CityFloodMap.Com: Trends in Canadian Short‐Duration Extreme Rainfall Data Contradict Insurance Bureau Statements

Trends in Canadian Short‐Duration Extreme Rainfall Data Contradict Insurance Bureau Statements

Rain intensity data in Canada show "lack of a detectable trend signal", despite recent statements by the Insurance Bureau of Canada ("Extreme weather events driven by climate change have increased in frequency and severity," said Don Forgeron, President and CEO, IBC, at a November 26, 2015 Economic Club of Canada event in Edmonton).

In this post we explain the rainfall intensity trend information available in Version 2.3 of Environment Canada's Engineering Climate Datasets, released in December 2014. That data is available here:

ftp://ftp.tor.ec.gc.ca/Pub/Engineering_Climate_Dataset/IDF/

(September 2, 2019 - the above ftp site is no longer - Version 3.00 IDF Files may be accessed here: http://climate.weather.gc.ca/prods_servs/engineering_e.html; the Version 2.30 trend file "idf_v2-3_2014_12_21_trends.txt", is available here: https://drive.google.com/open?id=0B9bXiDM6h5ViQ0xnSDR2cUl3WXc)

The "What's New" document describe "A new set of graphs that represents the historical trends of IDF properties has been produced for each station". These trend graphs have been featured in many posts on the www.cityfloodmap.com blog. Could it be scientists in Canada are muzzled from getting these facts out as that do not support climate change mitigation policies (cap and trade, Bill 172)? If storms are no worse, why look at emissions to address flooding?

The document Notes_on_EC_IDF.pdf available in Environment Canada's doc.zip package, entitled "Documentation on Environment Canada Rainfall Intensity-Duration-Frequency (IDF) Tables and Graphs Version V2.30 December, 2014", describes this new trend data in the appendix as follows:

Appendix: Trend Graphs

Trend graphs are included in release V2.30. The single station rainfall annual maximum series (AMS) were examined for any detectable trend, at a significant level 5%, for each of the durations examined. Figure A-3 is an example of such trend plots. The open circles in each plot represent the AMS rainfall amounts for the duration analyzed. These data are usually very scattered representing the variability of the climate. For most stations, the AMS does not feature any significant trends (Trend: N) but in some instances, an increasing (Trend: +) or decreasing (Trend: -) with time are noticeable. The slope with confidence levels are given in each duration plot.

The method and implications for trend analyses of IDF stations across Canada were reported in the paper: Mark W. Shephard, Eva Mekis, Robert J. Morris, Yang Feng, Xuebin Zhang, Karen Kilcup & Rick Fleetwood (2014): Trends in Canadian Short-Duration Extreme Rainfall: Including an Intensity-Duration-Frequency Perspective, Atmosphere-Ocean, DOI: 10.1080/07055900.2014.969677

So let's look at some trend indicators "+","-" and"N" for actual gauges. The Toronto City gauge has decreasing rainfall intensity trends for all durations from 5 minutes to 24 hours. The following mark-up shows that the 5 minute to 2 hour duration trends are flagged as "Trend :N", meaning not statistically significant. For longer durations, 6, 12 and 24 hours durations on the bottom three charts, the trends are flagged as "Trend :-", where the "-" means statistically significant. So some Toronto rainfall trends are decreasing to a large degree that is beyond the intrinsic variability expected in observed rainfall.

Toronto weather station 6158355 has decreasing extreme rainfall trends for all durations since 1940.
Over longer storm event durations, the decreasing storm intensity is statistically significant.

The noted Atmosphere-Ocean paper summarizes the Canadian extreme rainfall trend analysis in the abstract. Highlighted text below shows that there is no statistically significant trend overall in Canada, with the exception of parts of Newfoundland for durations of 1 - 2 hours:

Abstract
Short-duration (5 minutes to 24 hours) rainfall extremes are important for a number of purposes, including engineering infrastructure design, because they represent the different meteorological scales of extreme rainfall events. Both single location and regional analyses of the changes in short-duration extreme rainfall amounts across Canada, as observed by tipping bucket rain gauges from 1965 to 2005, are presented. The single station analysis shows a general lack of a detectable trend signal, at the 5% significance level, because of the large variability and the relatively short period of record of the extreme short-duration rainfall amounts. The single station 30-minute to 24-hour durations show that, on average, 4% of the total number of stations have statistically significant increasing amounts of rainfall, whereas 1.6% of the cases have significantly decreasing amounts.
However, regional spatial patterns are apparent in the single station trend results. Thus, for the same durations regional trends are presented by grouping the single station trend statistics across Canada. This regional trend analysis shows that at least two-thirds of the regions across Canada have increasing trends in extreme rainfall amounts, with up to 33% being significant (depending on location and duration). Both the southwest and the east (Newfoundland) coastal regions generally show significant increasing regional trends for 1- and 2-hour extreme rainfall durations. For the shortest durations of 5–15 minutes, the general overall regional trends in the extreme amounts are more variable, with increasing and decreasing trends occurring with similar frequency; however, there is no evidence of statistically significant decreasing regional trends in extreme rainfall amounts. The decreasing regional trends for the 5- to 15-minute duration amounts tend to be located in the St. Lawrence region of southern Quebec and in the Atlantic provinces. Additional analysis using criteria specified for traditional water management practice (e.g., Intensity-Duration-Frequency (IDF)) shows that fewer than 5.6% and 3.4% of the stations have significant increasing and decreasing trends, respectively, in extreme annual maximum single location observation amounts. This indicates that at most locations across Canada the traditional single station IDF assumption that historical extreme rainfall observations are stationary (in terms of the mean) over the period of record for an individual station is not violated. However, the trend information is still useful complementary information that can be considered for water management purposes, especially in terms of regional analysis.

