Infrastructure Resiliency and Adaptation for Climate Change and Today’s Extremes
1. OWWA/WEAO Joint Climate Change Committee
Climate Change and Infrastructure Design
Hitting a Moving Target
Infrastructure Resiliency and
Adaptation for Climate Change
and Today’s Extremes
Robert J. Muir, M.A.Sc., P.Eng.
Manager, Stormwater
City of Markham
November 16, 2017 - Mississauga, Ontario 1
2. Crystal Ball or Rear View Mirror
What do you really need for extreme
weather adaption?
2
3. If “Weather Zoltar” Predicted Future IDF,
… does it make a ‘BIG’ difference to your
adaptation priorities ?
If I had a
Weather
oltar – short
Story. 3
4. We have always had flooding
Engineers don’t let that stop them in
in their quests …
4
5. Is predicting future weather a priority ?
Or understanding existing risks?
(like repeated operational ones)
http://http://www.cityfloodmap.com/2015/12/toronto-go-train-flood-avoidable-july-8.html
5
7. data shows mostly an old extreme
http://www.cityfloodmap.com/2017/09/toronto-island-flooding-2017-were-lake.html
74
74.2
74.4
74.6
74.8
75
75.2
75.4
75.6
75.8
1915 1925 1935 1945 1955 1965 1975 1985 1995 2005 2015
MonthlyLevel(m)
Year
Lake Ontario Historical May - August Levels
May Average
August Average
Source:
1918-2016 http://www.tides.gc.ca/C&A/network_means-eng.html
2017 http://tides-marees.gc.ca/C&A/pdf/Bulletin1708.pdf
May 5 cm above record
August not a record
7
8. Do roadways and basements flood due to
climate change factors ?
Or old intrinsic design limitations?
2017 1962
8
9. Climate Change or ‘Hydrology Change’?
http://www.cityfloodmap.com/2016/03/lost-rivers-newtonbrook-tributary.html 9
10. Climate Change or ‘Hydrology Change’?
http://www.cityfloodmap.com/2016/03/lost-rivers-newtonbrook-tributary.html 10
14. Expanded Urban Areas (Pink)
… +1000 % in Rouge Watershed since 60’s
http://www.cityfloodmap.com/2016/08/land-use-change-drives-urban-flood-risk.html
14
15. Expanded Urban Areas (Pink)
… +1000 % in Rouge Watershed since 60’s
http://www.cityfloodmap.com/2016/08/land-use-change-drives-urban-flood-risk.html
15
16. Expanded Urban Areas (Pink)
… +1000 % in Rouge Watershed since 60’s
http://www.cityfloodmap.com/2016/08/land-use-change-drives-urban-flood-risk.html
16
17. Expanded Urban Areas (Pink)
… +1000 % in Rouge Watershed since 60’s
http://www.cityfloodmap.com/2016/08/land-use-change-drives-urban-flood-risk.html
17
18. Expanded Urban Areas (Pink)
… +1000 % in Rouge Watershed since 60’s
http://www.cityfloodmap.com/2016/08/land-use-change-drives-urban-flood-risk.html
18
19. Urban ‘Pink Splat’… like Cat in the Hat
… don’t think we’ll get
old hydrology back. 19
20. Given known design limitations
and significant, quantifiable
hydrology stresses affecting
urban flooding ….
20
21. .. should we “Blame it on the Rain”
like Milli Vanilli did ? 21
22. or critically assess local rain data trends?
http://www.cityfloodmap.com/2015/12/trends-in-canadian-shortduration.html
22
23. … but “What about 2005 & 2013…” ?
… the storms were extreme !”
Finch 2005
Ferarri 2013
23
24. 2013 was a record for all “July 8’s”
.. but we do not design infrastructure
for particular calendar days.
20
40
60
80
100
120
140
1945 1955 1965 1975 1985 1995 2005 2015 2025
MaximumRainfallDepth(mm)
Pearson Airport Maximum Daily Rainfall 1950 - 2016
http://www.cityfloodmap.com/2016/01/toronto-climate-change-extreme-rainfall.html
http://www.cityfloodmap.com/2015/11/thinking-fast-and-slow-about-extreme.html
24
25. … but “IBC says 40 year storms are
happening every 6 years !!!!!!!!!”
