Environmental RTDI Programme 2000–2006 CLIMATE CHANGE: Regional Climate Model Predictions for Ireland (2001-CD-C4-M2) Prepared for the Environmental Protection Agency by Community Climate Change Consortium for Ireland Authors: Ray Mc Grath, Elisa Nishimura, Paul Nolan, Tido Semmler, Conor Sweeney and Shiyu Wang ENVIRONMENTAL PROTECTION AGENCY An Ghníomhaireacht um Chaomhnú Comhshaoil PO Box 3000, Johnstown Castle, Co. Wexford, Ireland Telephone: +353 53 60600 Fax: +353 53 60699 E-mail: [email protected]Website: www.epa.ie
52
Embed
CLIMATE CHANGE: Regional Climate Model Predictions for ... · Regional Climate Model Predictions for Ireland (2001-CD-C4-M2) ... from a new regional climate modelling facility located
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Environmental RTDI Programme 2000–2006
CLIMATE CHANGE:
Regional Climate Model Predictions for Ireland
(2001-CD-C4-M2)
Prepared for the Environmental Protection Agency
by
Community Climate Change Consortium for Ireland
Authors:
Ray Mc Grath, Elisa Nishimura, Paul Nolan, Tido Semmler,
Conor Sweeney and Shiyu Wang
ENVIRONMENTAL PROTECTION AGENCY
An Ghníomhaireacht um Chaomhnú ComhshaoilPO Box 3000, Johnstown Castle, Co. Wexford, Ireland
C4I is supported and funded by the Environmental Protection Agency (under the National DevelopmentPlan), Met Éireann, Sustainable Energy Ireland, and the Higher Education Authority.
This work was carried out in collaboration with the CosmoGrid Project, funded under the ProgrammResearch in Third Level Institutions (PRTLI) administered by the Irish Higher Education Authority undNational Development Plan and with partial support from the European Regional Development Fund.
Support from Met Éireann staff, particularly those in the IT and Climate and Observations Divisions, acknowledged.
The C4I Project was supported by the Meteorology and Climate Centre, University College DNumerical integrations were carried out on the high-performance computer facility at UCD and European Centre for Medium-Range Weather Forecasts (ECMWF) in Reading. C4I is registered as aProject with ECMWF.
Support from the Rossby Centre, Sweden, with the modelling work, is also acknowledged.
DISCLAIMER
Although every effort has been made to ensure the accuracy of the material contained in this pubcomplete accuracy cannot be guaranteed. Neither the Environmental Protection Agency nor the aaccept any responsibility whatsoever for loss or damage occasioned or claimed to have been occaspart or in full, as a consequence of any person acting, or refraining from acting, as a result of acontained in this publication. All or part of this publication may be reproduced without further permisprovided the source is acknowledged.
ENVIRONMENTAL RTDI PROGRAMME 2000–2006
Published by the Environmental Protection Agency, Ireland
PRINTED ON RECYCLED PAPER
ISBN:1-84095-166-4Price: €15 05/05/500
ii
Details of Project Partners
Ray McGrath, Met Éireann, Glasnevin Hill, Dublin 9Ireland
CLIMATE CHANGE – Regional climate model predictions for Ireland
5 The Impact of Climate Change on River Flooding underDifferent Climate Scenarios
Summary of contents: Application of a standard hydrological
model to the Suir catchment area shows that the model is
capable of capturing the variability of river discharge with
reasonable accuracy when driven by observations (calibration) or
high-resolution data from a Regional Climate Model (validation).
When driven by the projected precipitation data from the RCM for
the period 2021–2060, the hydrological model shows a
significant increase in the more intense discharge episodes, and
an increase in the frequency of extreme discharges.
5.1 Introduction
The IPCC has stated that mean surface temperatures
may rise 0.3–0.6°C per decade in the 21st century (IPCC,
2001). As increased temperatures will lead to greater
amounts of water vapour in the atmosphere and an
accelerated global water cycle, it is reasonable to expect
that river catchment areas will be exposed to a greater risk
of flooding. Many impact studies have already been
carried out to assess such risks in other countries
(Bergstrom et al., 2001; Pilling and Jones, 2002; Arnel,
2003). This study examines the risks for Ireland using the
Suir catchment area as a test case.
The Land Surface Parameterisation (LSP) scheme is an
important part of the RCM; it acts as a bridge connecting
the atmosphere and water cycle. Significant efforts have
been made to improve the representation of the land
surface–atmosphere interaction during the last two
decades, particularly for the hydrological component.
