Are More 100 Year Storm Happening? Yes and No. A Proliferation of Rain Gauges Can Now Record More 100 Year Storms, But Fixed Locations Show No Increase

There are many sensational media stories about ghost storms and ninja storms hitting urban areas, and a steady claim that we are experiencing more extreme rainfall, that is, higher intensities for a given probability (called return period), or greater frequency of given design intensities. Often it is stated that we are experiencing more 100 year storms today and that is a "new normal" brought on by a changing climate.

How does the number of climate stations, or rain gauges, that are in operation affect the number of observed extreme events. Well, let's look at Toronto for example.  Several past extreme events were reported in the Staff Report on Impact of July 8, 2013 storm on the City's Sewer and Stormwater Systems dated September 6, 2016: (https://www.toronto.ca/legdocs/mmis/2013/pw/bgrd/backgroundfile-61363.pdf)

During the May 12, 2000 extreme rainfall event, Toronto operated 16 rain gauges as shown on the staff report map below.

Fifteen years later, during the August 19, 2005 storm, the City operated 31 rain gauges as shown below, so almost double the number of rain gauges.  Look at the higher density of gauges in north Toronto where many higher August 19, 2005 rainfall depths were observed.

Then 8 years later, during the July 8, 2013 storm the city operated even more rain gauges, i.e., 35 in total.

And then a few years later, on August 7, 2018, the city operated 43 rain gauges - even more than 2013. I don't have a map but here is a super-cool graph summarizing Toronto Open Data rainfall totals at those gauges over a period of 5 minutes to 24 hours.

And now today as of July 17, 2019, Toronto has 45 active rain gauges as shown in the following map presented to the Ministry of Environment Conservation and Parks' stormwater stakeholder group participating in development of minimum standards for ECA pre-approval.

So let us summarize the trend in the number of rain gauges in the chart below.

Astute blog readers will notice that the number of rain gauges has increased almost 300% since the year 2000. Yes, almost three times the number of rain gauges now. Obviously, more extreme events can be observed and recorded when the number of rain gauges increases dramatically.

The following table shows that in the year 2000, there was a rain gauge every 39.4 square kilometres (16 gauges per 630.2 square kilometres). By 2019, there is a gauge ever 14 square kilometres.

***

So what is happening at fixed locations where rain intensities are measured? In Toronto and Mississauga, many trends are downward according to the Engineering Climate Datasets:

 

As a result, design intensities for short durations have been decreasing since 1990:

***

To recap, many more rain gauges today mean we 'see' more storms - these are typically needed to support basement flooding Municipal Class EA studies (rainfall needed to calibrate hydrologic and hydraulic simulation models), to guide operational activities too.  Many municipalities have installed rain gauges to support inflow and infiltration management programs.

We have a "finer mesh net" to catch these events and add them to our records - we have almost 3 times more rain gauges in Toronto since 2000.

But no. Storm are not becoming more intense. If we see more of them, it is because we are looking harder for them with more extensive monitoring efforts. Given this expanding intensive network of rain gauges today, it is not uncommon, statistically speaking, to observe many 100-year storms over a short time period.  This earlier post explores those statistics in the GTA - https://www.cityfloodmap.com/2019/03/are-six-100-year-storms-across-gta-rare.html.

Extreme Rainfall Trends Toronto and Mississauga - Extending Annual Maximum Series with Environment Canada Data

Environment and Climate Change Canada (ECCC) updates the Engineering Climate Datasets periodically including annual maximum series (AMS's) that reflect annual rainfall extremes over various durations, and also the derived rainfall statistics (intensity-duration-frequency (IDF) curves)) used in engineering design.

Municipalities updating their design standards and practitioners involved in hydrotechnical studies can wait for official updates or complete them in-house. Raw data is available from Environment Canada for a small fee, and can be screened for data gaps / errors and then processed to identify the maximum rainfall each year over the standard periods of 5 minute to 24 hours.

