Reliability Study of Shoreline Change Using Monte Carlo Method

advertisement
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
Time
Programme
Chair
08:00-08:30 Registration
08:30-08:40 Opening remark
08:40-09:00 Group photo
09:00-09:30 Uncertainty analysis of flood mapping by using
satellite precipitation and hydrologic models
Pao-Shan Yu, NCKU, Taiwan
09:30-10:00 Mustafa Altinakar, NCCHE, USA
10:00-10:30 Assessing the element of surprise of recordbreaking flood events
Thomas Kjeldsen, University of Bath, UK
Keh-Chia Yeh,
NCTU, Taiwan
10:30-10:50 Break
10:50-11:20 Extensive monitoring of sediment transport for
reservoir sediment management
Chih-Ping Lin, NCTU, Taiwan
11:20-11:50 Development and validation of CCHE2D dam
break process model
Yafei Jia, NCCHE, USA
Wen-Cheng Liu,
TTFRI, Taiwan
11:50-12:20 The analysis and application of artificial neural
networks for early warning systems in floodrelated applications
Andrew P Duncan, University of Exeter, UK
12:20-13:30 Lunch
13:30-14:00 Research of radar sciences and engineering at the
university of oklahoma – Advanced Radar
Research Center (ARRC)
Tian-You Yu, University of Oklahoma, USA
14:00-14:30 Integrated Coastal Process Modeling and Impact
Assessment of Flooding and Sedimentation due to Mustafa Altinakar,
Typhoons in Taiwan
NCCHE, USA
Yan Ding, NCCHE, USA
14:30-15:00 The application of ensemble rainfall forecasts to
social-economic impact assessment during
emergency response
Jiun-Huei Jang, NCDR, Taiwan
15:00-15:20 Break
15:20-15:50 Towards efficient modeling
Yaoxin Zhang, NCCHE, USA
15:50-16:20 Linking fluvial and landslide erosions along a
Thomas Kjeldsen,
University of Bath,
UK
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
meandering river in Southern Taiwan
Yi-Chin Chen, NCUE, Taiwan
16:20-16:50 Investigation of the evolution of riverbed and pier
scour depths by using water-surface velocity radar
and wireless tracers
Jian-Hao Hong, TTFRI, Taiwan
16:50-17:00 Closing remark
Meeting Venue
Hao-Ran International Conference Hall, National Chiao Tung University Library
Phone: 03-5712121 # 52636
http://www.lib.nctu.edu.tw/
Local Contact Persons
Josh Yang, Ph.D.
Taiwan Typhoon and Flood Research Institute
National Applied Research Laboratories
Phone: 02-23219660 #118
Email: tshyang@narlabs.org.tw
Chung-Ta Liao, Ph.D.
Disaster Prevention & Water Environment Research Center
National Chiao Tung University
Phone: 03-5712121 #55268
Email: zeromic@gmail.com
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
Speaker List
Pao-Shan Yu,
Professor & Dean, Department
of Hydraulic and Ocean
Engineering, National Cheng
Kung University, Taiwan
Thomas Kjeldsen,
Senior Lecturer, Department of
Architecture and Civil
Engineering, University of
Bath, UK
Mustafa Altinakar,
Director, National Center for
Computational Hydroscience
and Engineering, USA
Chih-Ping Lin,
Distinguished Professor,
Department of Civil
Engineering, National Chiao
Tung University, Taiwan
Yafei Jia,
Research Professor and
Assistant Director, National
Center for Computational
Hydroscience and Engineering,
USA
Andrew P Duncan,
Associate Research Fellow,
University of Exeter Centre for
Water Systems, UK
Tian-You Yu,
Professor, School of Electrical
and Computer Engineering,
Advanced Radar Research
Center, and School of
Meteorology, University of
Oklahoma, USA
Yan Ding,
Research Associate Professor,
National Center for
Computational Hydroscience
and Engineering, USA
Jiun-Huei Jang, Assistant
Division Head, National Science
and Technology for Disaster
Reduction, Taiwan
Yaoxin Zhang, Research
Scientist, National Center for
Computational Hydroscience
and Engineering, USA
Yi-Chin Chen, Assistant
Professor, Department of
Geography, National Changhua
University of Education, Taiwan
Jian-Hao Hong, Associate
Researcher, Taiwan Typhoon
and Flood Research Institute,
Taiwan
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
UNCERTAINTY ANALYSIS OF FLOOD MAPPING BY USING SATELLITE
PRECIPITATION AND HYDROLOGIC MODELS
Pao-Shan Yu1, Soroosh Sorooshian2, Cheng-Shang Lee3, Kuo-Lin Hsu4, Tao-Chang Yang5,
Chen-Min Kuo6, Hung-Wei Tseng7
Ph.D., Dean and Distinguished Professor, Hydraulic and Ocean Engineering, National Cheng Kung University,
Taiwan. Email: yups@mail.ncku.edu.tw
2Ph.D.,
Distinguished Professor, Civil and Environmental Engineering, University of California, Irvine, Irvine, USA.