However, because of the different spatial resolutions of
the climate and hydrological models, it is still difficult to
couple the models directly. For the hydrological model,
the most important processes in the context of climate
change and river flooding are known to be precipitation
and evapotranspiration. In this study, the precipitation
values from different RCM simulations are used to drive
the hydrological model in the Suir catchment area. For
evapotranspiration data, the monthly mean climate values
from Johnstown Castle are used as proxies in the
catchment area.
5.1.1 The HBV hydrological modelThe hydrological discharge model (HBV) of the SMHI is
used in this study (Bergstrom, 1995; Lindstrom et al.,
1997). The model is a semi-distributed, conceptual
hydrological model using sub-basins as the primary
hydrological units; it takes into account area-elevation
distribution and basic land-use categories (forest, open
areas and lakes). The sub-basins option is used in
geographically or climatologically heterogeneous basins
or large lakes. It has been widely used in Europe and
other parts of the world in applications such as
hydrological forecasting, water balance mapping and
climate change studies.
5.1.2 Data sets
To investigate the influence of climate change on regional
water resources and flooding, three global data sets were
used to drive the RCM. For the past climate (1961–2000),
ERA-40 and ECHAM4 data were used, while for the future
climate simulation (2021–2060) the model was driven by
ECHAM4 data consistent with the SRES-B2 scenario
(see Appendix IV for details). As the ERA-40 data are
based on observations and are generally regarded as
providing an accurate description of the atmosphere, they
provide an excellent means for testing the performance of
the climate model in a hydrological application. To
consider the effect of the different boundary data on the
future climate simulation run, the ECHAM4 past climate
simulation was used as a control.
5.2 Results
5.2.1 Calibration
In the HBV model, the parameters with the largest
uncertainty are related to the soil moisture
parameterisation scheme. The main parameters are FC
(maximum soil moisture storage in millimetres), LP
(fraction of FC above which potential evapotranspiration
occurs and below which evapotranspiration will be
reduced) and the coefficient BETA (determining the
relative contribution to run-off from a millimetre of
precipitation at a given soil moisture deficit). These
parameters are dependent on the properties of the
catchment, such as the land-use type, the wilting point
and soil porosity. Because of the uncertainty, the Monte
Carlo Random Sampling (MCRS) method is popularly
25
R. McGrath et al., 2001-CD-C4-M2
used for the parameter estimation. However, as the HBV
program source code is not available, the above method
is difficult to apply. In order to overcome this obstacle,
quasi-stratified sampling in the form of Latin Hypercube
Sampling (McKay et al., 1979) was used. In this method,
the limited sampling numbers can produce similar results
to the Monte Carlo approach (Yu et al., 2001).
For the calibration of the Suir catchment run, observed
precipitation data for the period January 1960 to
December 1964 and monthly mean climate
evapotranspiration data were used to drive the HBV
model. The actual catchment area and rainfall stations are
shown in Fig. 5.1. Note that the calibration period includes
relatively dry and wet years. Although insufficient
observation data coverage limited the duration of the
calibration to 5 years, it should be sufficient according to
the model documentation of SMHI, which recommends
the use of 5–10 years of data. The performance of the
model was judged using a modified R2 statistical
correlation measure, defined as follows.
where QC represents computed discharge, QR is
observed discharge and QRmean is the mean of QR over
the calibration period.
For the calibration of Suir catchment data, the R2 value
(unity for perfect performance) was 0.787.
Figure 5.2 presents the calibration results. Except for the
peak values, which are slightly underestimated, the
variation in the simulated discharge coincides with the
observed discharge fairly well.
5.2.2 ValidationIn the validation run, the parameter values are kept the
same as in the calibration, but simulations are repeated
with independent input series from different present-day
simulation results. Results are shown in Fig. 5.3. The
evolution of the simulated discharge shows good
agreement with observed data, similar to the calibration
run and with peak values underestimated. On the whole,
the simulation is a little worse compared to the calibration
run with an R2 value of only 0.545, while the correlation
coefficient reaches 0.79. This confirms that the model
simulates the evolution of the discharge well, whereas the
underestimated peak values caused the R2 value to be
relatively low. Figure 5.4 shows the return values of the
annual extremes of the observation and ERA-40-driven
simulation. The distribution of return values for the
different return periods show good agreement, although
they are systematically underestimated by 15–20% in the
simulation.
5.2.3 FutureFigure 5.5 shows the impact of climate change on the river
discharge. For the past climate, the ECHAM4-driven
Figure 5.1. Suir catchment area and rainfall stations.