To review local design standards to account for any changes in rainfall intensity, my work team obtained raw data for the Pearson Airport and Toronto City (Bloor Street) climate stations in late 2017 to extend the ECCC analysis. The official Version 2.3 datasets extend to 2007 for Toronto City and 2013 for Pearson Airport - the added raw data extends to cover most of 2017 and 2016 respectively. After screening for anomalies, extended AMS's were analyzed using a Gumbel distribution to generate updated IDF curves.

Selected AMS charts for Pearson Airport and Toronto City stations are shown below:

Toronto City Maximum Annual 24-Hour Rainfall 1940-2017

Toronto City Maximum Annual 1-Hour Rainfall 1940-2017

Pearson Airport (Toronto International) Mississauga Maximum Annual 24-Hour Rainfall 1950-2016

Pearson Airport (Toronto International) Mississauga Maximum Annual 1-Hour Rainfall 1950-2016

Pearson Airport (Toronto International) Mississauga Maximum Annual 5-Minute Rainfall 1950-2016

What do the charts show?

• Long duration rainfall intensities are decreasing (24-hour period)
• Moderate duration intensities are mixed up and down (1-hour period)
• Short duration intensities are decreasing

Note these are not strong trends, and for example the r-squared value for the 1-hour Pearson chart is only 0.002. A previous post shows what happens to design intensities based on these observed rainfall trends, although only using the ECCC datasets and not extended records - see previous post. This is a nice summary considering 21 stations in southern Ontario with over 30 years of record:

We even have some records that go back 100 years like in Kingston, Ontario. Those trends charts show no change in annual extremes since the early 1900's:

Colleagues share that ECCC is updating the analysis for about 100 station records later the year. We'll see if there is any change in AMS trends and significance and derived IDF values compared to the current Version 2.3 Engineering Climate Datasets.

A recent op ed in the Financial Post suggests that analysis of data up to 2012 is not sufficient to assess rain trends as the CBC/Radio-Canada Ombudsman Guy Gendron recently did. Based on the analysis here, adding a few more years to the record is not going to change the overall picture. Its best to focus on other factors affecting flood risk and not the past rainfall trends in regions like southern Ontario. What are some of these other factors?

1) Expanded urbanization as shown in this post showing southern Ontario urban area growth since the mid 1960's and also in this post quantifying urbanization in GTA watersheds,

2) More extensive foundation underpinning, lowering some basements into harms way (closer to sewer back-up levels) as shown in this post on Toronto underpinning permits,

3) System modifications to reduce overflows for environmental protection like in this post referring to infrastructure impacts in Toronto Area 32,

4) Operational decisions that ignore known risks and put people in harm's way like in this post reviewing the July 8, 2013 GO Train flood in the Don River Floodplain,

5) Encroachment on overland flow paths, i.e., lost rivers in urban areas, putting properties at risk of pluvial flooding as this presentation analyzing flooding within overland flow path areas in Toronto in the May 200, August 2005 and July 2013 storms.

Detailed spatial analysis shows that most basement flooding can be explained by 2 factors of i) sanitary sewer inflow and infiltration rates (normalized for catchment area and design return period in a calibrated hydrodynamic model), and ii) the percent full of the sanitary sewers during extreme events - these 2 factors numerically explain over 60% of insurance back-up risks at a postal-code scale of accuracy.

What does this mean? Municipalities need to i) reduce extraneous flows in a cost effective manner in the short term, ii) upgrade sanitary sewer capacity where residual flows are high compared to capacity, iii) upgrade critical storm sewers where design standards are limited and overland flooding stresses adversely affect properties (pluvial flooding at the surface, inflow stresses below the surface), iv) offer private property isolation subsidies (backwater valves and foundation drain disconnection) to provide timely cost-effective risk reduction.