Email: soroosh@uci.edu
3Ph.D.,
Professor, Atmospheric Science, National Taiwan University, Taiwan. Email: cslee@as.ntu.edu.tw
4Ph.D.,
Associate Professor, Civil and Environmental Engineering, University of California, Irvine, Irvine, USA.
Email: kuolinh@uci.edu
5Ph.D.,
Associate Research Professor, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan.
Email: tcyang58@hotmail.com
6Ph.D.,
Assistant Research Fellow, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan.
Email: jemkuo@mail.ncku.edu.tw
7Ph.D.,
Post-Doctoral Fellow, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan. Email:
hungwei1618@gmail.com
This study aims at proposing an approach to apply WRF (Weather Research and Forecasting
Model) rainfall forecasting, radar rainfall and satellite rainfall to physiographic inundationdrainage model for providing a real-time flood forecasting of Dianbao River in Taiwan. The
Dianbao River is a low-relief catchment which is easily affected by the flood disaster. Since the
lacks of reliable rainfall forecasting and inundation model, this study tried to derive a selection
strategy to refine the rainfall forecasting for better flood simulation.
Various WRF rainfall forecasting results provided by Taiwan Typhoon and Flood Research
Institute (TTFRI) are used in this study. WRF can provide 78hr forecasting, but the results among
different models are quite different due to their non-isolated boundary condition. Thus, the realtime radar rainfall and satellite rainfall can be used to verify the estimation of WRF. Once the
WRF estimations are reliable, the WRF forecasting results can be used to derive the flood
inundation depth for the study area. So, the chosen of WRF is the key step for the flood
estimation. This study integrated QPESUMS radar rainfall and PERSIANN satellite rainfall to
provide better rainfall forecast. The idea is picking up the available WRF rainfall forecasting
form PERSIANN or QPESUMS in sea area while the typhoon has been generated. Base on the
6hr-delay rainfall forecasting from 21 sets of WRF model, a pattern recognition method is used to
compare the PERSIANN observation to the WRF forecasting for the same time period in every
6hr. With assigning some weighting factors for the 7-12hr WRF rainfall forecasting base on the
error between WRF and PERSIANN, we can generate the reliable rainfall forecasting. Also, we
may select some reliable rainfall forecasting results for uncertainty analysis. Through the flood
inundation map produced by physiographic inundation-drainage model, decision makers can
identify flood prone areas and make emergency preparedness.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
UNCERTAINTY ANALYSIS OF FLOOD MAPPING BY USING SATELLITE
PRECIPITATION AND HYDROLOGIC MODELS
Pao-Shan Yu1, Soroosh Sorooshian2, Cheng-Shang Lee3, Kuo-Lin Hsu4, Tao-Chang Yang5,
Chen-Min Kuo6, Hung-Wei Tseng7
Ph.D., Dean and Distinguished Professor, Hydraulic and Ocean Engineering, National Cheng Kung University,
Taiwan. Email: yups@mail.ncku.edu.tw
2Ph.D.,
Distinguished Professor, Civil and Environmental Engineering, University of California, Irvine, Irvine, USA.