R2 Σ QR QRmean–( )2 Σ QC QR–( )2–( )
Σ QR QRmean–( )2------------------------------------------------------------------------------------------------=
26
CLIMATE CHANGE – Regional climate model predictions for Ireland
Figure 5.2. Observed and simulated (using observed precipitation data to drive the HBV model) discharge (m3/s).
Figure 5.3. Observed and simulated (ERA-40-driven simulation) discharge (m3/s).
Figure 5.4. Return values of observed (a) and simulated (ERA-40-driven simulation) (b) maximum annual
discharge.
(a) (b)
27
R. McGrath et al., 2001-CD-C4-M2
control simulation slightly over-predicts the discharge
compared to the observations (Fig. 5.5a), and the timing
of peak events is slightly shifted. Note, however, that we
cannot expect an exact agreement concerning the timing
of peak events between observation and simulation as the
ECHAM4 data are based on averaged atmospheric
conditions. However, the number of peak values is similar,
which gives us confidence in the future projections. In the
future run (Fig. 5.5b), the frequency and intensity of heavy
discharges (e.g. >350 m3/s) have clearly increased
compared to the control run (Fig. 5.5a). The return value
analysis (Fig. 5.6) also shows similar results with the 10-
year return value increasing from about 290 to 360 m3/s.
Figure 5.7 shows the effect of climate change on the
Figure 5.5. Simulated discharge using the ECHAM4-driven Regional Climate Model simulation for the past and
future climate and observed discharge (m3/s).
(a) (b)
Figure 5.6. Return values of maximum annual discharge using the ECHAM4-driven Regional Climate Model
simulation for (a) present-day climate (1961-2000) and (b) future climate (2021-2060).
28
CLIMATE CHANGE – Regional climate model predictions for Ireland
annual cycle of the discharge: it remains unchanged in the
dry season, but increases by up to 20% in December and
January.
5.3 Conclusions
Application of the HBV model to the Suir catchment area
shows that the hydrological model is capable of capturing
the local variability of river discharge with reasonable
accuracy when driven by observations (calibration) or the
output from an RCM driven by high resolution re-analysis
data (validation). In both cases, the same model
parameters were used in the HBV model.
Using the simulation data from the RCM for the future
period (2021–2060), the hydrological model shows a
significant increase in the more intense discharge
episodes, a pattern that is also shown in the return values
of extreme discharge. This has implications for future
planning to reduce impacts of flooding.
Figure 5.7. Annual cycle of simulated discharge driven by ECHAM4 data for the past (red) and future (green)
climate (m3/s).
29
R. McGrath et al., 2001-CD-C4-M2
6 Cyclone Statistics and Tracks in the Climate Simulations:Past and Future
Summary of contents: To assess the ability of the Regional
Climate Model to realistically reproduce the frequency and
intensity of cyclones, data from the RCM simulation (1961–2000)
driven by standard re-analysis meteorological data, were
compared against the re-analysis data, using an algorithm
developed to identify and track cyclones. Analysis of the
frequency and intensity of cyclones shows that the RCM is in
good agreement with the re-analysis data. The number of weak
cyclones/storms with core pressures between 990 and 1000 hPa
is overestimated by 20 to 30% but the number of intense
cyclones/storms with core pressures of less than 950 hPa is very
well captured.
For the future period, climate data from the ECHAM4-driven
RCM simulation were used (based on the SRES-B2 emission
scenario of GHG concentrations). Results for the reference
(1961–2000) and the future (2021–2060) simulations were
compared. The total number of cyclones with core pressure less
than 1000 hPa and their seasonal and spatial distributions are
similar for both periods. The frequency of very intense cyclones/
storms with core pressures less than 950 hPa shows substantial
changes: a 15% increase in the future simulation with even
stronger increases in winter and spring seasons. Only the
autumn numbers show a decrease in these systems. The tracks
of these storms are also displaced further south relative to those
in the reference simulation.
6.1 Introduction
The damage and disruption caused by intense low-
pressure cyclones or storms can cause significant
hardship for ecosystems and normal socio–economic
activities, particularly in transport, energy and agriculture.
This is most evident in tropical and subtropical regions
which are impacted by hurricane systems. There are
considerable ongoing efforts to improve the prediction of
the intensity and movement of such storms, which even
outside of the hurricane zones can be very damaging and
economically costly. Understanding the impacts of
projected changes in such systems as a result of climate
change is an important task in predicting impacts. Here,
analyses of projected changes in these areas are
presented.