What to do where? It all starts with risk screening, as illustrated in our previous post describing tiered screening for riverine, sanitary and storm systems risks prepared for the Intact Centre on Climate Adaptation for their existing communities 'best practices' document, and another post describing such tiered screening with quantified risk factors prepared for Green Communities Canada's Urban Flooding Collective project.  What about a city-wide perspective on how much to budget for a comprehensive program of flood risk reduction incorporating these tactics? See a recent post that explored the cost-effectiveness of various municipal-wide strategies - look for more details at the 2019 WEAO Annual Conference, and look for even more in future national standards on benefit/cost analysis for flood mitigation we are developing for the National Research Council.

Catastrophic Losses in Canada - Have Flood Damages Increased Significantly Or Have Changing Data Sources Affected Trends?

Disaster Losses Are Up

Catastrophic loss trends have been reported regularly in Canada, often in relation to flood damages. These have often linked to climate change effects as well as other factors that may include aging infrastructure (not a significant factor in our mind), or urbanization and intensification (the true overriding factor in many urban centres). This post looks at how trends have changed in relation to changes in data sources.

GDP Adjusted Losses are Down

A blog post by the Institute for Catastrophic Loss Reduction (ICLR) discusses loss trend reporting by the Insurance Bureau of Canada. ICLR discusses but dismisses the calls for adjusting losses for growth, which is commonplace in Munich RE NatCatSERVICE analysis and reporting, and which is promoted by may others (this includes my paper in the Journal of Water Management Modelling which evaluated losses adjusted for net written premiums, and Roger Pielke Jr.'s work, such as reported here in Five Thirty Eight - see charts to the right - that also calls for evaluating trends considering GDP growth).

The ICLR notes "Normalizing disaster loss data to include such factors as growth in population, economic activity and building stock is not a simple undertaking. Further, there are many problems with using simple measures like GDP or insurance premium growth as a normalizer. For these and other reasons, I don’t want to go ‘there’ at this point ...".

So ICLR is content to us the following chart that does not include GDP adjustments:

Losses in Canada Unadjusted for GDP Growth - 1983-2007 Data per IBC Survey, 2008- Data per CatIQ.

The ICLR notes a change in the data source for the above graph: "Bureau data begins at 1983. From that year to 2007, IBC uses data it collected itself through various company surveys conducted immediately after significant natural disaster events. It also uses various data from Property Claim Services (PCS), Swiss Re, Munich Re and Deloitte. After 2007, the Bureau only uses data from Catastrophe Indices and Quantification Inc. (CatIQ)."

How does the change in data affect reported losses? We can look at how the increase in losses has been reported, for example by the ICLR in 2016:

Catastrophic Loss Trends in Canada. Effects of change in data source on reported losses pre 2008.

Below the ICLR chart, the timing of the change in data is shown. This indicates that the change in reported annual losses from $400M average up to 2008 to $1B average after corresponds to the change in data source in 2008.

More recently the Intact Centre on Climate Adaptation (ICCA) has reported trends in losses on TVO's The Agenda as shown in the chart below:

Intact Centre on Climate Adaptation cites changes in insurable claims on TVO (chart shown), with ICLR's noted change in data sources added below (IBC data up to 2007 and CatIQ data from 2008 onward).

Again, the change in data is added below the ICCA chart. The lower losses of $200-500M up to 2008 and higher losses typically over $1B from 2009 onward correspond to this change in data source.

Adjusting for data sources or for GDP does not really change priorities for flood risk and catastrophic loss reduction. Better characterization of the GDP-adjusted trend can give us insight into the effectiveness of past mitigation efforts, more-resilient design standards that are common in modern practice. Without such GDP adjustment, one would think that everything is built as disaster-prone as it was in the past. Also, understanding the cause of the trend in losses will help focus adaptation or mitigation efforts in the proper place - if increases are explained by GDP growth as opposed to changes in extreme weather (shown to not be a factor) efforts will be placed on adaptation infrastructure built to old, less-resilient design standards as opposed to mitigation (e.g., GHG reduction).

A more wordy comment has been added to the ICLR blog post.