Email: soroosh@uci.edu
3Ph.D.,
Professor, Atmospheric Science, National Taiwan University, Taiwan. Email: cslee@as.ntu.edu.tw
4Ph.D.,
Associate Professor, Civil and Environmental Engineering, University of California, Irvine, Irvine, USA.
Email: kuolinh@uci.edu
5Ph.D.,
Associate Research Professor, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan.
Email: tcyang58@hotmail.com
6Ph.D.,
Assistant Research Fellow, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan.
Email: jemkuo@mail.ncku.edu.tw
7Ph.D.,
Post-Doctoral Fellow, Hydraulic and Ocean Engineering, National Cheng Kung University, Taiwan. Email:
hungwei1618@gmail.com
This study aims at proposing an approach to apply WRF (Weather Research and Forecasting
Model) rainfall forecasting, radar rainfall and satellite rainfall to physiographic inundationdrainage model for providing a real-time flood forecasting of Dianbao River in Taiwan. The
Dianbao River is a low-relief catchment which is easily affected by the flood disaster. Since the
lacks of reliable rainfall forecasting and inundation model, this study tried to derive a selection
strategy to refine the rainfall forecasting for better flood simulation.
Various WRF rainfall forecasting results provided by Taiwan Typhoon and Flood Research
Institute (TTFRI) are used in this study. WRF can provide 78hr forecasting, but the results among
different models are quite different due to their non-isolated boundary condition. Thus, the realtime radar rainfall and satellite rainfall can be used to verify the estimation of WRF. Once the
WRF estimations are reliable, the WRF forecasting results can be used to derive the flood
inundation depth for the study area. So, the chosen of WRF is the key step for the flood
estimation. This study integrated QPESUMS radar rainfall and PERSIANN satellite rainfall to
provide better rainfall forecast. The idea is picking up the available WRF rainfall forecasting
form PERSIANN or QPESUMS in sea area while the typhoon has been generated. Base on the
6hr-delay rainfall forecasting from 21 sets of WRF model, a pattern recognition method is used to
compare the PERSIANN observation to the WRF forecasting for the same time period in every
6hr. With assigning some weighting factors for the 7-12hr WRF rainfall forecasting base on the
error between WRF and PERSIANN, we can generate the reliable rainfall forecasting. Also, we
may select some reliable rainfall forecasting results for uncertainty analysis. Through the flood
inundation map produced by physiographic inundation-drainage model, decision makers can
identify flood prone areas and make emergency preparedness.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
ASSESSING THE ELEMENT OF SURPRISE OF RECORD-BREAKING
FLOOD EVENTS
Thomas Kjeldsen1
Ph.D., Senior Lecturer, Department of Architecture and Civil Engineering, University of Bath, UK.
Email: T.R.Kjeldsen@bath.ac.uk
The occurrence of record-breaking flood events continuous to cause damage and disruption
despite significant investments in flood defences, suggesting that these events are in some sense
surprising. This study develops a new statistical test to help assess if a flood event can be
considered surprising or not. The test statistic is derived from annual maximum series (AMS) of
extreme events, and Monte Carlo simulations were used to derive critical values for a range of
significance levels based on a Generalized Logistic distribution. The method is tested on a
national dataset of AMS of peak flow from the United Kingdom, and is found to correctly
identify recent large event that have been identified elsewhere as causing a significant change in
UK flood management policy. No temporal trend in the frequency or magnitude of surprising
events was identified, and no link could be established between the occurrences of surprising
events and large-scale drivers.
