Future climate change in Northern Ireland:

1
Future climate change in Northern Ireland From
downscaling to dissertations
Thomas Crawford and David Favis-Mortlock Environme
ntal Change Research Cluster
2
Get reading!
If youre doing your dissertation on climate
change in NI, there are several texts that are
essential to have read. Youre simply asking for
trouble if you havent read them. 1. The SNIFFER
report (QUB, DFM editor, Dr Betts climate
section, agriculture section) (available to
download) 2. The EHS climate change indicators
for NI report (available to download) 3. Soil and
Environment Northern Ireland (QUB, climate and
agriculture sections) 4. Northern Ireland
Environment and Natural Resources (QUB, as
above) 5. EHS website To get up to speed with
ideas about climate change in the UK / Ireland,
see 1. UKCIP reports / website 2. Met Office /
Met Éireann websites 3. IPCC reports (global)
3
Who can you believe?
The future climate projections presented here
consist of two basic components 1. General
Circulation Model (GCM) global projections. These
are created by leading atmospheric physics
centres across the globe (e.g. Met Office). These
are fully referenced in the published literature
and form part of the IPCC future climate change
assessments. There is nothing we can do to change
/ modify these. They are they best tools
currently available for projecting global scale
climate change. However, different models create
different projections 2. Downscaled scenarios
for NI. Created at QUB as part of Ph.D. research.
The highest resolution climate change scenarios
currently available for NI. Created using
well-established downscaling techniques.
Methodology has strengths and weaknesses, but
represents a replicable, defensible regional
impacts modelling approach.
4
The current rainfall regime in NI
5
(No Transcript)
6
The big picture
To get an idea about the big picture of future
climate change in NI, we must look at grid
square scale changes. This tells us about the
general changes in seasonal / monthly temperature
and rainfall. We must look at a range of GCM
predictions uncertainty.
7
Making global predictions relevant at the
regional scale
Evidently we need to make large scale projections
relevant for a land area that is small and
heterogeneous NI. You do not need to know
about the details of the techniques used to
achieve this. The suite of methods are referred
to as downscaling techniques. The information
you will see today was created using delta
change and statistical downscaling
approaches. Other, more general information is
available for the UK as a whole UKCIP reports.
8
The big picture changes predicted for NI as a
whole
9
General statements
N.B. Uncertainty models / scenarios / different
sites / time periods So, if youre making general
statements, pick a time period and model(s)
etc. All months are likely to experience a
temperature increase, the severity of this
increase is model dependent Winter rainfall is
likely to increase. This may be manifested as a
combination of more frequent wet days and more
precipitation on those wet days. Summer
rainfall is likely to decrease. Combined with
temperature increase, this will place pressure
on water resources. There may be an increase in
late summer / early autumn rainfall associated
with increased convectional activity.
10
The big picture changes predicted for NI as a
whole (2060s)
The general, monthly changes can be applied to a
gridded climatology (5km), to help appreciate the
spatial aspects of change. This makes comparative
statements possible.
11
The big picture changes predicted for NI as a
whole
Comparative statements By the 2060s, the ECHAM
GCM predicts that the Comber / Hillsborough area
of County Down will have the same mean daily
precipitation in January that is currently
experienced by the lower slopes of the
Mournes By the 2060s, the HAD GCM predicts
that parts of the upland Sperrins will receive
the same mean daily precipitation in July that is
currently experienced by their lower slopes
bordering Lough Neagh However, these only refer
to changes in rainfall be careful with your
statements. Comparative statements should be
created with your audience in mind i.e. the type
of person to be answering your questionnaire /
something that they will be able to appreciate
etc.
12
Seasonality of rainfall receipt
In terms of annual rainfall amount, no
statistically significant long-term trend is
evident. However, an increasing trend towards
contrasting behaviour of winter and summer
precipitation is evident. Thus, when discussing
changes in rainfall, it is necessary to look at
sub-annual time scales e.g. seasonal /
monthly This is a function of A marked
decrease in summer rainfall (mean daily) A
moderate increase in winter rainfall (more
frequent wet days and more precipitation in
those wet days) Think of the potential impacts
for farmers wet winter following a dry summer?
13
Annual
Winter
Winter - summer
Summer
Observed data
14
Seasonality of rainfall receipt the future?
