Risk of flooding in the city of Saskatoon in light of climate change and the uncertainty in the associated rainfall intensity estimates
Date
2020-12-13
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
0000-0002-9213-7922
Type
Thesis
Degree Level
Masters
Abstract
Adaption is an efficient measure for coping with climate change impacts and, as such, it has been included in urban planning. Guidelines for adaptation planning suggest updating Intensity Duration Frequency (IDF) curves using climate change models (GCMs). GCMs, however, add uncertainty to projections due to a combination of different factors in the modelling process. To deal with this uncertainty, projections for many cities includes an ensemble of climate models, which gives a range of potential climate changes. However, cities do not appear to have decision-making methodology for choosing a storm for climate change adaptation for new urban drainage systems. First, most of the cities are already developed and adaptation consist of integrated strategies. Second, there are many doubts about the storm event to be used for adaptation. While for many city planners using the average storm intensity might be cost-effective; studies suggest this option might show characteristics that do not correspond to any of the single model outputs, or that the average storm could even result in a lower value than the most likely expected change.
The present study assessed the impacts of a broad range of projected rainfall intensities on the stormwater collections systems in two neighbourhoods in the City of Saskatoon. The objectives of the study include to examine the variation of both construction costs and flood damage costs over the range of the rainfall intensities projected to occur under climate change for the two neighbourhoods; to conduct a cost-benefit analysis to facilitate the selection of the rainfall intensity for design within the climate model results; and to recommend a design storm for the design of the stormwater infrastructure in new developments to adapt to climate change. The two neighbourhoods tested were Willowgrove and Hampton Village.
First, the selection of rainfall intensities from climate change modelling results for testing was conducted. Design storms for modelling were the 1-in-2-year, 1-hour; 1-in-5-year, 1-hour; and 1-in-100-year, 24-hour storms. Rainfall intensities were selected for each storm from the range of intensities projected previously by Nazemi and Goswami (2020). Projections include rainfall intensities for four projected periods: the 2011-2040, 2041-2070, 2071-2100, and 2011-2100.
Next, construction costs for stormwater systems designed to handle these rainfall intensities, based on the current City of Saskatoon (2019a) design standards, were estimated. For this, the minor system was modelled under each minor storm and the major system was modelled under the major storm. The size of different components of the systems was changed to accommodate the runoff water produced for each rainfall intensity. Cost-non-exceedance probability curves were then developed for the minor and major system, separately. From this curves, it was possible to identify a steep increase in construction costs for rainfall intensities associated with non-exceedance probabilities greater than about 0.9. These probabilities all correspond to a rainfall intensity of 16.5 mm/h for the 1-in-2-year, 1-hour storm. For the 1-in-100-year 24-hour major storm these probabilities correspond to rainfall intensities of approximately 132 to 144 mm/day.
Next, flood damage costs in the two neighbourhoods were estimated if the test rainfall intensities were to occur and the stormwater systems had been designed to meet the current City’s standards and design storms. For this, a flooding analysis was conducted for both minor and major storms and flood maps were developed. Once the location of ponding water was identified from flood maps, basic costs of flood-damaged infrastructure were calculated for each tested rainfall event. Then, these costs were related to the non-exceedance values for that rainfall intensity within the climate change model results for the major storm. Flood damage costs increase at an exponential rate in both neighbourhoods. The period 2071-2100 has the largest range of predicted flood damage costs.
Then, a weighted damage analysis was conducted for both neighbourhoods to include the hazard due to the whole range of projected rainfall intensities from climate modelling. The highest weighted damage corresponds to a maximum rainfall intensity of 132 mm/day and 120 mm/day for Willowgrove and Hampton Village respectively, which it is suggested as the minimum rainfall intensity for design to cover the highest hazards. Then, cost-benefit analyses were conducted to examine the feasibility of choosing a rainfall intensity for design of the major system different from that the intensity defined in the City’s current standards. The analysis showed that net present values for choosing a rainfall intensity larger than the defined in the City of Saskatoon standards, show negative values for Willowgrove but positive values for Hampton Village.
It is therefore recommended that one consistent standard be applied for the city that corresponds to a probability of non-exceedance within the climate model results of 0.9. It is recommended that a rainfall intensity of 16.5 mm/h be used for the minor storm. It is also recommended to use a rainfall intensity of 140.8 mm/day for the major storm in design standards.
Description
Keywords
Intensity-Duration-Frequency (IDF) curves, climate change, adaptation, cost assessment, stormwater collection system, design storm, risk assessment, hydraulic modelling
Citation
Degree
Master of Science (M.Sc.)
Department
Civil and Geological Engineering
Program
Civil Engineering