4.3.2.2: Projected changes and
uncertainties
In the
future, extreme precipitation is projected to increase in Canada. Averaged for
Canada, extreme precipitation with a return period20 of 20 years in the late
century climate is projected to become a once in about 10-year event in
2031–2050 under a high emission scenario (RCP8.5) (see Figure 4.20). Beyond mid-century,
these changes are projected to stabilize under the low emission scenario
(RCP2.6), but to continue under the high emission scenarios (RCP8.5). An
extreme event that currently occurs once in 20 years is projected to become
about a once in five-year event by late century under the high emission
scenario (RCP8.5). In other words, extreme precipitation of a given magnitude
is projected to become more frequent. Moreover, the relative change in event
frequency is larger for more extreme and rarer events. For example, an event
that currently occurs once in 50 years is projected to occur once in 10 years
by late 21st century under a high emission scenario (RCP8.5). The amount of
precipitation with a certain recurrence interval is projected to increase. The
amount of 24-hour extreme precipitation that occurs once in 20 years on average
is projected to increase by about 5% under a low emission scenario (RCP2.6) and
by 12% under a high emission scenario (RCP8.5) by 2031–2050, and to increase as
much as 25% by 2081–2100 under a high emission scenario (RCP8.5). Differences
in the projected percentage changes in annual maximum 24-hour precipitation
among regions of Canada for the same emission scenario and time period are
notably small. The median value for every region is in general within the range
of 25th to 75th percentiles of other regions, except under the high emission
scenario toward the late century (see Table 4.6).
While
results from global climate model projections (such as those above) are useful
for impact assessment and adaptation planning, there is an important caveat,
especially for extreme precipitation. It is difficult to interpret these
projections at local scales. The spatial resolution of global climate models is
coarse (typically 100–250 km). The precipitation extremes in a model
therefore represent averages over an area of several thousand square
kilometres, and so convey different information than may be required for
practical applications. More importantly, climate models may not include all of
the physical processes that produce local intense rainstorms. This affects the
confidence we can have in statistical downscaling products that transform
precipitation from coarse resolution models to smaller areas. While regional
climate models may operate at much smaller scales, conventional regional
climate models, which are used to conduct most of the dynamical downscaling, do
not accurately simulate important processes such as convection. These
limitations must be kept in mind when using projections for the purpose of
regional and local adaptation; in particular, the projected values given by
these global or regional climate models should not be interpreted literally as
the measured amount of precipitation at a point location.
Estimating
changes in short-duration extreme precipitation at a point location is complex
because of the lack of observations in many places and the discontinuous nature
of precipitation at small scales. Projection of such extreme precipitation is
also difficult because of the shortage of simulations by models with a very
high resolution that resolve the physical processes that produce those extreme
events (Zhang et al., 2017). Nevertheless, multiple lines of evidence support
high confidence in projecting an increase in extreme precipitation globally.
These lines of evidence include attribution of an observed increase in
high-latitude total precipitation to human influence, consistency in projected
future increases in extreme precipitation by multiple models, and the physical
understanding that warming would result in an increase in atmospheric moisture.
It is likely that extreme precipitation will increase in Canada in the future,
although the magnitude of the increase is much more uncertain.
4.4: Attribution Of Extreme Events
Anthropogenic
climate change has increased the likelihood of some types of extreme events,
such as the 2016 Fort McMurray wildfire (medium confidence) and the extreme
precipitation that produced the 2013 southern Alberta flood (low confidence).
There has been
an increase in costly extreme weather and climate events worldwide (WMO, 2014)
and across Canada (Kovacs and Thistlethwaite, 2014; OAGC, 2016; OPBO, 2016).
Much of this rise is due to greater exposure to the effects of such extreme
events, as Canada’s population and the value of its supporting infrastructure
have both increased considerably. Changes in the intensity and frequency of
damaging extreme weather and climate events due to climate change (IPCC, 2013)
may also be playing a role. These extreme weather/ climate events attract
attention because they are rare and often have notable impacts on our society
and economy.