EXTENSIVE MONITORING OF SEDIMENT TRANSPORT FOR RESERVOIR
SEDIMENT MANAGEMENT
Chih-Ping Lin1
1
Ph.D., Distinguished Professor, Department of Civil Engineering, National Chiao Tung University, Hsinchu,
Taiwan. Email: cplin@mail.nctu.edu.tw
Sedimentation is a serious threat to long-term water resource management worldwide. In
particular, reservoir sedimentation is becoming more serious in Taiwan due to geological
weathering and climate change in watersheds. Large amount of sediments transports to reservoirs
during storm events at hyperpycnal concentration. Full-event monitoring of sediment transport in
a reservoir plays an important role in sustainable reservoir management. This presentation begins
by reviewing existing surrogate techniques in need for monitoring suspended-sediment transport
with high concentration range and wide spatial coverage. More commercially available
techniques suffer from particle size dependency and limited measurement range. A relatively new
technique based on time domain reflectometry is introduced. It possesses several advantages,
including particle-size independence, high measurement range, durability, and cost-effective
multiplexing. Its application in an extensive SSC monitoring program for reservoir management
is demonstrated through a case study in Shihmen reservoir, Taiwan. Monitoring stations were
installed at the major inflow river mouth and outlet works with fixed protective structures to
provide inflow and outflow sediment-discharge records. To capture the characteristics of density
currents, a multi-depth monitoring station was designed and deployed on floating platforms in the
reservoir. Some of the data collected during Typhoons are presented as an example to
demonstrate the effectiveness and benefits of the TDR-based monitoring program.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
DEVELOPMENT AND VALIDATION OF CCHE2D DAM BREAK PROCESS
MODEL
Yafei Jia1, Yaoxin Zhang2
1
Ph.D., Research Professor and Assistant Director, National Center for Computational Hydroscience and Engineering,
USA. Email: jia@ncche.olemiss.edu
2
Research Scientist, National Center for Computational Hydroscience and Engineering, USA.
Email: yzhang@ncche.olemiss.edu
Flooding due to earth embankment breaching often results in detrimental impact on the people
and their properties in the flooding zone. The embankment breaching is often caused by
overtopping of the excessive water in a reservoir/river; the breaching process is dominated by the
shape, soil property of the embankment and the flow conditions of the reservoir/river.
A practical numerical simulation model for overtopping embankment breaking process is
developed in this study. Because we would like the model to represent the hydrodynamics and
soil erosion processes as much as possible, , the key physical-empirical dam breaking
mechanisms of earth embankment are adopted and implemented into a depth integrated free
surface flow model CCHE2D (Jia, et al. 2002). A special function representing the shape of the
breaching channel profile is introduced which greatly simplifies the effort of modeling. The shear
stress of the flow over the cohesive earth of the breaching embankment is simulated by the 2D
flow model.
The developed model is validated using experiment data collected by Henson et al. (2005). The
simulated flooding hydrograph, headcut migration and breaching embankment profiles agree with
the observation very well. For general users’ applications, the graphic user interface of the
CCHE2D model is modified to include this capability of the developed model. Because the
breaching model is associated to a 2D general flow model, it is possible to simulate multiple
embankment breachings in general and complex flow situations; the applicability of the dambreak models are thus broadened significantly.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
THE ANALYSIS AND APPLICATION OF ARTIFICIAL NEURAL
NETWORKS FOR EARLY WARNING SYSTEMS IN FLOOD-RELATED
APPLICATIONS
Andrew Paul Duncan1
1
Ph.D., Associate Research Fellow, University of Exeter Centre for Water Systems, UK.
Email: apd209@exeter.ac.uk
Artificial Neural Networks (ANNs) have been comprehensively researched, both from a computer
scientific perspective and with regard to their use for predictive modelling in a wide variety of applications
including flood prediction and the environment. Yet their adoption for live, real-time systems remains on
the whole sporadic and experimental. A plausible hypothesis is that this may be at least in part due to their
treatment heretofore as “black boxes” that implicitly contain something that is unknown, or even
unknowable. It is understandable that many of those responsible for delivering Early Warning Systems
(EWS) might not wish to take the risk of implementing solutions perceived as containing unknown
elements, despite the significant computational advantages that ANNs offer. This presentation therefore
builds on existing efforts to open the box and develop tools and techniques that visualise, analyse and use
ANN weights and biases especially from the viewpoint of neural pathways from inputs to outputs of
feedforward networks. In so doing, it aims to demonstrate novel approaches to self-improving predictive
model construction for both regression and classification problems. This includes Neural Pathway
Strength Feature Selection (NPSFS), which uses ensembles of ANNs trained on differing subsets of data
and analysis of the learnt weights to infer degrees of relevance of the input features and so build simplified
models with reduced input feature sets and improved predictive performance.
