|Thursday, June 03|
Flood Susceptibility Modeling using Geospatial-based Multi-Criteria Decision Making in Ungauged Basins
* Roya Sahraei, University of Tehran, Iran (Islamic Republic of)
Yousef Kanani-Sadat, Canada
Abdolreza Safari, Canada
Saeid Homayouni, Canada
"Floods are frequent natural hazards that threaten human and wild lives and responsible for substantial economic and environmental damages. The frequency and severity of floods have increased during the last decades due to aggregative reasons such as climate change, population growth, deforestation, and human intervention. Floods subsequently can cause other disasters, such as landslides, erosion, ground cavity, etc. Although floods cannot be certainly prevented, flood-prone areas can be modelled and predicted. One of the useful tools that can be used for spatial planning and developing the cities is flood susceptibility maps (FSMs), which aim to provide information about flood-prone areas to establish an early warning system, emergency plan, environmental and water management, and execution of flood management strategies. There are various types of factors that affect the severity of flood events, and they must be considered in analyzing flood susceptibility mapping. Flood conditioning factors with spatial and temporal aspects and different sources and dimensions can be provided by remote sensing and analyzed using geographic information systems. The integration of these efficient technologies has made a fundamental contribution to the multidimensional analysis of natural hazards such as landslides, groundwater resources, and flood susceptibility. This study aimed to implement an analytical framework to identify flood potential zones in large scale ungauged areas. We applied this framework to model the flood susceptibility in a region located in southeastern Iran. This region has been dealing with data scarcity, and despite severe flood events, there is no study regarding flood susceptibility assessment. We used the Multi-Criteria Decision Making (MCDM) approach combined with geospatial analysis and remote sensing observations for this susceptibility analysis. Nine flood conditioning factors that have an impact on flood behaviour were selected and used for the susceptibility modelling, namely: Slope, Digital Elevation Model (DEM), Normalized Difference Vegetation Index (NDVI), Curve Number (CN), Distance to the river (DR), Topographic Wetness Index (TWI), Topographic Position Index (TPI), Flow Accumulation (FA), and Modified Fournier Index (MFI). At first, the criteria’s weights were estimated using the Analytic Hierarchy Process (AHP) approach and based on experts’ knowledge. The resultant weights reveal that FA, TWI, and DR were the most influential flood susceptibility criteria. After calculating these weights, the criteria’s layers and their weights were aggregated through geospatial analysis, which resulted in generating FMS. The final flood susceptibility map was classified into five susceptibility categories to ease the interpretation of estimated flooding maps. The historical flood points were then used to evaluate the FSMs using the area under the receiver operating characteristic curve (AUROC) and statistical measures such as the Kappa index. Validation results (AUROC= 94.53 and Kappa=76.35) show that the implemented method is efficient. Moreover, it was observed that 74.29% of the historically flooded areas are in the “high” and “very high” classes of the FSMs. These susceptibility maps can help the decision-makers and managers to allocate the mitigation equipment and facilities in data-scarce and ungauged large-scale areas. "
How is the overall uncertainty on flood projections influenced by local hydroclimatic conditions?
* Mariana Castaneda-Gonzalez, École de technologie supérieure, Canada
Annie Poulin, Canada
Rabindranarth Romero-Lopez, Canada
Richard Turcotte, Canada
The uncertainties associated to the simulation of flooding scenarios remain a challenge for assessing future potential risks and implementing adaptation and mitigation strategies. Various efforts have been made to identify the main uncertainty sources by using large ensembles of models and methodologies at the basin scale. However, the different methodological choices and limited diversity of studied basins make difficult to generalize the findings over different studies and from contrasting regions. Thus, we investigate how the uncertainty contribution of different elements of the hydroclimatic modelling chain are impacted by the hydroclimatic regimes of the region. Four uncertainty sources are evaluated, (1) climate simulations, (2) post-processing methods, (3) hydrological models, and (4) probability distributions for return levels estimation, over 96 diverse North American basins. A GCM-ensemble (twenty-two simulations) and a CRCM-ensemble (three simulations) are coupled with two postprocessing methods, three lumped hydrological models and six probability distributions to estimate six flood indicators (2-, 5-, 10-, 20-, 50-, and 100-year return periods) per dataset. Uncertainty contributions on these six flood indicators are quantified through a variance decomposition approach for different seasons and periods (1976-2005, 2041-2070 & 2070-2099). Results indicate that uncertainty contributions were systematically influenced by the basin’s hydroclimatic conditions, especially over the future periods. Basins where flooding is dominated by snowmelt showed larger uncertainty contributions from hydrological models, while rainfall-dominated basins showed larger uncertainty contributions from the climate simulations ensembles. The results also showed that uncertainty contributions were sensitive to the season, period and flood indicator. These findings provide some insights of the link between uncertainty contribution and the dominant hydrological process of each basin, highlighting the need to consider their influence on flooding projections.