***

A paper Trend Analysis of Normalized Insured Damage from Natural Disasters, published in:
Climatic Change, 113 (2), 2012, pp. 215-237, by Fabian Barthel and Eric Neumayer, Department of Geography and Environment and The Grantham Research Institute on Climate Change and the Environment, London School of Economics and Political Science explores "Normalized" / GDP adjusted damages, exploring trends for different types of events.

As noted in their abstract:

"As the world becomes wealthier over time, inflation-adjusted insured damages from natural disasters go up as well. This article analyzes whether there is still a significant upward trend once insured natural disaster loss has been normalized. By scaling up loss from past disasters, normalization adjusts for the fact that a hazard event of equal strength will typically cause more damage nowadays than in past years because of wealth accumulation over time. A trend analysis of normalized insured damage from natural disasters is not only of interest to the insurance industry, but can potentially be useful for attempts at detecting whether there has been an increase in the frequency and/or intensity of natural hazards, whether caused by natural climate variability or anthropogenic climate change."

The following charts from the paper show an increase in deflated (non-normalized) damage losses over time, and virtually no change in normalized losses.

Global deflated insured losses from natural disasters

Global normalised insured losses from all disasters

Similarly, the following charts illustrate normalized trends for convective storm events (4165 disasters) showing a decrease, all storms including winter and other storms but excluding tropical cyclones (4369 disasters) showing a decrease, and for tropical cyclones (874 disasters) showing an increase.

Global normalized insured losses from convective events

Global normalized insured losses from all storm events except tropical cyclones

Global normalized insured losses from tropical cyclones

***

The Government of Canada has reported that the majority of loss increases have been due to growth (more exposed people, assets and wealth), and that climate change 'may' be having an effect - this contrast many media and insurance industry comments. The true driver of increased losses was reiterated in the just-released Canada in a Changing Climate: National Issues Report (see post: https://www.cityfloodmap.com/2021/06/national-issues-report-identifies.html). Canadian loses have been normalized for growth and show a moderate increase over time - the report notes that earlier data may be incomplete, which would affect the normalized trend as well (more complete older data could decrease the trend).

Short Duration Frequent Rainfall Show No Change in Southern Ontario IDF Design Intensities - No Change in Averages Suggests No Change in Extremes

It is often stated that changes in average conditions are an indicator of changes in extreme conditions. This makes sense for rainfall statistics as a change in typical conditions, such as an increase in rainfall intensities, can be accompanied by higher extreme values as well (i.e., the whole distribution shifts). Since extreme values are somewhat elusive to those recording rainfall intensities at Canadian climate stations - that is, they are rare and may not be readily observed in short records or sparsely-spaced climate stations - we can look at the trends in the more abundant and frequent short duration rainfall statistics as an indicator of where the extreme values are heading.

The following table summarizes trends in short duration rainfall intensities for long term Southern Ontario climate stations (below latitude of 44 degrees). Stations have at least 30 years of record. The change in 2-year 5 minute rainfall intensity and 5-year 10 minute rainfall intensity have been calculated using a starting point of then Environment Canada's 1990 IDF tables (obtained from Environment and Climate Change Canada in 2017), and an ending point of the Version 2.3 Engineering Climate Datasets.

Change in average and frequent rainfall intensities in southern Ontario.

The review indicates that there has been no increase in frequent short duration rainfall intensities. In fact the most frequent 2-year (i.e., average), 5-minute duration rainfall intensities have decreased somewhat. This is welcome news considering the potential for frequent storms to cause erosion in southern Ontario streams. This also suggests that extreme rainfall intensities have not changed as a result of the average rainfall intensities changing. That is, there is no consistent shift in the average rainfall intensities at long term climate stations.

The Insurance Bureau of Canada and the Institute for Catastrophic Loss Reduction have reported that average rainfall intensities have shifted by an entire standard deviation (thus making extreme 40 year storms become more frequent 6 year storms) - this has been refuted by Environment and Climate Change Canada (see Canadian Underwriter editor's note). The data in the above table indicate no such shift.

