#11 Real-impact Attribution
Imagine instead of attributing meteorological events, the
attribution community could also assess real impacts?
Dr.Otto introduced this to me as “another
interesting frontier” where the aim is “not to attribute meteorological events
but real impacts” that the community are hoping to work on more, or “do at
all”. She added that “there are very few publications as it’s rather tricky to
do this in a robust way” but linked me up with some that had been published.
Remember way back in post 4 we looked at the 2003 Extreme Heat Wave
that hit Europe and was responsible for up to 70,000 deaths. This paper claims
to be different because it: "explicitly quantifies" the role of human
activity on heat-related mortality in an event attribution framework.
 |
| Brighton Beach? - 2003 - 38.5 degrees |
What was discovered?!
Anthropogenic climate change:
- Increased the risk of heat
mortality by 70 percent in central Paris, and 20 percent in London (which
had experienced a relatively lower extreme heat)
- Attributable for 506/735
deaths caused by the heatwave in Paris, and 64/315 in Greater London
However, the models struggled to predict future trends as they
will never be truly robust without a complete incorporation and understanding
of societal and physical factors i.e. extent of urbanisation and demographics,
in order to be able to plausibly project future heat related mortality.
A similar issue r.e. future projections is shown in the
study on the 2014 southern England floods, where the results suggested that for
events as seen in 2014, anthropogenic climate change is making over 1000 properties more at risk from
flooding...but in reality the range of results
spanned from 4000-less to
8000-more properties at risk...struggling to take into account the
continuing development on the Thames flood plain and the future of possible
investment strategies.
A 'standard' event attribution study had previously proved this flooding was made 7 times more likely by anthropogenic warming, a lot more definitive than the conclusions drawn from the real-impact study. So, I think using the real-impact attribution alone may, at the moment, be too weak an attribution method if we are trying to enforce legislation changes. However, in time, with real, tangible impacts discussed, it may actually become the best tool!
A 'standard' event attribution study had previously proved this flooding was made 7 times more likely by anthropogenic warming, a lot more definitive than the conclusions drawn from the real-impact study. So, I think using the real-impact attribution alone may, at the moment, be too weak an attribution method if we are trying to enforce legislation changes. However, in time, with real, tangible impacts discussed, it may actually become the best tool!
The standard theory is that human-induced climate change
will increase global terrestrial carbon uptake. However, evidence is showing that extreme events like droughts and storms can actually decrease regional
ecosystem carbon stocks and thus actually negate this expected terrestrial
carbon uptake.
So did this concur with the recent real-impacts?
Sippel et al., 2017 created a factorial experiment to show
the impacts of ecosystem productivity extremes (EPE) in 6 European regions. They proved climate extremes DO have the
power to cause not just changes, but 'extreme' changes in the functioning of
terrestrial ecosystems. Extreme heat did increase spring productivity and
thus increased carbon uptake, however, these effects were either neutralised or
undone by decreased summer productivity and depleted soil moisture further
increasing summer losses. However, the
simulations ignore a number of ecosystems processes and potential feedbacks
between these e.g. wind disturbance, pests, nitrogen and phosphorous
limitations.
So, as Dr.Otto suggests, this is not a simple task, but it could
become a pretty relevant one with regards to mitigating the effects of extremes
on everyday life.
Hi Louise, a question for you regarding this post.
ReplyDeleteI am confused as to a couple of the stats you use.
You mention future projections of floods in southern England, with 1,000 properties more at risk from flooding.
Then go on to say the range of results span from 4,000 less to 8,000 more properties at risk.
Perhaps I am misreading this post, but can you clarify how you can have 4,000 less than 1,000?
Surely then we are talking negative numbers where in reality the lowest number of properties at risk must be ZERO?
Hi Bryan, really glad you raised this! Essentially, the model results were given in this format, but the idea is that the result span was huge and included both an increase in properties at risk (9000 more than current) and a reduction in those already at risk (up to 3000 less). These results, therefore, should hopefully show how difficult it is to accurately model future 'real-impacts' as a result of extreme weather in an anthropogenic climate.
DeleteAnd actually, I think the point you raise perfectly highlights the issues of 'phrasing' we have discussed over the weeks, especially in the EEA post. Scientists are still debating over the best way to phrase research questions and results, so glad to see you getting involved!
DeleteBest,
Louise