6.1.1 Tracking cyclones
The following algorithm was used to track the movement
of cyclones.
On a horizontal grid of Nx × Ny points a low-pressure
centre is identified at grid point (i, j) if these criteria are
met:
• the pressure value, p(i, j), is less than a given
threshold, e.g. 1000 hPa;
• the point is a local minimum of pressure.
A point (i, j) is defined to be a local minimum of surface
pressure if its pressure value p(i, j) is less than the
pressure values of all points contained in the surrounding
four grid boxes (i.e. the surrounding 80 grid points, see
Fig. 6.1). This corresponds to p(i, j) being the minimum
pressure within a distance of approximately 53 km in all
directions. Since for each point we examine the four
surrounding grid boxes, we do not take into account the
boundary points and the three nearest points to the
boundaries.
For example, a point will be identified as the centre of a
cyclone if for the pressure value p(i, j), we have:
p(i, j) < 1000 hPa
and
p(i, j) < p(i + k, j + l), k = –4,…..,+4 (k ≠ 0)
where
4 < i < Nx – 3 and 4 < j < Ny – 3.
6.1.2 Cyclone statisticsThe algorithm was applied to the RCM results and the
ERA-40 re-analysis data. As the ERA-40 data are only
available at 6-h intervals, the same interval was used for
the RCM data. Figure 6.2a shows the frequency of lows
as a function of core-pressure value for both data sets.
The data bins along the x-axis have an increment of 1
hPa.
Although both curves in Fig. 6.2a follow a similar trend,
the RCM returns substantially more lows, particularly at
higher pressure values. The discrepancy arises from the
fact that the RCM tends to split a large-scale low-pressure
system with only one local minimum in the ERA-40 data
30
CLIMATE CHANGE – Regional climate model predictions for Ireland
into several local minima that are counted as separate
lows in our algorithm (Fig. 6.3). This is a common
occurrence in RCM simulations and since the local
minima obviously belong to the same system we
introduce an extra constraint in the algorithm: two local
minima are considered to belong to the same system if
they are less than 1000 km apart. The distance of 1000
km was chosen since this is a typical extension of a low-
pressure system. With this constraint the low-pressure
system is assigned the minimum pressure value and its
position is taken to be the position of this minimum. The
revised statistics (Fig. 6.2b) show much better agreement
with ERA-40.
Figure 6.4a shows the annual number of lows with a core
pressure of less than 1000 hPa. The patterns are very
similar for the RCM and ERA-40 data with the model
numbers being slightly higher. This is also reflected in the
ratio plot shown in Fig. 6.4b.
The seasonal number of lows over the 40-year period is
depicted in Fig. 6.4c. Both the RCM and the ERA-40 data
follow the same trend with the greatest and lowest number
of systems occurring in winter and summer, respectively.
Figure 6.4d shows the spatial distribution of the cyclones
split into quadrants. The percentages are broadly similar
but the regional model has slightly more lows in the
eastern half of the area compared with ERA-40. As
expected, most of the lows fall into the northern half of the
area.
The time-series plot of the number of lows (Fig. 6.5)
shows that the RCM is capable of capturing the seasonal
variations.
Figure 6.1. Locating a low point.
Figure 6.2. Frequency of lows as a function of core pressure (a) without using a distance constraint, and (b)
using a distant constraint. Frequency is shown for the ERA-40 re-analysis data (black) and the Regional Climate
Model driven by the re-analysis data (red).
(a) (b)
31
R. McGrath et al., 2001-CD-C4-M2
Figure 6.3. An example of (a) one local minimum in ERA-40 and (b) three local minima in the Regional Climate
Model.
Figure 6.4. (a) ERA-40 and Regional Climate Model annual number of lows, (b) ratio of annual number of lows
(RCM/ERA-40), (c) total seasonal number of lows, and (d) spatial distribution of lows with a core pressure of less
than 1000 hPa in the regional model and ERA-40 for the time period 1961–2000.
(a) (b)
(a) (b)
(c) (d)
32
CLIMATE CHANGE – Regional climate model predictions for Ireland
6.1.3 Intense cyclones
In this section, statistics are presented for cyclones with
core pressures less than 950 hPa. Note that with the lower
pressure threshold the RCM results agree more closely
with the ERA-40 re-analysis data. This is evident from the
plot depicting the annual number of lows (Fig. 6.6a). The
seasonal and spatial distributions of the cyclones are
shown in Figs 6.6b and 6.6c, respectively. As expected,
most of the intense cyclones occur in winter and are
located in the northern half of the area. A time-series plot
of the seasonal number of lows (Fig. 6.7) shows good
agreement between the RCM and ERA-40. Note the
increase in frequency in the 1990s.