Case studies are carried out for prediction of flooding at multiple nodes in urban drainage networks
located in three urban catchments in the UK, which demonstrate rapid, accurate prediction of flooding
both for regression (flood depths and volumes) and classification (severity levels). By exploiting the
similarities between hydrograph shapes at different nodes in the sewer network we demonstrate that it is
possible to build predictive models for multiple sewer nodes using a single multi-output ANN. Predictive
skill is shown to reduce beyond the time of concentration of each sewer node, when actual rainfall is used
as input to the models – indicating the need for predictions of rainfall to be used to achieve operationally
useful prediction times.
Results from ANN model ensembles generally exhibit improved performance, when compared with single
ANN models. Also ensembles with reduced input feature sets, using NPSFS, demonstrate as good or
improved performance when compared with the full feature set models.
Conclusions are drawn about a new set of ANN-based tools and techniques, including NPSFS and
visualisation techniques for inspection of ANN weights, the adoption of which it is hoped may lead to
improved confidence in the use of ANN for live real-time flood-related Early Warning System
applications.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
RESEARCH OF RADAR SCIENCES AND ENGINEERING AT THE
UNIVERSITY OF OKLAHOMA – ADVANCED RADAR RESEARCH CENTER
(ARRC)
Tian-You Yu1
1
Ph.D., Professor, School of Electrical and Computer Engineering, Advanced Radar Research Center, and School of
Meteorology, University of Oklahoma, USA. Email: tyu@ou.edu
The Advanced Radar Research Center (ARRC) at the University of Oklahoma was established in
2005 by leveraging the legacy of weather radar research, development, application, and operation
in Norman achieved through unique synergy among university, government laboratories and
private sectors. Since then, the ARRC has grown to be arguably one of the largest academic
research centers in the US focused on advancements in weather radar. Currently, ARRC has 18
faculty members, eight engineers in radar software, hardware, and mechanical design and
development, five administrative staff and more than 100 undergraduate and graduate students,
postdocs, and visiting scholars from meteorology, hydrology, and engineering. The ARRC’s
mission is solving challenging radar research problems, preparing students to become the next
generation of scientists and engineers, and serving to empower economic growth and
development in the field of weather radar. Recently, ARRC has expended its portfolio to include
radar research in the areas of Defense, Security and Intelligence. ARRC’s areas of emphasis exist
in rapid hardware prototyping, advanced signal processing, antennas, hydrometeorology, clear-air
sensing, UAS sensors, severe weather, applied electromagnetics, and microwave engineering.
Through the collaborative nature instilled in its members, the ARRC has proven effective at
developing synergy between science and engineering in the field of radar. In the National
Weather Center and in its extensive laboratory and radar facilities, meteorology, hydrology and
engineering faculty and students work side-by-side to learn from each other and to tackle
challenging problems in remote sensing, microwave engineering, and applied electromagnetics.
In this presentation, the ARRC facilities will be briefly introduced and the ARRC’s areas of
expertise will be discussed with the focus on the innovative development and application of
radars to mitigate the impact of natural hazards such as severe storms, flash flood, etc.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
INTEGRATED COASTAL PROCESS MODELING AND IMPACT
ASSESSMENT OF FLOODING AND SEDIMENTATION DUE TO TYPHOONS
IN TAIWAN
Yan Ding1
1
Ph.D., Research Associate Professor, Interim Associate Director of Administration, National Center for
Computational Hydroscience and Engineering, The University of Mississippi, University, MS 38677, USA.
Email: ding@ncche.olemiss.edu
Hazardous storms and typhoons/hurricanes can be devastating by causing flooding water
inundations, shoreline erosions, navigation channel refilling, and casualties. For risk analysis and
coastal protection planning, it is essential to simulate and predict multiscale physical processes
during storms due to severe changes in waves, storm surges, sea levels, river flooding flows, and
sediment transport and morphology from rivers, estuaries, to coasts. For planners and decisionmakers to assess socio-economic and environmental impacts of extreme tropical storms and
climate changes, integrated coastal process modeling has become a major approach to facilitate
multiple-purpose engineering practices in developing cost-effective coastal flood management
plans, as well as designing erosion control structures.