It is a commonly held belief that rainfall intensities have increased dramatically as a result of climate changes effects. Recently the Globe and Mail reported "It is hard to ignore the growing relationship between climate change and the resulting impact of severe flooding events." .. actually its hard to explain the role of changing climate given rainfall intensity data in some regions. It may be best to ignore rainfall and focus on other flood risk drivers like urbanization and intensification.

Datasets from Environment and Climate Change Canada refute the belief that rainfall is becoming more extreme.

***

The following tables show the 2 to 5 year IDF trends for 5 to 10 minutes (first table), and 5 to 10 year trends for 1 hour and 2 hours (second table)

1990 (pre-version 1) IDF Dataset Worksheets have been prepared for Ontario stations:
Ontario Disk 1 Volume Tables :		https://drive.google.com/open?id=0B9bXiDM6h5ViWV9HeXZIWDZxTXM
Ontario Disk 2 Volume Tables :		https://drive.google.com/open?id=1vhaXcC3MidpgHCmSbXdgqy53pUyAXu0g
Ontario Disk 3 Volume Tables :		https://drive.google.com/open? id=0B9bXiDM6h5ViZVpJMEZzWnNDV28
Ontario Disk 4 Volume Tables :		https://drive.google.com/open?id=0B9bXiDM6h5ViZEVoOE8xT0oyZ2M

Less Extreme Short Duration Rainfall in Kitchener-Waterloo - IDF trends do not show climate change impacts that would affect urban flooding.

The University of Waterloo civil and environmental engineering department analyzed design rainfall intensity trends - that is intensity-duration-frequency (IDF) statistics used in infrastructure design - with the goal of potentially updating City of Kitchener and City of Waterloo design rainfall values. Why do this? Because climate change is predicted in some models to increase the intensity and frequency of extreme rainfall. Results are in the following report:

Update of Intensity-Duration-Frequency (IDF) Curves for the City of Waterloo and the City of Kitchener Prepared By: Donald H. Burn, Ph.D., P.Eng. Department of Civil and Environmental Engineering University of Waterloo, August 2012.

The executive summary states there is no significant change and some decreasing trends for short duration rain intensities - that is, the design parameters that affect urban infrastructure flood risks:

"Annual maximum rainfall data for durations ranging from five minutes to 24 hours were analyzed for trends using the Mann-Kendall non-parametric trend test. Although no statistically significant trends were identified, there were noticeable patterns in the magnitude and direction of the trends in the rainfall data, as a function of the rainfall duration. Based on the available rainfall data for the period 1971 to 2007, new intensity-duration-frequency (IDF) curves were developed for the Waterloo Wellington A climate station; the curves can be used as IDF curves for the City of Waterloo and the City of Kitchener. The Pearson Type III (PE3) distribution was identified as the preferred distribution function for the data and formed the basis for estimating the quantiles required to form the IDF curves. The rainfall intensity values for the new IDF curves tend to be lower than the corresponding values for the existing curves for rainfall durations of up to one to two hours and generally slightly higher than the rainfall intensity values for the existing curves for the longer rainfall durations. The results indicate that the existing IDF curves for the City of Waterloo and the City of Kitchener are likely somewhat conservative for rainfall durations less than two hours, although the impacts of climate change could result in more severe events in the future."

The University of Waterloo findings are consistent with the 'general lack of detectable trend signal' in past rainfall observations as reported by Environment and Climate Change Canada (ECCC). ECCC reported even some regional decreasing trends (St. Lawrence region of southern Quebec and the Atlantic Provinces) for the short duration intensities affecting urban drainage systems:

Trends in Canadian Short‐Duration Extreme Rainfall: Including an Intensity–Duration–Frequency Perspective Mark W. Shephard, Eva Mekis, Robert J. Morris, Yang Feng, Xuebin Zhang, Karen Kilcup & show all Pages 398-417, Published online: 19 Nov 2014 (Atmosphere-Ocean).