… is an “Insurance Fact”! 25
26. Telling the Weather Story - Gordon McBean - Empire Club presentation - YouTube
http://www.slideshare.net/RobertMuir3/storm-intensity-not-increasing-factual-review-of-engineering-datasets
Actually, not a fact …
.. confused prediction with past data.
26
27. Test : Infographics or Data
… can you tell the difference?
Environmental Commissioner
Engineering Climate Datasets v2.3
27
29. Toronto IDF Intensities Decreasing
• As annual
maximum values
drop, extreme
IDF intensities
decrease too.
• Toronto City
“Bloor Street”
trends are lower
for all durations
and for all return
periods.
http://www.cityfloodmap.com/2016/01/toronto-climate-change-extreme-rainfall.html
www.cityfloodmap.com
Source:
Environment Canada Engineering Climate Dataset
ftp://ftp.tor.ec.gc.ca/Pub/Engineering_Climate_Dataset/IDF/
Up to 2007 per Dataset v2.3, to 2003 per Dataset v1, to 1990 per hardcopy records
CityFloodMap.Com, 2016
29
30. Mississauga IDF Intensities Decreasing
• Mississauga
Airport 5-15
min. intensities
decreasing for
all return
periods.
• 24 hour trends
are downward
despite July 8,
2013 storm.
30
32. Canada-wide “lack of a detectable trend
signal”, some regional decreases.
http://www.tandfonline.com/doi/abs/10.1080/07055900.2014.969677?journalCode=tato20 32
33. “Rear View Mirror”
data says:
Old standards were
sometimes a ‘wreck’
Hydrology changes
have ‘monster’ impact
Rain intensity change
was a ‘yappy puppy’
33
34. Has rainfall increased with temperatures?
.. no, research at MIT/Columbia and U.
of Western show some opposite trends.
http://www.ldeo.columbia.edu/~clepore/publications_pdf/grl52319.pdf
https://www.slideshare.net/glennmcgillivray/iclr-friday-forum-updating-idf-curves-for-future-climate-march-24-2017
Decreases
above 5 degrees C.
Clausius-Clapeyron
theory has not
been observed.
‘Flat’ trends
compared
to theory.
http://www.cityfloodmap.com/2017/06/does-higher-temperature-increase-rain.html 34
35. What and where do we need to adapt
for tomorrow’s expected extremes?
… same as for “design standard
adaptation” for yesterday’s & today’s
extremes.
35
37. Cities must focus on areas with overland flow limitations
too (these drive many flood reports and claims).
<- Not in these areas ….
… but
in these
overland
flow
issue areas
beyond
valleys.
37
38. Why focus beyond valleys ?
… because flood records show this is the priority.
https://www.slideshare.net/RobertMuir3/urban-flood-risk-from-flood-plains-to-floor-drains 38
39. August 19, 2005 Storm - Percentage of
Properties Flooded
Markham Adaptation Priorities are Pre 1980’s Areas
Pre 1980
1.0 % Flooded
Post 1980
0.15 % Flooded
* Considers all properties city-wide (storm intensity
was high across the majority of the City).
Standard Adaptation:
downspout
disconnection, MH
sealing, storm capacity
upgrades (limit sanitary
inflows), backwater
valves, etc.
39
40. July 16, 2017 Storm - Percentage of
Properties Flooded
July 2017 Storm Confirmed Design Standard
Adaptation Priorities
Pre 1980
2.4 % Flooded
1980 - 1990
0.6% Flooded
Post 1990
0.04% Flooded
• Considers north-east quadrant of the City where the
storm intensities were highest (25,527 properties). Post-1990 servicing flooded 7 of 15,889;
1980-1900 servicing flooded 21 of 3366; pre-1980 servicing flooded 151 of 6272.
Most effective to focus
on these old areas –
new ones are resilient
to 2095. See CWWA
2017 future stress test
analyses.