This presentation gives a brief review on integrated coastal process modeling including an
integrated modeling system, CCHE2D-Coast, developed in the National Center for
Computational Hydroscience and Engineering. Then it focuses on the application cases of this
model for coastal and estuarine planning by assessing the impact of coastal flooding and
sedimentation due to typhoons which made landfalls in Taiwan. It demonstrates model validation
by simulating hydrodynamic and morphodynamic processes (sediment transport and bed changes)
due to recent multiple typhoons in the Tamsui and Touchien Estuaries. Simulation results of the
model serve the purpose of assessing coastal flooding risks, navigation channel refilling, and
shoreline erosions, and identifying cost-effective engineering protection plans in the estuaries.
THE APPLICATION OF ENSEMBLE RAINFALL FORECASTS TO SOCIALECONOMIC IMPACT ASSESSMENT DURING EMERGENCY RESPONSE
Jiun-Huei Jang1
1
Ph.D., Assistant Division Head, National Science and Technology for Disaster Reduction, Taiwan.
Email: jamesjang@ncdr.nat.gov.tw
Typhoon Soudelor brought tremendous rainfall to Taiwan during 8/7—8/8 in Aug., 2005, causing
severe river water surge, flash flooding and debris flow to the north area of Taiwan. Taking
Soudelor as an example, this study investigates the application of grid-based ensemble rainfall
forecast products to evaluate the socio-economic impacts related to meteo-hydrological disasters
in real time. Efforts are specially put on the discussion of prediction uncertainty and accuracy
from the angle of lead-time in emergency responses.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
TOWARDS EFFICIENT MODELING
Yaoxin Zhang1, Yafei Jia2
1
Research Scientist, National Center for Computational Hydroscience and Engineering, USA.
Email: yzhang@ncche.olemiss.edu
2
Ph.D., Research Professor and Assistant Director, National Center for Computational Hydroscience and Engineering,
USA. Email: jia@ncche.olemiss.edu
CFD (Computational Fluids Dynamics) analyses are playing more and more important roles in
water resource related problems. However the analyses are often time-consuming when they are
applied to large-scale, long-term, and computation-intensive problems. The situation becomes
more challenging in the mega data time, when a huge amount of data requires efficient
computation, treatment and interpretation. CCHE modeling system developed at NCCHE
(National Center for Computational Hydro-science and Engineering) is a widely used software
for water resource problems. Recently, advanced models and techniques have been developed to
enhance its efficiency. The following modeling enhancements will be presented in the workshop.
1) CCHE1D channel network model: when seeking the average hydrology/hydraulic
solutions, 1D model is the most suitable and efficient tool for large-scale and long-term
problems. The newly developed CCHE1D GUI is focused on 64-bit version to satisfy
users’ mega data requirements.
2) Sub-domain method: 2D and 3D are generally more accurate modeling methods. To
improve their computation efficiency for large domains, a novel sub-domain method has
been developed. The method allow us to obtain dense and sophisticated 2D/3D results
efficiently at the interested locations in a large domain covered with a coarse mesh.
Parallel computing: as an important technology and standard solution, parallel computing is able
to significantly enhance efficiency by distributing computation loads to multiple processers
working at the same time. In CCHE modeling system, a 2D parallel computing module for
GPGPU (General-purpose computing on Graphics Processing Unit) has been developed for both
general flow and dam-break flow/flooding simulations. Applications have demonstrated its high
computing efficiency.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
LINKING FLUVIAL AND LANDSLIDE EROSIONS ALONG A
MEANDERING RIVER IN SOUTHERN TAIWAN
Yi-Chin Chen1, Kang-Tsung Chang2, Jui-Yi Ho3
1
Ph.D., Assistant Professor, Department of Geography, National Changhua University of Education, Taiwan.