Most GCMs agree that there will be an increased
seasonality of rainfall receipt during the 21st
century. This can therefore be described as a
robust prediction. In this instance, the winter
/ summer ratio is used. For example, the Hadley
GCM (Met Office) predicts that (for Helens
Bay) The winter / summer ratio is currently
1.5. By the end of the century, the ratio may
be 2.5. There may be some individual years
when the ratio is extreme very dry summer or
very wet winter (combinations) Variability is
much more pronounced
15
Detailed projections
It is possible to make more detailed projections
using a method known as statistical
downscaling. It produces future projections of
daily rainfall for individual sites. This mean
that we can make much more detailed statements
about certain areas. In the previous slide, we
looked at the general projection for seasonality
at Helens Bay. However, what will a much more
detailed analysis of future projections reveal?
16
(No Transcript)
17
(No Transcript)
18
Detailed projections
So, for many sites in NI, the following may be
said The distribution of daily rainfall is
likely to change. For example, whilst there may
be a certain change in mean daily rainfall, the
more extreme quantiles may change in a complex
way. The Q90 level is likely to increase in
winter and decrease in summer. The maximum 5 day
total may experience dramatic changes in winter.
This is particularly important for NI. An
increase in the amount of rainfall received
during wet spells in winter months will impact
upon surface runoff / flooding / harvest and
planting times. Peaks over threshold are also
likely to increase in winter. This means that
individual intense rainfall events will become
more frequent. This may also mean that these
events contribute much more to the overall total
of rainfall receipt. Think about the impact of
intense rainfall on runoff process / how they
will combine with max 5 day totals / what they
mean for crops etc.
19
A sample impacts analysis
Future climate change will certainly impact, at
least in some way, on almost all natural
systems. One example impacts analysis in NI is
the relationship between future climate change
and hydrological regimes. How will future changes
in rainfall characteristics change the way rivers
in NI behave?
20
A sample impacts analysis
In general, changes to mean daily flow follow
each GCMs projections of changes to mean daily
rainfall. However, projections vary between
catchments, and there are dramatic differences
between models.
21
A sample impacts analysis
Water resource management in NI is primarily
concerned with the disposal of excess water. For
example, see the current news coverage about
water shortages in SE England rarely a problem
here. So, impacts analysis should look at future
wet extremes floods etc.
22
A sample impacts analysis
Floods in NI are often less of a problem than in
other parts of the UK. However, they are still
socially and economically severe at the local /
regional scale. How might a change in flood
frequency impact farmers? See Foresight report.
23
Questionnaire issues
In my undergrad dissertation, I investigated the
way in which changes in rainfall chemistry may
impact upon agricultural practice. Part of this
involved talking with farmers / farming
groups. Key ideas from questionnaire work Make
sure you ask more than 30 people think of the
stats that you can work with (237 module) A wide
spectrum of responses individuals could appear
focussed and certain in their outlook, but there
may be no overall consensus within the sample
group (despite being from a small area) Think of
controls on response gender / age / type of
farmer / economics etc. Some farmers may be
surprisingly up-to-date / scientific /
critical do you know that I collected 175mm
in October this year? Thats nearly three times
as much as last October
24
Questionnaire issues
For the farmers I interviewed, rainfall was
evidently a topical issue. However, recent wet
months were a potential source of bias in
questionnaire response. What would your current
feelings on rainfall / climate be if you had just
lost a field of crop because of a run of intense
rainfall events? Think of including both direct
/ closed (obvious?) and indirect / open
(ambiguous?) questions this varies the scope
for response. Deliberately ambiguous questions
can provoke the most interesting answers reveal
what is most important for an individual. Do not
underestimate the power of economics for shaping
answers. For example, think of recent problems
for pig farmers / poultry farmers etc.
25
Sample questions

• Start with age / location / gender etc.
• How well do you feel that those involved with
  the agricultural community are informed about
  environmental issues?
• To what extent do you feel at risk from
  environmental change?
• At present, what is the greatest risk to the
  success of your agricultural practice?
• Have you observed any evidence of change to the
  environmental conditions in the area that you
  farm?
• Do you believe that sources of pollutants are
  close to your farm or far away?
• Would changes in rainfall characteristics
  (amounts / intensity / seasonality) influence
  your farming practice?