40
41. Future IDF ‘Stress Test’ – City-wide Sanitary System
• Future U. of R. A1B 50%
2065-2095 scenario.
• Chicago Storm -
increase IDF values by
30% over 2 hrs.
• Use calibrated
/monitored values.
• Apply calibrated inflow &
infiltration response.
• Dynamic / gradually
varied flow model
(InfoWorks).
• Sanitary system HGL
less than 2.0 m below
grade (near basement).
IDF
Data
Hyetograph
Selection
Dry Weather
Sanitary Flow
I&I Transformation
(Extraneous Flow)
Hydraulic Performance
/ Freeboard
Damages / Risk – Assess
Need for Adaptation
Today’s
IDF
Future
IDF
Sewer
Surcharge
(%)
7.4 12.1
+ 64 %
MH
Surcharge
(%)
1.8 3.5
+ 94%
Impacts in old
subdivisions,
new subdivisions
resilient
Most
Critical
± 2000 %
Risk
Less
Critical
41
42. Fully-Separated Sanitary (post 1980’s)
Not Stressed with Future IDF (High Resiliency)
Shallow Sewer (limited freeboard for dry weather)
Existing 100-Year IDF - MH Freeboard < 2 m
Future 100-Year IDF – MH Freeboard <2 mToday’s IDF
Future IDF
Freeboard < 2 m = Back-up Risk
Impacts in old
subdivisions
with existing
limitations
Pre 1980
80’s
90’s
42
43. 0
50
100
150
200
250
300
0 50 100 150 200 250 300 350 400 450 500
RainfallIntensity(mm/hr)
Time (minutes)
Markham Storm (3 Hr AES) 100 Year
TRCA Storm (AES 12 Hr) 100 Year
100-Year Storm Hyetographs Are Highly Variable / Uncertain
Risk
“Gap”
Sanitary sewer
surcharge and flood risk
vary by 2000% based on
changes in storm
pattern.
43
48. 0
50
100
150
200
250
300
350
400
1975 1995 2015 2035 2055 2075 2095
Post 80’s systems are
resilient for 20 – 30 %
higher IDF (storm &
sanitary networks)
So focus on known problems & priorities to adapt
Observed Predicted
Today’s Standard
RainIntensitymm/hr(5minute100-year)
“Design Standard
Adaptation”
in old areas. Requires
250% or more
conveyance capacity.Pre 80’s Standard
48
49. Conclusions
• There are significant industry gaps in understanding:
• Past extreme rainfall trends (decreasing in southern Ontario)
• Key quantifiable drivers for flood risk reduction
weather vs. hydrology change
weather vs. hydraulics constraints (overland design)
• System vulnerability varies with design standards:
• Post 1990 standards significantly decreased flood risk to 1/60th
of pre-1980’s standards risks.
• Historical land use practices with limited design standards
drive specific, local adaptation priorities (no generic fixes).
• “Design Standard Adaptation” adds resiliency
• Addressing older system limitations provides future climate
adaptation co-benefit … regardless of future rainfall IDF
(uncertain & not converging).
• Highest incremental benefits for infrastructure improvements
to older systems – marginal benefits decrease at higher levels.49
50. Thank You
Questions ?
More Rob :
Blog: www.CityFloodMap.com
Podcast: Open During Construction on iTunes
Twitter: @RobertMuir_PEng
More City of Markham :
Web: www.markham.ca
Twitter: @CityofMarkham
50
Notas del editor
Thank you for sticking around for the last presentation – good things come to those who wait.
Hopefully a fresh persepctive.
Crystal ball or rear view mirror.
What do you really need for extreme weather adaptation?
Is knowing future extreme weather patterns essential to guide where and how much to adapt – what we do with infrastructure - or can we look back to see the risks factors to address?
Make a wish.
If you had a Weather Zoltar arcade machine to predict future intensity duration IDF statistics.
Put in a quarter, and our comes future IDF.
If you had such as machine, does its output make a Big difference in how your city should adaptation?