Email: yichinchen@cc.ncue.edu.tw
2
Ph.D., Professor, Department of Geography, National Taiwan University, Taiwan. Email: ktchang@ntu.edu.tw
3
Ph.D., Assistant Researcher, Hydrotech Division, Taiwan Typhoon and Flood Research Institute, National Applied
Research Laboratories, Taiwan. Email: juiyiho@narlabs.org.tw
Fluvial erosion is an important geomorphologic process that induces both vertical incision on the
stream bed and lateral erosion on the toe of adjacent hillslopes. After the materials are removed,
rock fall and landslide are triggered on hillslopes. Although it has been long expected that fluvial
erosion can trigger landslide, relatively few studies have been conducted on quantifying the
effects of fluvial erosion on landslide and the application of these factors to landslide prediction.
This study linked the effects of fluvial erosion on landslide erosion in the particular meanders
landscape in the Jhoukou river watershed, southern Taiwan. A semi-automatic model was
developed to extract various fluvial factors, e.g. sinuosity, stream power index, and stream order,
and to build the spatial linkage of river to hillslope by using geographic information system
techniques. To quantify landslide erosion rate, the area of landslides for 11 events in 2001-2010
were mapped from satellite images or orthophotos, and a volume-area relation was then used to
estimate the landslide volume. The results showed that stream sinuosity, slope gradient, and unit
stream power were significantly correlated with the landslide erosion rate. The rates on the
undercut, slip-off, and head-valley hillslopes were 36.9 mm/yr, 26.5 mm/yr, and 30.4 mm/yr,
respectively, and were different significantly among each other. Moreover, landslide erosion rates
increased with sinuosity or stream order on the undercut slope, but decreased on the slip-off slope.
This suggests that the effects of fluvial erosion play a more important role on the meandering or
downstream river than that on the straight or upstream river by eroding the materials on the
undercut slopes and depositing sediment on the slip-off slopes. Furthermore, comparing the
infinite slope models with and without using fluvial erosion factors, the usage of the fluvial
erosion factors can improve the model performance, especially for the downstream area. This
study highlights the need to understand more about the fluvial effects on landslide and
topography evolution in mountainous areas.
International Workshop on Computation, Uncertainty and Risk Assessment in Hydroscience and
Engineering
INVESTIGATION OF THE EVOLUTION OF RIVERBED AND PIER SCOUR
DEPTHS BY USING WATER-SURFACE VELOCITY RADAR AND
WIRELESS TRACERS
Jian-Hao Hong1
1
Ph.D., Associate Researcher, Hydrotech Division, Taiwan Typhoon and Flood Research Institute, National Applied
Research Laboratories, Taiwan. Email: dinohong@narlabs.org.tw
Scour around bridge piers and along river reaches has long been an intrigue topic for researchers.
Its development, especially during floods or some other large hydrological events, has
particularly received plenty of attention. Most research focuses on either building a prediction
model or developing a scour monitoring system. However, due to the difficulties in obtaining
scour measurements during floods or spontaneous simulations from numerical models, it always
poses a great challenge for the administrators or agencies in the right timing for bridge closure or
re-opening. This study conducted field measurements of bridge pier and channel bed scour at
Mingchu Bridge which crosses the middle section of the Choshui River in Taiwan. Numbed
bricks and wireless tracers were used to measure the maximum scour depth and temporal
variations of the scour depths during floods. A surface velocity radar and a water-level gauge
were also installed on the bridge deck to obtain flow information. Scour data were collected
separately during a monsoon and Typhoon Matmo in 2014, with the respective peak flow
discharges of 1,446 and 4,980m3/s. The corresponding maximum general and pier scour depths
reached 1.76 m and 2.53 m during the monsoon, and 3.245 m and 4.125 m during the typhoon. A
quick estimation algorithm for temporal variations of general scour depth was developed, based
on the effectively cumulative stream power concept and calibrated by using the field data.
Temporal variations of total pier scour depth then could be determined by superimposing the
estimation on the local pier scour depth. By examining with the data from these two events, the
results showed reasonable agreement with the field measurements. With the quick estimation
developed in this study, it would be possible to install guidelines for river and bridge
management. More field data are needed to further test the reliability and capability of the
proposed method, and a more robust scour monitoring system shall be developed in the future.
Download