What do the University of Waterloo findings look like? These intensity-duration curves for 5-year, 25-year and 100-year rain frequencies show that for duration less than 2 hours, the storm severity has been decreasing:

IDF update to assess climate change impacts for sewer design shows decreasing 5-year rainfall intensities for durations less than 120 minutes (2 hours) - small flashy urban drainage systems now have lower extreme rainfall risk than with previously higher rainfall. Therefore earlier designs using older, higher rainfall design intensities have a safety factors against extreme weather risks for frequent events, and climate change has not adversely affected Kitchener-Waterloo erosion risks or erosion damage potential (erosion risks are often governed by frequent storm stresses). 

IDF update to assess climate change impacts for sewer design shows decreasing 25-year rainfall intensities for durations less than 90 minutes (one and a half hours) - small flashy urban drainage systems subject to nuisance flooding now have lower extreme rainfall risks. Therefore earlier storm drainage designs using older, higher rainfall design intensities, have a safety factors against extreme weather risks, and climate change has not adversely affected Kitchener-Waterloo flood risks or damage potential. Wastewater systems that have extraneous flow stresses (i.e., inflow and infiltration) and that respond to short duration rainfall appear to have lower capacity stresses with climate change IDF update. A Class Environmental Assessment Master Plan study in Kitchener has shown that peak wastewater flows in the trunk systems (e.g., Ottawa, Manchester, Montgomery trunks) are most highly correlated to short duration rainfall intensities. A Class Environmental Assessment study of the Sandrock Greenway trunk has also shown that smaller local wastewater systems respond to short duration rainfall intensities, even without high inflow potential in catchments with fully and partially-separated foundation service catchments).

IDF update to assess climate change impacts for sewer design shows decreasing 100-year rainfall intensities for durations less than 90 minutes (one and a half hours) - small flashy urban drainage systems now have lower extreme rainfall risk today compared to previously higher intensities. Therefore earlier designs using older, higher rainfall design intensities have a safety factors against extreme weather risks for rare storm events, and climate change has not adversely affected flood risks or flood damage potential. 

 Why do derived IDF values decrease? Because the observed maximum rainfall amounts in the Annual Maximum Series (AMS) have been decreasing. The following two graphs show trends in annual maximum rainfall recordings from 1971 to 2007. The 5-minute duration maximum rainfall has been decreasing and the 2-hour maximum rainfall have been flat - it is over this range that IDF intensities have been shown to be decreasing n the earlier graphs.

***
Comparing University of Waterloo analysis with politician statements:

Prime Minister Justin Trudeau recently stated "We are a government grounded in science". If so, why has he stated that 100 year storms may occur every 10 years or sooner following 2017 flooding in Gatineau? There is no data to support the storm frequency statement made by the Prime Minister.

Comparing University of Waterloo analysis with insurance industry statements:

The insurance industry, interested in flood damages and the effect on business activities, has stated that storms that happened every 40 years are now occurring every 6 years. Environment and Climate Change Canada has clearly refuted this frequency shift, most recently in Canadian Underwriter saying its studies do not support the insurance industry statement. As noted in the recent article:

"Associate Editor’s Note: In the 2012 report Telling the Weather Story, commissioned to the Institute for Catastrophic Loss Reduction by the Insurance Bureau of Canada, Professor Gordon McBean writes: “Weather events that used to happen once every 40 years are now happening once every six years in some regions in the country.” A footnote cites “Environment Canada: Intensity-Duration-Frequency Tables and Graphs.” However, a spokesperson for Environment and Climate Change Canada told Canadian Underwriter that ECCC’s studies “have not shown evidence to support” this statement."

The following detailed review of insurance industry statement and comparison with Environment and Climate Change Canada's Engineering Climate Datasets clearly shows how the insurance industry has confused projected rainfall intensity shifts with trends from past observations:

Storm intensity not increasing - factual review of engineering data - Canada and Ontario from Robert Muir

Climate-driven variability in the occurrence of major floods across North America and Europe - Journal of Hydrology Review

An assessment of major flood trends was conducted by a group of researchers including Environment and Climate Change Canada and University of Waterloo Civil and Environmental Engineering. They found no evidence that climate change has to date increased the occurrence of floods.