It is
generally not feasible to answer the question, Did human-induced climate change
cause a particular weather or climate event? Often, that event could have
occurred in the absence of human effects. Instead, recent research has focused
on whether human activity has influenced the probability of particular weather
or climate events or, in some cases, the strength or intensity of the events.
As the climate changes, largely due to anthropogenic influences, the likelihood
of a particular class of events — all events as extreme as or more extreme than
the one defined in the study — also changes (NASEM, 2016). In this sense, an
extreme event may be attributable to causes external to the natural climate
system. Thus, a new branch of climate science, called event attribution, has
emerged that evaluates how the probability or intensity of an extreme event, or
more generally, a class of extreme events, has changed as a result of increases
in atmospheric GHGs from human activity.
A growing
number of extreme events in Canada and worldwide are being examined in this way
(e.g., Herring et al., 2017; NASEM 2016). Several of these event attribution
analyses are relevant to Canadians (see Table 4.7). Two examples are
highlighted in this section, including a description of methods of analysis in
Box 4.3.
4.4.1: Attribution of two recent
events
4.4.1.1: 2013 Southern Alberta flood
In June
2013, an extreme flood event in southern Alberta became Canada’s costliest
natural disaster to that date, with significant damage to property and
infrastructure throughout the region, including in the City of Calgary. The
flood displaced almost 100,000 people and resulted in $6 billion in damage,
including $2 billion in insured losses (ECCC, 2017).
A storm
producing heavy rainfall over the region triggered the flooding event in the
Bow River basin, but a combination of both meteorological and hydrological
factors led to the extreme flooding. A recent study (Teufel et al., 2017)
assessed the contributions of several of these factors, including anthropogenic
GHG emissions.
The study
used the Canadian Regional Climate Model (CRCM5) to run large ensembles of
high-resolution simulations for North America. To assess the contribution of
human climate change, simulations of the model were run using present-day
levels of GHGs and also using pre-industrial levels to represent the time
before humans had a discernable impact on the climate.
To estimate
the probability of the event, return periods were calculated for three-day
rainfall totals during May and June exceeding the observed amount. The return
period for the observed event in the present-day climate in the Bow River was
estimated to be about 60 years. Using climate projections, the return period is
estimated to be reduced to about 20 years by the late 21st century (under both
an intermediate emission scenario [RCP4.5] and a high emission scenario
[RCP8.5]), implying that the type of extreme rainfall that led to the southern
Alberta flooding event will become much more common in the future.
Estimated
return periods were compared between the present-day and pre-industrial
climates, in order to determine the human contribution (see Figure 4.22).
Including human GHG emissions resulted in shorter return periods (the event is
more likely) for three-day maximum precipitation over the entire southern
Alberta region than for pre-industrial levels (see Figure 4.22a). Weather and
climate variability tends to be larger for smaller regions, resulting in a
smaller ratio between the anthropogenic influence and natural internal
variability. Therefore, the anthropogenic influence on the rainfall events over
the smaller Bow River basin region was less pronounced (see Figure 4.22b).
There is no discernable difference for maximum three-day surface runoff, and
thus no anthropogenic influence is detected for this variable (see Figure
4.22c). The authors acknowledge the uncertainties in modelling complex surface
hydrological processes and suggest that any increase in rainfall could be
offset by decreases in snowpack or frozen ground conditions.
In summary,
human influence was detected for the flood-producing rainfall, particularly
over the larger region, but human influence could not be detected for the flood
itself. A flood event is the result of many factors in addition to the amount
of rainfall, including the ground conditions, characteristics of the snowpack,
and size and orientation of the storm. As a result, two events with the same
rainfall amounts do not necessarily produce floods of the same magnitude. The
complex hydrological processes that occur after the rainfall reaches the ground
add additional uncertainty which decreases the ability to detect human
influence.
Increased
GHG emissions, largely due to human activities, result in increased
temperatures. Increased temperatures allow more moisture to be available in the
atmosphere for precipitation, leading to increased intensity of extreme
rainfall events. The 2013 southern Alberta flood was the result of a
combination of many factors, and this study demonstrated that human-induced
emission of GHGs had increased the likelihood of an extreme amount of
precipitation in southern Alberta, at least as large as the amount observed
during this event.
No comments:
Post a Comment