It is encouraging that the results compare well for intense
cyclones not only in the annual numbers, but also in the
spatial and seasonal distributions, since extreme events
are interesting in terms of impacts.
Figure 6.6. (a) Annual, (b) seasonal and (c) spatial distribution of the number of lows with a core pressure of less
than 950 hPa in ERA-40 and the Regional Climate Model.
(a) (b)
(c)
Figure 6.5. Seasonal number of lows with a core pressure of less than 1000 hPa from ERA-40 and the Regional
Climate Model for 1 961–2000.
33
R. McGrath et al., 2001-CD-C4-M2
6.2 Cyclone Tracking
In this section, the spatial movement of cyclones with a
core pressure less than 950 hPa is examined. To track the
movement of the systems the cyclone centres are located
at output times t and t + ∆, where ∆ is the data output
frequency, set to 6 h to suit the ERA-40 data. A cyclone at
time t is considered to be the same cyclone identified at
time t + ∆ if the estimated speed of movement, based on
the great circle distance between the positions, is less
than 120 km/h. In addition, only cyclones that exist for at
least 24 h are considered.
Applied to the RCM and ERA-40 fields (Fig. 6.8) we can
see that the RCM is in good agreement with ERA-40.
Results for one particular storm that caused widespread
damage when it crossed Ireland and the UK on 25
January 1990 are shown in Fig. 6.9. The RCM track of this
so-called Burns’ Day storm, while a little too far south,
agrees well with the ERA-40 track. However, the
intensification of the system lags ERA-40 although it does
eventually catch up with an extreme pressure of 950 hPa
(Fig 6.10).
6.3 Simulation of the Future Climate
This section examines the impact of climate change on
the frequency and intensity of cyclones, using the two
ECHAM4-driven RCM simulations for 1961–2000
(reference simulation) and 2021–2060 (future simulation)
described in Chapter 4. Data at 3-h intervals were used.
Figure 6.7. Seasonal number of lows with a core pressure of less than 950 hPa from ERA-40 (top) and the
Regio nal Cl imate Model (bott om) for 1961–2000.
(a)
(b)
Figure 6.8. Tracks of storms with a core pressure of less than 950 hPa and with a lifetime of at least 24 h from (a)
ERA-40 and (b ) the Regio nal Climate Model for th e time period 1961–2000.
(a) (b)
34
CLIMATE CHANGE – Regional climate model predictions for Ireland
Figure 6.11 shows the frequency of lows as a function of
core-pressure value for both sets of data. The differences
between the simulations are generally small but for the
higher core-pressure values the number of cyclones in the
future simulation is slightly less compared with the
reference simulation.
Figure 6.12 shows the seasonal and spatial distribution of
lows with a core pressure of less than 1000 hPa. Again,
the differences between the simulations are small and
similar results are found when the core-pressure
threshold is reduced to 960 hPa.
However, with a core-pressure threshold of 950 hPa a
different pattern emerges (Fig. 6.13). Even if there is a
substantial decadal variability (Fig. 6.13a), there are at
least three peaks in the annual number of lows in the
future simulation, which exceed all the peaks in the
reference simulation. Furthermore, the intense storms are
more frequent in winter and spring (Fig. 6.13b), whereas
there is a substantial decrease in autumn. Although the
spatial distributions are similar (Fig. 6.13c), the future
simulation shows more of these intense cyclones in the
south-east quadrant, which is the quadrant with the
highest land proportion and therefore more important in
terms of impacts. In the future simulation, the storm tracks
also extend further south relative to the reference
simulation (Fig. 6.14).
6.4 Conclusions
The RCM has been shown to realistically reproduce the
frequency and intensity of cyclones in the current climate.
The number of intense cyclones with core pressures of
less than 950 hPa is very well captured by the model.
Comparison of results for the reference (1961–2000) and
future (2021–2060) simulations show that for cyclones
with core pressures less than 1000 hPa the total numbers
and the seasonal and spatial distributions are similar
during both periods. However, the frequency of very
intense cyclones with core pressures less than 950 hPa,
Figure 6.9. Tracks of th e Burn s’ Day stor m fro m ERA-
40 (black arrows ) and the Region al Cl imate Model
(red arrows) fro m 00 UTC 25 January 1990 to 00 UTC
26 January 1990.
Figure 6.10. Core-pressure v alues of the Burn s’ Day
stor m fro m ERA-40 and the Regio nal Clim ate Model
for t he perio d 00 UTC 25 January 1990 to 00 UTC 26
Januar y 1990.