Some cities have put a quarter – of a million dollars - into similar machines.
If we look to the past, we see we have always had extreme events and flooding.
Engineers can’t predict the details but that doesn’t stop them in their everyday quests.
Like getting to the beer store. Or upgrading infrastructure.
Is predicting future weather even a priority?
Lets say to address catastrophes like the flooded GO Train in 2013.
Or could we manage some existing risks by critically understanding operational risks with today’s weather.
Sometimes history is just repeating itself and weather is not the key. The GO Train has been flooding right after the Richmond Hill line started in the late 1970’s.
Toronto Island Flooding, is that a new normal?
Or and old extreme?
To answer this, we really have to look at data instead of just exercising our availability biases.
Anyone remember the flood of 1973?
Or the flood of 52 or 47 ?
Probably not – I see a crowd of young good looking people.
For you folks, data shows 2017 – while above average - was just centimeters above previous extremes.
Make you wonder why evictions from this known hazard zone were stalled in the 80’s.
Do roadways and basements flood due to climate change factors ?
That is Windsor in
Are effective CSO controls keeping too much wet weather flow in the pipes for extreme events, is the 100 year floodplain level of service enough Essex County?
What have we had more of recently – climate change or hydrology change.
This shows the Newtonbrook area in North York
It developed through the 40s 50s and 60s and upstream in Vaughan through the 80’s – so parts of our cities have not really been subject to full hydrology runoff stresses until recently – this area was flooding by the 80’s.
What have we had more of recently – climate change or hydrology change.
This shows the Newtonbrook area in North York
It developed through the 40s 50s and 60s and upstream in Vaughan through the 80’s – so parts of our cities have not really been subject to full hydrology runoff stresses until recently – this area was flooding by the 80’s.
And some systems are designed to flood occasionally because of lack of overland flow path.
We only started that in the late 1970’s.
So we have inherited systems with ‘Lost Rivers’ and overland flow risks. This maps shows the overland flow spread areas, and reported flooding. One keen insurance company found 3 time the claims in 2013 in overland flow zones.
And some systems are designed to flood occasionally because of lack of overland flow path.
We only started that in the late 1970’s.
So we have inherited systems with ‘Lost Rivers’ and overland flow risks. This maps shows the overland flow spread areas, and reported flooding. One keen insurance company found 3 time the claims in 2013 in overland flow zones.
And we have more than doubled the hard surfaces in our subdivisions.
That drives higher flood risks in our watersheds.
Intensification has doubled impervious due to infills in one Markham neighbourhood since the 50’s – Bayview Glen.
And urbanization change has occurred at a considerable, measurable pace with almost 1000% more urban are in the Rouge Watershed from the mid 60’s to the 1999 – other areas that have experience significant flooding in the GTA has experienced the same.
And urbanization change has occurred at a considerable, measurable pace with almost 1000% more urban are in the Rouge Watershed from the mid 60’s to the 1999 – other areas that have experience significant flooding in the GTA has experienced the same.
And urbanization change has occurred at a considerable, measurable pace with almost 1000% more urban are in the Rouge Watershed from the mid 60’s to the 1999 – other areas that have experience significant flooding in the GTA has experienced the same.
And urbanization change has occurred at a considerable, measurable pace with almost 1000% more urban are in the Rouge Watershed from the mid 60’s to the 1999 – other areas that have experience significant flooding in the GTA has experienced the same.
And urbanization change has occurred at a considerable, measurable pace with almost 1000% more urban are in the Rouge Watershed from the mid 60’s to the 1999 – other areas that have experience significant flooding in the GTA has experienced the same.
So this pink urban splat .. looks like Cat in the Hat.
And I don’t think we’ll get old hydrology back.
Given known design limitations and hydrology stresses
Given known design limitations and hydrology stresses ….
Or critically assess local data for trends.
These annual maximum rain trends in Toronto are from Environment Canada’s Engineering Climate Dataset.
Toronto long term trends are down.
But Rob! What about 2005 and 2013 .. those storms were extreme.
We lost Finch. I lost my Ferrari !!!
