughness
estimation was accomplished by combining land use data with tables of Manning’s n
values available in HEC (2008b). The hydrographs from HEC-HMS were incorporated
90
DPM
21,1
in the HEC-RAS model by assigning representative cross-section for each group of related
sub-basins as entry of lateral inflow to the main river. The representative cross-sections are
points where a group of associated sub-basins collectively drain to that point.
The HEC-RAS model was simulated using a sub-critical flow. Flow hydrograph in
the upstream, lateral inflow in the middle cross-sections, and normal depth in the
downstream are the boundary conditions used for the simulation. Since the river is
properly georeferenced, the output of the simulation was easily transferred to the
ILWIS. The extent was then digitized to produce a spatial flood extent map.
2.3 Model calibration and val idation
Basin parameters such as infiltration coefficients, time of concentration and baseflow
were iteratively adjusted and modified to produce a best fit between the model and the
observed events. The hydrographs generated from the basin model are calibrated with
observed discharge. The availability of discharge data, however, is scarce in the area.
Only two rainfall events have their corresponding discharge data. For the channel model,
the rain event associated with Typhoon Reming (November 29-December 4, 2006) was
used for calibration and validation. Delineated flood map as a result of the channel model
is validated by actual flood extents during that particular typhoon. This flood extent was
obtained from remote sensing image from the Dartmouth Flood Observatory. Figure 4
summarizes the method employed in model development.
3. Results and discussion
3.1 Geomorphological analysis
The maps derived from DEM-hydroprocessing provided a quick overview and
assessment of the areas that will likely to experience flooding (Figure 5). Flow
accumulation model derived from DEM processing indicates that most areas that
reported to have experienced flooding based on historical accounts fit with the regions
of flow accumulation. Table I shows the parameters obtained from the
geomorphological method and calculations using the mentioned empirical formula.
3.2 Basin model
A continuous record of data particularly for flood discharge in the area is scarce. Only
two rainfall events have corresponding discharge measurements, which were used in
the model calibration. After a number of simulations, the results of the basin model
fairly fit the observations – the hydrograph shape and timing of peaks matched. The
adjustments of initial loss and other parameters of each sub-basin were done to
accurately represent the rainfall run-off relationship over the area. For the transform
method, the time of concentration and storage coefficient were also modified. The
modification indicated that the time of concentration is a more sensitive parameter.
The changes done in the time of concentration yielded a more accurate timing of peaks.
3.3 Channel model
Relatively fewer modifications on the channel model were done to fit peak flood timing.
Most of the calibration was accomplished by adjusting the Manning’s nvalues in the
left over bank (LOB), channel, and right over bank (ROB). LOB and ROB roughness
range from 0.05 to 0.045, while that of the channel is 0.020-0.025.
The final output of the models is a map showing the flood extent over the basin
during Typhoon Reming. A segment of the Bicol River reach was processed and flood
inundation results were generated. The extent was superimposed on the reach map
91
Community-
based flood early
warning
digitized in ILWIS (Figure 6). The flood extent in the MODIS satellite image processed
by the Dartmouth flood observatory was digitized and superposed on the basin map.
The results from the satellite image showed that the model overestimates flooding
throughout the river over banks but underestimates the lateral extent. However, this
discrepancy may be attributed to the time when the image was acquired which was on
the latter part of the flood event (December 4, 2006). In general, the model-derived
points where flooding occurs strongly match the satellite data.
3.4 Test run of the early warning system
The Typhoon Preparedness Center (TPC) was established and stationed in Naga City,
Bicol Region. This serves as the center where flood forecasts will be issued during
typhoons. The operation was first tested during Typhoon Dante, which made landfall
on May 1 and left the country on May 5, 2009. The three-hourly rainfall was recorded in
TPC via text messages from the volunteers. These records were used in the simulation
and a flood bulletin was issued using the model results as guide. The timing of
the flood was predicted to occur at the tenth hour after the high-intensity rainfall
SRTM DEM
Field survey data
Hydrologic model for rainfall
events:
HEC-HMS
Output stream hydrographs
Calibration and
validation
Calibration and
validation
Output: stream breach points,
delineated flood extent, and
flood timing rainfall events
Hydraulic model for each
rainfall event:
HEC-RAS
DEM processing and storage
ILWIS
academic 3.3
Output sub-basins, major rivers
and tributaries, flow
accumulation, flow direction,
flow length, slope
Derivation of
parameters of each
sub-basin
Manual entry of
rainfall data
Control specifications
Field survey data: river
cross-sections and other
geomorphologic data, and
interview data
Figure 4.