Integrated flood impact modeling system for the Ottawa River
* Hamza Ousoukhman, Polytechnique Montréal, Canada
Ahmad Shakibaeinia, Canada
Elmira Hassanzadeh, Canada
This study presents an integrated hydraulic-hydrologic modeling system for flood impact prediction. In this system, the Delft3D two-dimensional hydrodynamic model is connected with a hydrologic model and observation data to provide an automatic exchange of data and results. The weather data and watershed characteristics provide input to the hydrological model to predict streamflow conditions, which are then automatically fed into the hydrodynamic model. The hydrodynamic model simulates the flood characteristics e.g., water level, 2D depth-averaged velocity field, and flood extent. The system is applied to the Ottawa River. The bathymetric data of various sources are combined and interpolated to the Delf3D hydrodynamic model. The hydrodynamic model is validated against the measured water level and velocity profiles, showing a good agreement. This model is then coupled with a hydrological model and data sources through a modular unified modeling platform. The final system is tested for a real-time flood impact forecasting scenario.
A North American flood model for spatial dependence and diversification problems
* Manuel Grenier, Université du Québec à Montréal, Canada
"Authors and affiliations: Manuel Grenier*, M.Sc. Student of Actuarial Science Mathieu Boudreault*, Associate Professor of Actuarial Science David Carozza*, Postdoctoral Fellow *Department of Mathematics, Université du Québec à Montréal, Montréal (Québec), Canada Abstract: Following the Calgary flooding in 2013, the Canadian insurance industry has gradually introduced overland flood insurance products since 2015. Sold as an optional rider to the basic homeowner’s insurance, nowadays nearly 80% of insurers offer such protection with an overall take-up rate of about 30-40%. However, the profitability and long-term stability of flood insurance, in addition to the industry’s participation to risk-sharing with governments, depends on its ability to geographically diversify its portfolio exposure and how and where climate change may impact the latter. To replicate large scale spatial dependence and determine its impacts on portfolio flood risk management, we design a North American flood model driven by climate and hydrology to represent flood occurrences and impacts in terms of population displaced. The North American flood model is fitted to past observations and used to simulate large forward-looking event sets using the Canadian Regional Climate Model (CRCM) over time horizons representing present (1981-2020) and future climates (2021-2060, 2061-2100). Using statistical and machine learning techniques, we link flood occurrence and population displaced (Darthmouth Flood Observatory, Brakenridge et al., 2010) over Pfafstetter level 8 basins, to past precipitation and temperature patterns (Funk et al., 2015 & Shi, 2007), land use data (ESI, 2017), as well as basin characteristics (Linke et al., 2019) to explain how flooding may occur at this spatial scale. Using outputs from the event sets, here we will present how spatial dependence affects financial management of flood risk in North America and discuss how climate change may impact flood insurance in both Canada and the United States. References: Brakenridge, G.(2010). Global active archive of large flood events. Dartmouth Flood Observatory, University of Colorado. Retrieved from http://floodobservatory.colorado.edu/Archives/index.html ESA. Land Cover CCI Product User Guide Version 2. Tech. Rep. (2017). Available at: maps.elie.ucl.ac.be/CCI/viewer/download/ESACCI-LC-Ph2-PUGv2_2.0.pdf Funk, C., Peterson, P., Landsfeld, M. et al. The climate hazards infrared precipitation with stations—a new environmental record for monitoring extremes. Sci Data 2, 150066 (2015). https://doi.org/10.1038/sdata.2015.66 Linke, S., Lehner, B., Ouellet Dallaire, C. et al. Global hydro-environmental sub-basin and river reach characteristics at high spatial resolution. Sci Data 6, 283 (2019). https://doi.org/10.1038/s41597-019-0300-6 Shi, W. (2007). Global daily surface air temperature analyses. Presentation. Retrieved from ftp://ftp.cpc.ncep.noaa.gov/precip/PEOPLE/wd52ws/892globaltemp/CPC-GLOBAL-T.pdf "