2013 had a record for July 8ths definitely – but we do not design infrastructurefor particular calendar days.
Check out the data on my blog at the first link.
And at the second link I breakdown how the media takes short cuts when reporting on extremes.
But Rob! The insurance Bureau of Canada says 40 year storms are happening every 6 years. Telling the weather story says this is happening in some part of Canada.
Actually not a fact at all – this 40 to 6 year frequency shift was based on an arbitrary 1 standard deviation shift in a standard normal bell curve as presented at the Empire Club of Canada by the author.
I am bound by confidentiality on 3 files settled through advertising standards Canada using this inaccutate fact in advertising – so don’t ask me any questions on the advertising. But lets look at the data I shared.
An I need to test you to see if you can tell the difference between infographics and data.
Which one of these is data.
OK lets look at that good stuff for southern Ontario.
Southern Ontario extreme rain trends are decreasing.
What would make me say such as silly thing.
Answer- The Engineering Climate Datasets.
More significant decreases than increases and more statistically significant decreases.
Follow the links. Get the data of check out the review on my blog.
As observed rainfall decreases so do extreme value statistics – IDF curve intensities. This is the Toronto Bloor Street trend – look at the bottom row, the 100 year rain intensity, its down over 5% since I graduated.
Mississauga IDF intensities down too even considering 2013.
This is repeated in countless studies from the University of Waterloo, University of Guelph, professional studies for cities.
And Canada-wide – in the rear view
Has rainfall increased with temperatures?No research has shown sometimes the opposite – so when someone says 1 decrees Celsius equals 7% more water vapour so we must get more rain, they are referring to the Clausius-Claperon equation – and obervations show this has not happened.
So fundamentally it makes sense that we may not see rain changes like temperatures changes.
What and where do we need to adapt for tomorrow’s expected extremes?… same as for “design standard adaptation” for yesterday’s and today’s extremes.
I’ll show you the Markham example.
Based on spatial analysis of flood reports from 2000, 2005 and 2013 in Toronto.
This adaptation is needed most in partially separated sewer service areas with no CSO relief.
So in Markham in 2005 most flooding was in old areas – pre 1980’s areas with limited design standards - so we are doing the most cost effective adaptation measures first. Downspout disconnection. And storm capacity upgrades. We are also starting a backwater valve program and looking at sanitary capacity upgrades because we cannot control all the I&I.
And this summers storms confirmed our priorities. New subdivisions post 1990’s are resilient already.
I also presented the future resiliency of storm and sanitary sewer systems at CWWA last week.
If we use a different hyetograph, like Toronto’s May 200 storm or Ottawa’s 2004 100 year storm we would have about half the surcharge percentage.
This shows the Markham 100 year design storm and the conservation authority 100 year floodline mapping storm – there is a big difference, and one is more conservative for flashy urban systems.
What and where do we need to adapt for tomorrow’s expected extremes?… same as for “design standard adaptation” for yesterday’s and today’s extremes.
I’ll show you the Markham example.
Although local trends are downs we do look at future climate IDF to stress test sewer systems.
Its easy pick the starting point, pick the future climate scenario and stress test any system that is designed below the future standard. We are doing this in Markham.
Although local trends are downs we do look at future climate IDF to stress test sewer systems.
Its easy pick the starting point, pick the future climate scenario and stress test any system that is designed below the future standard. We are doing this in Markham.
But it is not that easy – there is no simple future scenario and they all diverge up and down even.
But it doesn’t matter – back to the question of adaptation. And lets scale this chart to show the old standards.
This shows todays standard and a range of possibilities.
We have tested that new systems can handle a 20-30 % increase in IDF if we use inlet controls and modern standards.
But some old systems were designed down here. @ year sewers and no overland flow path – it is a no brainer that we will achieve significant risk reduction by increasing the standards in those areas even to todays rain intensities – and the greatest marginal return on investment will be on in the most deficient areas. There will also be diminishing relative marginal returns above that.
Most cities have decades of this basic design capacity adaptation to do to bring up old system to todays standards.