Model development
scheme. The SRTM DEM
was processed using
ILWIS 3.3 where the
indicated outputs were
derived and manual entry
of the rainfall data was
done. After the control
specifications were set, the
hydrologic model was run.
The river cross-sections
derived from field works
and elevation data, as well
as the output hydrographs
from the hydrologic model
were then used for the
hydraulic model in
HEC-RAS. Finally, the
plan views and cross-
section of flood extent
derived from the hydraulic
model were overlayed on
the river to produce the
flood inundation maps.
Calibration and validation
in all the models were
also done
92
DPM
21,1
(14.7 mm/hr) during the rainfall’s first three hours. The channel model also predicts
that flooding will start in the southern towns of Bula and Milaor along the Bicol River.
Maximum flood in Bula were recorded to have reached 0.3-1.0 meters, confirming the
model’s prediction.
From the simulated models and the updates of rainfall record from text messages,
the TPC was able to issue three-hourly advisories. A total of 16 advisories were issued
during the most critical hours of the typhoon, which is on May 1-3. The first advisory
was issued on May 1, at 05:00 hours. At 1,200 hours, the third advisory had included
Legend
Legend
1
3
4
5
2
Legend
Legend
Central BRB
Bicol River
Pawili River
Tributaries
Sub-basins
1
16,632
33,264
49,895
66,526
N
NE
E
SE
E
SW
W
N
0 5 10 km
N
0 5 10 km
N
0510km
N
0510km
NW
(a)
(c)
(b)
(d)
Notes: (a) flow direction map; (b) flow accumulation map; (c) river network map;
(d) merged sub-catchment map
Figure 5.
The DEM hydro-
processing derived maps
Minimum Maximum
Physiographic parameters
Sub-basin area (km
2
) 1.2 25.6
Mean slope of the river reach (m/m) 0.002 0.05
Loss rate
Initial loss (mm) 0.36 1.2
Moisture deficit 0.38 0.5
Suction (mm) 41.95 290.2
Conductivity (mm/hour) 0.30 60.8
Impervious (percent) 0 15
Transform
Time of concentration (hour) 0.1 5
Storage coefficient (hour) 0.3 2.3
Tabl e I.
Estimated parameters
obtained from
geomorphological method
93
Community-
based flood early
warning
flooding to be imminent in Bula and Milaor. People therefore were advised to stay alert
and be ready for evacuation if necessary. At 1,500 hours, flooding have occurred
parallel to what the model have predicted. Advisory at this hour was issued with an
emphasis that flooding will continue in these areas. The flood alert level was sustained
until advisory 11, issued on May 2, at 1,500 hours. With the help of the advisories
issued, people were able to prepare for the imminent flooding in the study area.
4. Conclusions
Combining geomorphological methods with basin and channel modeling proved to be
an efficient way to establish an early warning system as far as technical requirements
are concerned. The use of public access software and DEM has maximized the utility of
these available resources for flood inundation modeling in central BRB. The use of
SRTM DEM for regional geomorphological analysis is optimized when integrated with
basin and channel models. This integration consequently enables a more accurate
delineation of flood extents particularly during typhoon events. The models also
produced reliable results as validated by the actual data. This study also confirms the
sufficiency of on-line sources for developing flood models. The methods of this study
can be easily transferable to develop regional flood studies and alert system for hazard
mitigation to other places in the country.
The inclusion of the communities as part of the early warning system itself was
highly successful. While the same rainfall data could have been gathered using
automated weather stations, manually gathered information can be just as accurate as
long as proper instruments and training are provided. The communities’ sense of
ownership of the system also had an impact on their level of understanding of flooding.
This methodology is also promising as it is relatively low cost and relies highly on
volunteerism.
Legend
km
N
Mt Isarog
01234
Ragay Hills
Bicol River
HEC-RAS model
Dartmouth satellite data
Figure 6.