"Abstract:
Concern over the potential impact of anthropogenic climate change on flooding has led to a proliferation of studies examining past flood trends. Many studies have analysed annual-maximum flow trends but few have quantified changes in major (25–100 year return period) floods, i.e. those that have the greatest societal impacts. Existing major-flood studies used a limited number of very large catchments affected to varying degrees by alterations such as reservoirs and urbanisation. In the current study, trends in major-flood occurrence from 1961 to 2010 and from 1931 to 2010 were assessed using a very large dataset (>1200 gauges) of diverse catchments from North America and Europe; only minimally altered catchments were used, to focus on climate-driven changes rather than changes due to catchment alterations. Trend testing of major floods was based on counting the number of exceedances of a given flood threshold within a group of gauges. Evidence for significant trends varied between groups of gauges that were defined by catchment size, location, climate, flood threshold and period of record, indicating that generalizations about flood trends across large domains or a diversity of catchment types are ungrounded. Overall, the number of significant trends in major-flood occurrence across North America and Europe was approximately the number expected due to chance alone. Changes over time in the occurrence of major floods were dominated by multidecadal variability rather than by long-term trends. There were more than three times as many significant relationships between major-flood occurrence and the Atlantic Multidecadal Oscillation than significant long-term trends."

The paper is available from the Journal of Hydrology. The authors are from across the world.

Glenn A.HodgkinsaPaul H.WhitfieldbDonald H.BurncJamieHannaforddBenjaminRenardeKerstinStahlfAnne K.FleiggHenrikMadsenhLuisMedieroiJohannaKorhonenjConorMurphykDonnaWilsong

a

U.S. Geological Survey, 196 Whitten Road, Augusta, ME 04330, United States

b

Environment and Climate Change Canada, 401 Burrard Street, Vancouver, BC V6C 3S5, Canada

c

University of Waterloo, 200 University Avenue West, Waterloo, ON N2L 3G1, Canada

d

Centre for Ecology and Hydrology, Maclean Building, Benson Lane, Wallingford, Oxfordshire OX10 8BB, United Kingdom

e

Irstea Lyon, Hydrology-Hydraulics, 5 rue de la Doua BP32108, 69616 Villeurbanne cedex, France

f

Albert-Ludwigs-Universität Freiburg, Fahnenbergplatz, 79098 Freiburg, Germany

g

Norwegian Water Resources and Energy Directorate, P.O. Box 5091, Majorstua, 0301 Oslo, Norway

h

DHI, Agern Allé 5, DK-2970 Hørsholm, Denmark

i

Technical University of Madrid, ETSI Caminos, Canales y Puertos, c/ Profesor Aranguren, 3 28040 Madrid, Spain

j

Finnish Environment Institute, SYKE, Freshwater Centre, P.O. Box 140, 00251 Helsinki, Finland

k

Irish Climate Analysis and Research UnitS (ICARUS), Department of Geography, Maynooth University, Maynooth, Co. Kildare, Ireland

Since rainfall intensities are not increasing dramatically across Canada, as demonstrated by Environment and Climate Change Canada (ECCC) in the Atmosphere-Ocean in 2014, saying rainfall intensities are stationary, it makes sense that major floods are not increasing either. Of course there may be local regional trends - ECCC found that there are decreasing rainfall intensities in regions such as the St Lawrence basin of southern Quebec and the Maritimes. Their Engineering Climate Datasets (version 2.3) also show twice as many statistically significant decreasing trends as increasing ones in southern Ontario.

The paper Climate-driven variability in the occurrence of major floods across North America and Europe focused on minimally altered catchments in order to isolate climatic as opposed to hydrologic drivers. It focused on large watersheds. As noted on this blog, urbanization is a key driver of flood risk in small urban catchments and is expected to have increased of the past  50-100 years in many Ontario cities. This is link to GIS mapping of changes: Ontario city urbanization affecting runoff and flood risk since from 1966 to 2000.