Figure 6.11. Number of lows as a function of core-
pressure v alue f or the reference 1961–2000 (blue
line) and future 20 21–2060 (red lin e) simulation s.
35
R. McGrath et al., 2001-CD-C4-M2
Figure 6.12. (a) Annual, (b) seasonal and (c) spatial distribution of the number of lows with a core pressure of
less than 1000 hPa for 1961–2000 and 2021–2060 as simulated by the Regional Climate Model driven by ECHAM4
data.
(a) (b)
Figure 6.13. (a) Annual, (b) seasonal and (c) spatial distribution of the number of lows with a core pressure of
less than 950 hPa for 1961–2000 and 2021–2060 as simulated by the Regional Climate Model driven by ECHAM4
data.
(c)
(a) (b)
(c)
36
CLIMATE CHANGE – Regional climate model predictions for Ireland
shows substantial changes: a 15% increase in the future
simulation with even stronger increases in winter and
spring. Only the autumn numbers show a decrease.
Previous studies on cyclone activity in the North Atlantic
region show different results for the total number of lows:
Zhang and Wang (1997) and Lambert (1995) found a
decrease; König et al. (1993) found no significant
changes; Hall et al. (1994) and Sinclair and Watterson
(1999) found an increase. However, the increased
intensity of cyclones in the future climate of the North
Atlantic area close to Ireland and the UK during the winter
months is a common feature in these studies (e.g.
Lambert, 1995; Geng and Sugi, 2003). In terms of
impacts, it is interesting to note that in the future
simulation the very intense cyclones (core pressures
below 950 hPa) extend further south relative to the
reference simulation (Fig. 6.14).
Figure 6.14. Tracks of storms with a core pressure of less than 950 hPa and with a lifetime of at least 12 h from
the Regio nal Cl imate Model drive n by ECHAM4 data for (a) t he period 1961–2000 and (b) th e peri od 2021–2060.
Arr ows show the direct ion of mo vement of in divid ual storm s.
(b)(a)
37
R. McGrath et al., 2001-CD-C4-M2
References
ACIA, 2005. Impacts of a Warming Arctic – Arctic ClimateImpact Assessment. Cambridge University Press,Cambridge, UK, 144 pp.
Arnell, N.W., 2003. Relative effects of multi-decadal climaticvariability and changes in the mean and variability of climatedue to global warming: future streamflow in Britain. Journalof Hydrology 270: 195–213.
Bergstrom, S., 1995. The HBV model. In: Computer Models ofWatershed Hydrology, Chapter 13. Water ResourcesPublications. pp. 443–476.
Bergstrom, S., Carlsson, B., Gardelin, M., Lindstrom, G.,Pettersson, A. and Rummukainen, M., 2001. Climatechange impacts on the runoff in Sweden – assessments byglobal climate models, dynamical downscaling andhydrological modelling. Climate Research 16: 101–112.
Chen, G.T.J., Jiang, Z.H. and Wu, M.C., 2003. Spring heavy rainevents in Taiwan during warm episodes and the associatedlarge-scale conditions. Monthly Weather Review 131: 1173–1188.
Davies, J.R., Kelly, M.P. and Osborn, T., 1997. Explaining theclimate of British Isles. In: Hulme, M. and Barrow, E. (eds)Climates of British Isles: Present, Past and Future. Springer,Berlin. pp. 386–406.
Geng, Q. and Sugi, M., 2003. Possible change of extratropicalcyclone activity due to enhanced greenhouse gases andsulfate aerosols – study with a high-resolution AGCM.Journal of Climate 16: 2262–2274.
Hall, N.M.J., Hoskins, B.J., Valdes, P.J. and Senior, C.A., 1994.Storm tracks in a high-resolution GCM with doubled carbondioxide. Quarterly Journal of the Royal MeteorologicalSociety 120: 1209–1230.
Horel, J.D., 1981. A rotated principal component analysis of theinterannual variability of the Northern Hemisphere 500-mbheight field. Monthly Weather Review 109: 2080–2092.
Hurrell, J.W., 1995. Decadal trends in the North AtlanticOscillation: Regional temperature and precipitation. Science269: 676–679.