Comparison of modeled
flood extent with the
Dartmouth satellite image
acquired for December 4,
2006
94
DPM
21,1
References
Alho, P., Roberts, M.J. and Kayhko, J. (2007), “Estimating the inundation area of a massive, hypothetical
Jokulhlaup from Northwest Vatnajokull, Iceland”, Natural Hazards, Vol. 41, pp. 21-42.
Bates, P.D. and De Roo, A.P.J. (2000), “A simple raster-based model for flood inundation
simulation”, Journal of Hydrology, Vol. 236, pp. 54-77.
Castrogiovanni, E.M., La Loggia, G. and Noto, L.V. (2005), “Design storm prediction and
hydrologic modeling using a web-GIS approach on a free-software platform”, Atmospheric
Research, Vol. 77, pp. 367-77.
Chow, V.T., Maidment, D.R. and Mays, L.W. (1988), Applied Hydrology, McGraw-Hill, New York, NY.
Chubey, M.S. and Hathout, S. (2004), “Integration of RADARSAT and GIS modeling for
estimating Red River flood risk”, GeoJournal, Vol. 59, pp. 237-46.
Coroza, O., Evans, D. and Bishop, I. (1997), “Enhancing runoff modeling with GIS”, Landscape
and Urban Planning, Vol. 38, pp. 13-23.
Correla, F.N., Rego, F.C., Saraiva, M. and Ramos, I. (1998), “Coupling GIS with hydrologic and
hydraulic flood modelling”, Water Resources Management, Vol. 12, pp. 229-49.
Demirkesen, A.C., Evrendilek, F., Berberoglu, S. and Kilic, S. (2007), “Coastal flood risk analysis
using landsat-7 ETM þImagery and SRTM DEM: a case study of Izmir, Turkey”,
Environmental Monitoring Assessment, Vol. 131, pp. 293-300.
HEC (2008a), Hydrologic Modeling System: Technical Reference Manual, US Army Corps of
Engineers Hydrologic Engineering Center, Davis, CA.
HEC (2008b), River Analysis System: Technical Reference Manual, US Army Corps of Engineers
Hydrologic Engineering Center, Davis, CA.
Horritt, M.S. and Bates, P.D. (2002), “Evaluation of 1D and 2D numerical models for predicting
river flood inundation”, Journal of Hydrology, Vol. 268, pp. 87-9.
Jasrotia, A.S. and Singh, R. (2006), “Modeling runoff and soil erosion in a catchment area, using
the GIS, in the Himalayan region, India”, Environmental Geology, Vol. 51, pp. 29-37.
Knebl, M.R., Yang, Z.-L., Hutchison, K. and Maidment, D.R. (2005), “Regional scale flood
modeling using NEXRAD rainfall, GIS, and HEC-HMS/RAS: a case study for the San
Antonio River Basin Summer 2002 storm event”, Journal of Environmental Management,
Vol. 75, pp. 325-36.
Lastra, J., Fernandez, E., Diez-Herrero, A. and Marqueinez, J. (2008), “Flood hazard delineation
combining geomorphological and hydrological methods: an example in the Northern
Iberian Peninsula”, Natural Hazards, Vol. 45, pp. 277-93.
Lee, K.T., Hung, W.C. and Meng, C.C. (2008), “Deterministic insight into ANN model performance
for storm runoff simulation”, Water Resource Management, Vol. 22, pp. 67-82. DOI 10.1007/
s11269-006-9144-x.
Liu, Y.B. and De Smedt, F. (2005), “Flood modeling for complex terrain using GIS and remote
sensed information”, Water Resources Management, Vol. 19, pp. 605-24.
Liu, Y.B., De Smedt, F., Hoffman, L. and Pfister, L. (2004), “Assessing land use impacts on flood
processes in complex terrain by using GIS and modeling approach”, Environmental
Modeling and Assessment, Vol. 9, pp. 227-35.
Ludwig, R. and Schneider, P. (2006), “Validation of digital elevation models from SRTM X-SAR
for applications in hydrologic modeling”, ISPRS Journal of Photogrammetry & Remote
Sensing, Vol. 6, pp. 339-58.