The following chart by CityFloodMap.Com aggregates all the trend indicators in the version 2.3 dataset, counting combinations of trend slope and statistical significance as follows:

"Trend :-", slope negative : Significant Decrease
"Trend :N", slope negative : Decrease
"-99.9" : No Data
"Trend :N", slope positive : Increase
"Trend :+", slope positive : Significant Increase

Across Canada, only a few percentage of weather stations have significant increases in recorded extreme rainfall, for any given duration of storm. IDF curves based on these overall stable trends would not change over time.

The Toronto City ("Bloor Street gauge") rainfall trends would fall into the light green "Decrease" range for 5 minutes, 10 minutes, 15 minutes, 30 minutes, 1 hour and 2 hours, and fall into the dark green "Significant Decrease" range for the 6 hour, 12 hour and 24 hour durations.

It is interesting to review the Pearson Airport gauge data trends given the record setting rainfall on July 8, 2013 that caused widespread urban flooding in Toronto. Even with the significant 2013 rainfall event, the 5 minute, 6 hour, 12 hour and 24 hour trends are negative, but not statistically significant, as indicated by the "Trend :N" flag. The 10 minute to 2 hour duration trends are positive but not statistically significant.

The confidence limits show that while there may be upward trends in the 15 minute extreme rainfall amounts - the slope is +0.02 mm/year - there is also a chance that the trend is downward by a significant amount (lower confidence band is -0.12 mm/yr). This help illustrate why the positive trend is not statistically significant, as rainfall patterns are highly variable and minor trends can occur due to the random nature of observations.

Many factors affect flood risks and damages, and those related to runoff have increased over decades in many Canadian cities. For example the Don River Watershed has increased in the amount of impervious, high-runoff cover from 15% in the 1950's to nearly 90% today and many overland flow paths, critical for conveying overland flow during extreme events, have been blocked or constrained.

In general, the Insurance Bureau of Canada has not relied upon data to make weather statements and has instead relied on theoretical statements, confusing projections with observations, and substituting temperature probability density functions with severe rainfall distributions:

http://www.slideshare.net/RobertMuir3/storm-intensity-not-increasing-factual-review-of-engineering-datasets

"Without knowledge action is useless and knowledge without action is futile."

Abu Bakr

Let's hope that IBC can improve knowledge of Environment Canada's Engineering Climate Datasets including the new trend analysis available - this is needed to better inform inaccurate statements on extreme rainfall trends and to focus instead on important flood risk factors. Cap and trade climate mitigation won't help reduce flooding because storms are not getting worse - so the government should focus on infrastructure upgrades in areas built before the 1980's (most cities), i.e., areas with lower levels of service in drainage design.

***

Here is a list of Ontario weather stations where the measured annual maximum 15 minute storm duration intensities are decreasing:

Station ID Name
6020LPQ ATIKOKAN (AUT)
6037775 SIOUX LOOKOUT A
6041221 CARIBOU ISLAND
6046770 PUKASKWA NATL PARK
6056907 RAYNER
6059408 WAWA (AUT)
6068158 SUDBURY SCIENCE NORTH
6073980 KAPUSKASING CDA ON
6079068 UPPER NOTCH
6084307 LAKE TRAVERSE
6085700 NORTH BAY A
6100971 BROCKVILLE PCC
6101901 CORNWALL ONT HYDRO
6104027 KEMPTVILLE CS
6104146 KINGSTON A
6104175 KINGSTON PUMPING STATION
6105978 OTTAWA CDA RCS
6106000 OTTAWA MACDONALD-CARTIER INT'L A (significant trend)
6107836 SMITHS FALLS TS
6110557 BARRIE WPCC
6111792 COLLINGWOOD
6112072 DORSET MOE
6116132 OWEN SOUND MOE
6116843 RAGGED RAPIDS
611E001 EGBERT CS
6127514 SARNIA AIRPORT
6133362 HARROW CDA AUTO
6135638 NIAGARA FALLS
6136606 PORT COLBORNE
6137287 ST CATHARINES A
6137362 ST THOMAS WPCP
6137730 SIMCOE
6139525 WINDSOR A
6140818 BLUE SPRINGS CREEK
6140954 BRANTFORD MOE
6149625 WOODSTOCK
6142286 ELORA RCS
6151042 BURKETON MCLAUGHLIN
6153194 HAMILTON A
6154820 MAIN DUCK ISLAND
6155722 OAK RIDGES
6155790 ORANGEVILLE MOE
6158355 TORONTO CITY
6158406 TORONTO BOOTH
6158665 TORONTO ISLAND A (significant trend .. so is 5 minute and 10 minute storm duration)

6158732 TORONTO LESLIE EGLINTON
61587PG TORONTO SENECA HILL
6158875 TRENTON A
6166418 PETERBOROUGH A
6166450 PETERBOROUGH STP
6169453 WEST GUILFORD

Extreme rainfall trends in Canada (Environment Canada Engineering Climate Datasets) are documented in the following posts:

Static Maps: http://www.cityfloodmap.com/2015/12/severe-storm-trends-canada-rainfall.html

Interactive Map: http://www.cityfloodmap.com/2015/12/canadian-extreme-rainfall-map-climate.html

Table Summaries: http://www.cityfloodmap.com/2015/12/canadian-extreme-rainfall-summary-by.html

Chart and Table: http://www.cityfloodmap.com/2015/12/top-weather-story-in-canada-2015-less.html

Long-term Station Table: http://www.cityfloodmap.com/2015/12/long-term-climate-change-short-term.html

Environment Canada Denies Changes: http://www.cityfloodmap.com/2015/10/bogus-statements-on-storms-in-cbcnewsca.html

Contradicting Insurance Industry Claims: http://www.cityfloodmap.com/2015/12/trends-in-canadian-shortduration.html