IPCC, 2001. Climate Change 2001: The Scientific Basis.Contributions of Working Group I to the Third AssessmentReport of the Intergovernmental Panel on Climate Change.Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van derLinden, P.J., Dai, X., Maskell, K. and Johnson, C.A. (eds).Cambridge University Press, Cambridge, UK, 881 pp.
IPCC, 2000. Special Report on Emission Scenarios. A SpecialReport of Working Group III of the Intergovernmental Panelon Climate Change. Nakicenovic, N. and Swart, S. (eds).Cambridge University Press, Cambridge, UK, 599 pp.
Jolliffe, I.T., 1987. Rotation of principal component: somecomments. Journal of Climatology 7: 507–510.
Kawamura, R., 1994. A rotated EOF analysis of global searsurface temperature variability with interannual andinterdecadal scales. Journal of Physical Oceanography 24:707–715.
Keane, T. and Collins, J.F., 2004. Climate, Weather and IrishAgriculture. pp. 27–62.
Kiely, G., 1999. Climate change in Ireland from precipitation andstreamflow observation. Advances in Water Resources 23:141–151.
Kiely, G., Albertson, J.D. and Parlange, M.B., 1998. Recenttrends in diurnal variation of precipitation at Valentia on thewest coast of Ireland. Journal of Hydrology 207: 270–279.
König, W., Sausen, R. and Sielmann, F., 1993. Objectiveidentification of cyclones in GCM simulations. Journal ofClimate 6: 2217–2231.
Lambert, S.J., 1995. The effect of enhanced greenhousewarming on winter cyclone frequencies and strengths.Journal of Climate 8: 1447–1452.
Lindstrom, G., Johansson, B., Persson, M., Gardelin, M. andBergstrom, S., 1997. Development and test of thedistributed HBV-96 model. Journal of Hydrology 201: 272–288.
Logue, J.J., 1984. Regional variation in the annual cycle ofrainfall in Ireland as revealed by principal componentanalysis. Journal of Climatology 4: 597–607.
McElwain, L. and Sweeney, J., 2003. Climate change in Ireland– recent trends in temperature and precipitation. IrishGeography 36: 97–111.
McKay, M.D., Conover, W.J. and Beckman, R.J., 1979. Acomparison of three methods for selection values of inputvariables in the analysis of output from a computer code.Technometrics 2: 239–245.
Moses, T., Kiladis, G.N., Diaz, H.F. and Barry, R.G., 1987.Characteristics and frequency of reversals in mean sea levelpressure in the North Atlantic sector and their relationship tolong-term temperature trends. Journal of Climatology 7: 13–30.
Namias, J., 1964. Seasonal persistence and recurrence ofEuropean blocking during 1958–1960. Tellus 16: 394–407.
Pilling, C.G. and Jones, J.A.A., 2002. The impact of futureclimate change on seasonal discharge, hydrologicalprocesses and extreme flows in the Upper Wyeexperimental catchment, mid-Wales. HydrologicalProcesses 16: 1201–1213.
Rex, D.F., 1951. The effect of Atlantic blocking action uponEuropean climate. Tellus 3: 100–112.
Richman, M.B., 1986. Rotation of principal component. Journalof Climatology 6: 293–335.
Richman, M.B., 1987. Rotation of principal component: a reply.Journal of Climatology 7: 511–520.
Roeckner, E., Arpe, K., Bengsson, L., Christoph, M., Claussen,M., Duemenil, L., Esch, M., Giorgetta, M., Schlese, U. andSchulzweida, U., 1996. The Atmospheric GeneralCirculation Model ECHAM-4: Model Description andSimulation of Present-Day-Climate. Max Planck Institute forMeteorology, Report No. 218, Hamburg, Germany, 90 pp.
38
CLIMATE CHANGE – Regional climate model predictions for Ireland
Rogers, J.C., 1985. Atmospheric circulation changes associatedwith the warming over the northern North Atlantic in the1920s. Journal of Climate and Applied Meteorology 24:1303–1310.
Rummukainen, M., Raisanen, J., Bringfelt, B., Ullerstig, A.,Omstedt, A., Willen, U., Hansson, U. and Jones, C., 2001. Aregional climate model for Northern Europe: Modeldescription and results from the downscaling of two GCMcontrol simulations. Climate Dynamics 17: 339–359. DOI:10.1007/s003820000109.
Schulz, J., Meywerk, J., Ewald, S. and Schluessel, P., 1997.Evaluation of satellite-derived latent heat fluxes. Journal ofClimate 10: 2782–2795.
Sinclair, M.R. and Watterson, I.G., 1999. Objective assessmentof extratropical weather systems in simulated climates.Journal of Climate 12: 3467–3485.