Machado, S.M. and Sajad, A. (2007), “Flood hazard assessment of Atrato River in Colombia”,
Water Resources Management, Vol. 21, pp. 591-609.
National Disaster Coordinating Council (NDCC) (2004), “‘Typhoon Durian Update’ Pacific
disaster management information network”, available at: www.coe-dmha.org/Durian/
Dur120406.htm (accessed June 25, 2009).
95
Community-
based flood early
warning
Pappenberger, F., Beven, K., Frodsham, K., Romanowicz, R. and Matgen, P. (2007), “Grasping the
unavoidable subjectivity in calibration of flood inundation models: a vulnerability
weighted approach”, Journal of Hydrology, Vol. 333, pp. 275-87.
Philippine Atmospheric Geophysical and Astronomical Services Administration (PAGASA)
(2007), “The Bicol River Basin”, available at: www.pagasa.dost.gov.ph/ffb/pabc.htm
(accessed November 23, 2007).
Pramanik, N., Panda, R.K. and Sen, D. (2009), “One dimensional hydrodynamic modeling of
river flow using DEM extracted river cross-sections”, Water Resources Management,
DOI 10.1007/s11269-009-9474-6.
Sanders, B.F. (2007), “Evaluation of on-line DEMs for flood inundation modeling”, Advances in
Water Resources, Vol. 30, pp. 1831-43.
Schumann, G., Matgen, P., Cutler, M.E.J., Black, A., Hoffmann, L. and Pfister, L. (2008),
“Comparison of remotely sensed water stages from LiDAR, topographic contours and
SRTM”, ISPRS Journal of Photogrammetry and Remote Sensing, Vol. 63, pp. 283-96.
Shahrokhnia, M.A. and Javan, M. (2005), “Performance assessment of Dorrodzan irrigation
network by steady state hydraulic modeling”, Irrigation and Drainage Systems, Vol. 19,
pp. 189-206.
Further reading
Aurelio, M.A. and Pena, R.E. (2002), “Geology and mineral resources of the Philippines”, Department
of Environment and Natural Resources, Mines and Geosciences Bureau, Vol. 1 p. 1-43.
Brakenridge, G.R., De Groeve, T., Nghiem, S.V. and Hiser, R. (2009), “Current flooding display,
Dartmouth flood observatory”, available at: www.dartmouth.edu/Bfloods/ (accessed June
25, 2009).
Ko
¨mu
¨s¸cu, A.U
¨., Erkan, A. and C¸elik, S. (1998), “Analysis of Meteorological and Terrain Features
Leading to the Izmir Flash Flood, 3-4 November 1995”, Natural Hazards, Vol. 18, pp. 1-25.
About the authors
Catherine C. Abon has been a Teaching Instructor/Instructor at the National Institute of
Geological Sciences, University of the Philippines, Diliman, since 2006, and is also studying for
an MS in Geology at the same institution. She also completed her BS in Geology at the National
Institute of Geological Sciences, in 2006. Catherine C. Abon is the corresponding author and can
be contacted at: catherineabon@gmail.com
Carlos Primo C. David has a PhD in Environmental Science and Geology from Stanford
University (1997-2003), and obtained his MS and BS in Geology from the National Institute of
Geological Sciences, University of the Philippines, Diliman. He is currently an Associate
Professor at the National Institute of Geological Sciences, and previously worked as a
hydrologist for the US Geological Survey, and as a Teaching Associate/Instructor at the National
Institute of Geological Sciences, University of the Philippines, Diliman.
Guillermo Q. Tabios III has a PhD in Environmental Science and Geology from Stanford
University (1997-2003), and obtained his MS and BS in Geology from the National Institute of
Geological Sciences, University of the Philippines, Diliman. He held an Associate Professor
position in the Department of Civil Engineering at the University of the Philippines, Diliman
(1995-2008) and was the Director of the National Hydraulic Research Center, University of the
Philippines, Diliman (2006-2009). He is currently a Professor at the Department of Civil
Engineering, University of the Philippines, Diliman.
To purchase reprints of this article please e-mail: reprints@emeraldinsight.com
Or visit our web site for further details: www.emeraldinsight.com/reprints
96
DPM
21,1