• Which type of agricultural activity will suffer
  most from future environmental change?
• Would you be willing to modify your farming
  practices in response to changes in environmental
  conditions?
• Make questions flow from more general to more
  focussed

26
Questionnaire issues
Comments made on evidence of environmental
change Unpredictable and unusual
weather Unseasonal weather Changes to the
pattern of weather in my lifetime. Weather is
much more unseasonal, namely periods of summer in
winter and winter in summer Rainfall seems to
be greater and concentrated into fewer days. Long
days of rain instead of odd showers But
also Changes to the amount of
wildlife Increase in the number of buildings
that dont fit in with the countryside
27
The way we think about weather / climate
You must also be sympathetic about the way people
think about the weather. Climate change in
Ireland is nothing new That the climate of
Ireland has suffered a considerable change,
almost within the memory of the present
generation, is not only a popular opinion, but is
a doctrine held by intelligent and philosophical
observors. We are told, that the winters, in this
island, have laid aside their ancient horrors,
and frequently assume the mildness and vegetative
powers of spring while summer is represented as
less favourable than formerly, less genial in
promoting vegetation, and less vigorous in
advancing to maturity the fruits of the earth
(Patterson, 1802) I have seen changes in
the pattern of weather in my lifetime. Weather is
much more unseasonal, namely periods of summer in
winter, and winter in summer (Mixed cereal
farmer, 2002)
28
The way we think about weather / climate
Individuals tend to focus on extreme
events There seems to be an inability or an
unwillingness to understand that unusual seasons
are part and parcel of this countrys any
countrys climate. We should also be aware that
the human memory is a notoriously poor means of
recalling weather events. Extremes tend to stick
in the memory, so we remember floods and
droughts, snowstorms and damaging gales,
heatwaves and intense cold. Hot summers like
1959, 1976 and 1995 stick in the memory, as do
severe winters such as 1947 and 1963, but we do
not retain specific memories of warm winters or
poor summers (Eden, 2003) So, in NI, change
is the norm, but change is not something that is
easily remembered. The memory of events will
evolve/erode over time, depending on the
individual, type / style of report, or indeed the
tendency for tales to grow in the telling! It may
often be possible for extremes to be
over-represented.
29
Memory and probability
Popular understanding of climate science can be
hindered by jargon-filled literature. Where this
is the case, extreme-event research is likely to
be misunderstood by the public. So, make your
questions / comments straightforward. Individuals
often state an observable change in the
frequency of extreme events in their lifetime,
but this is often related to complex factors
surrounding individual events. In this respect, a
full appreciation of the frequency and magnitude
of events is vital emphasis may be placed on
one-off, dramatic events (high magnitude, low
frequency), but more common events (high
frequency, low magnitude), which may nonetheless
have a greater cumulative impact, could well be
forgotten.
30
Memory and probability
With reference to objective probability, this
means that individuals may overestimate the
probability of relatively infrequent events such
as damaging urban floods, but underestimate the
probability of relatively frequent, but less
dramatic events such as flooding on agricultural
land. Geographical location may well be a
controlling factor in the recollection of events.
Patt and Schrag (2003) state that, peoples
interpretation of probability descriptors depends
on the background frequency of an event.
Mathematically this may provoke a wry smile, as a
flood-prone area with a 30 chance of having a
flood this winter and a usually dry area with a
30 chance have equal probabilities - but in real
terms it is understandable that individuals in
the flood-prone area will feel at more risk.
So, when you ask people questions be aware of
where they come from!
31
Testing memories!
It is possible to test (within reason) the
accuracy of some memories. One possible
example Q Have you observed an increase in
extreme rainfall events? A Yes, I have
noticed much more heavy winter rainfall during
the past five years Test Perform statistical
analysis on winter rainfall record from the site
closest to the farm. If there has been no upward
trend, what has conditioned this answer news /
media / economics / crises? However, be careful!
32
General advice for dissertations
Attempt a topic that is Different /
interesting Possible to achieve within the time
that you have! Involves some type of statistical
/ data analysis Has a point the who cares?
question! That there is at least some other
material on Interesting / relevant for you! Do
not attempt a topic that is Investigated every
year as a dissertation Irrelevant / of no real
importance Has never been studied before / by
anyone! Unrealistic given the time / resources
that you have