Stanhill, G., 1998. Long-term trends in, and spatial variation of,solar irradiances in Ireland. International Journal ofClimatology 18: 1015–1030.
Sweeney, J., Donnelly, A., McElwain, L. and Jones, M., 2002.Climate Change Indicators for Ireland. ERTDI Report SeriesNo. 2. EPA, Wexford.
Sweeney, J., Brereton, T., Byrne, C., Charlton, R., Emblow, C.,Fealy, R., Holden, N., Jones, M., Donnelly, A., Moore, S.,Purser, P., Byrne, K., Farrell, E., Mayes, E., Minchin, D.,Wilson, J. and Wilson, J., 2003. Climate Change Scenariosand Impacts for Ireland. ERTDI Report Series No. 15. EPA,
Wexford.
Torrence, C. and Compo, G.P., 1998. A practical guide towavelet analysis. Bulletin of the American MeteorologicalSociety 79(1): 61–78.
Uppala, S.M., Kållberg, P.W., Simmons, A.J., Andrae, U., daCosta Bechtold, V., Fiorino, M., Gibson, J.K., Haseler, J.,Hernandez, A., Kelly, G.A., Li, X., Onogi, K., Saarinen, S.,Sokka, N., Allan, R.P., Andersson, E., Arpe, K., Balmaseda,M.A., Beljaars, A.C.M., van de Berg, L., Bidlot, J., Bormann,N., Caires, S., Dethof, A., Dragosavac, M., Fisher, M.,Fuentes, M., Hagemann, S., Hólm, E., Hoskins, B.J.,Isaksen, L., Janssen, P.A.E.M., McNally, A.P., Mahfouf, J.-F.,Jenne, R., Morcrette, J.-J., Rayner, N.A., Saunders, R.W.,Simon, P., Sterl, A., Trenberth, K.E., Untch, A., Vasiljevic, D.,Viterbo, P. and Woollen, J., 2005. The ERA-40 reanalysis.Quarterly Journal of the Royal Meteorological SocietySubmitted.
Wilby, R.C., O'Hare, G. and Barnsley, N., 1997. The NorthAtlantic Oscillation and British Isles climate variability, 1865–1996. Weather 52: 266–27.
Yu, P.S., Yang, T.Ch. and Chen, S.J., 2001. Comparison ofuncertainty analysis methods for a distributed rainfall runoffmodel. Journal of Hydrology 244: 43–59.
Zhang, Y. and Wang, W.-C., 1997. Model-Simulated NorthernWinter Cyclone and Anticyclone Activity under aGreenhouse Warming Scenario. Journal of Climate 10:1616–1634.
39
R. McGrath et al., 2001-CD-C4-M2
Glossary of Terms
CosmoGrid Network of distributed computing
resources in Ireland (seewww.CosmoGrid.ie)
ECHAM4 European Centre Hamburg Model Version
4. Global climate model developed at theMax Planck Institute for Meteorology inHamburg, Germany
ECMWF European Centre for Medium-RangeWeather Forecasting
ERA-40 Re-analysis project of the European Centrefor Medium-Range Weather Forecasting;archive of re-analysis data
GCM Global Climate Model (also GeneralCirculation Model)
HBV Hydrological discharge model
HOAPS Hamburg Ocean Atmosphere Parameters
and Fluxes from Satellite Data. A satellite-derived global climatology of freshwaterflux (see www.hoaps.zmaw.de), produced
by the Meteorological Institute of theUniversity of Hamburg and the Max PlanckInstitute for Meteorology in Hamburg
IPCC Intergovernmental Panel on ClimateChange
RCM Regional Climate Model
SMHI Swedish Meteorological and HydrologicalInstitute
SRES Special Report on Emissions Scenarios.
The ECHAM4/OPYC3 simulation uses theSRES-B2 scenario, a scenario ofmoderately increasing greenhouse gas
concentrations, for the period 1990–2100.For further details, seewww.grida.no/climate/ipcc/emission/095.htm
SST Sea Surface Temperature
T159 Refers to the resolution of the global ERA-40 data in spherical harmonic format.
Triangular truncation at wave number 159corresponds to grid point resolutions ofapproximately 1.125 degrees. ECHAM4
data have a native horizontal resolution ofapproximately 2.8 degrees (T42)
UKCIP UK Climate Impacts Programme. Resourceprovides gridded observation data for
elements such as temperature andprecipitation
UNFCCC United Nations Framework Convention onClimate Change