35
temperature and the large scale fields from ERA-40 reanalysis, over the period 1960–
2002 has been used in order to set-up SDs model. Once the most skilful model is
selected for each season and predictand, this is then applied to the predictors simulated by AOGCMs experiments in the framework of A1B emission scenario, such as
to evaluate the local future scenarios of seasonal temperature. The SDs scheme here
proposed use as predictors a selection of fields between mean sea level pressure
(MSLP), 500 hPa geopotential height (Z500), and temperature at 850 hPa (T850),
already tested over Emilia-Romagna and N-Italy region (Tomozeiu et al. 2013).
These fields (predictors) derived from ERA40 re-analysis (http://www.ecmwf.int/
products/) have a spatial resolution of 1.125° × 1.125°, cover the window 90°W–90°E
and 0°–90°N and are referred to the period mid-1957 (September) to mid-2002
(August). As regards the AOGCMs predictors from the ENSEMBLES –STRAEM1
(Van der Linden and Mitchell 2009) runs had been used, over the period 1961–1990
(control-run) and 2021–2050 (A1B scenario). These fields are archived in the Climate
and Environmental Retrieval and Archive (CERA data base) of the World Data
Center System for Climate (WDC) and the access at the data is given by http://
ensembles.wdc-climate.de. The STREAM1 simulations (http://www.ensembles-eu.
org/), used in the present work have been performed with the methodology and the
forcing that were defined for the CMIP3 simulations contributing to the IPCC AR4
assessment. Thus, the experiments were done using a common set of agreed forcing
for historical simulations over the period 1860–2000, and for the three IPCC scenario
A1B, A2, B2, over the twenty-first century. The scenarios were started from an initial
condition obtained for year 2000 in the historical simulation. Several runs, produced
by the following modelling groups have been take into account in the present work:
INGV, NERSC, FUB, IPSL, METOHC (2 runs), MPIMET+DMI. The statistical
downscaling scheme (CCAReg scheme) was applied to each seasons and each predictands (seasonal minimum and maximum temperature). The presence of different
AOGCMs gives the opportunity to construct an Ensemble Mean (EM) of climate
projections. The results obtained by applying the outputs of the AOGCMs to the
CCAReg scheme at Bologna station are presented bellow.
1.3.5.3 Results
The future changes are presented in terms of probability density functions (PDFs)
of seasonal minimum and maximum temperature, which provide a good estimation
of changes not only in the mean but also in the extreme values. As could be noted
from Fig. 1.21, that presents the PDFs of changes in winter minimum temperature
as projected by the CCAReg scheme applied to each AOGCM, all the outputs
emphasizes an increase in the winter Tmin between 0.7 (BCCR and ECHAM5
models) up to 1.8 °C (IPSL and EGMAM run2), over the period 2021–2050 with
respect to 1961–1990.
As concerns the other seasons, the Ensemble Mean of changes in minimum temperature computed taking into account all runs, reveals an increase of temperature
in all seasons (see Fig. 1.22), around 1.5 °C during spring and autumn and 2.5 °C
during summer.
1 Forecasting Models for Urban Warming in Climate Change
temperature and the large scale fields from ERA-40 reanalysis, over the period 1960–
2002 has been used in order to set-up SDs model. Once the most skilful model is
selected for each season and predictand, this is then applied to the predictors simulated by AOGCMs experiments in the framework of A1B emission scenario, such as
to evaluate the local future scenarios of seasonal temperature. The SDs scheme here
proposed use as predictors a selection of fields between mean sea level pressure
(MSLP), 500 hPa geopotential height (Z500), and temperature at 850 hPa (T850),
already tested over Emilia-Romagna and N-Italy region (Tomozeiu et al. 2013).
These fields (predictors) derived from ERA40 re-analysis (http://www.ecmwf.int/
products/) have a spatial resolution of 1.125° × 1.125°, cover the window 90°W–90°E
and 0°–90°N and are referred to the period mid-1957 (September) to mid-2002
(August). As regards the AOGCMs predictors from the ENSEMBLES –STRAEM1
(Van der Linden and Mitchell 2009) runs had been used, over the period 1961–1990
(control-run) and 2021–2050 (A1B scenario). These fields are archived in the Climate
and Environmental Retrieval and Archive (CERA data base) of the World Data
Center System for Climate (WDC) and the access at the data is given by http://
ensembles.wdc-climate.de. The STREAM1 simulations (http://www.ensembles-eu.
org/), used in the present work have been performed with the methodology and the
forcing that were defined for the CMIP3 simulations contributing to the IPCC AR4
assessment. Thus, the experiments were done using a common set of agreed forcing
for historical simulations over the period 1860–2000, and for the three IPCC scenario
A1B, A2, B2, over the twenty-first century. The scenarios were started from an initial
condition obtained for year 2000 in the historical simulation. Several runs, produced
by the following modelling groups have been take into account in the present work:
INGV, NERSC, FUB, IPSL, METOHC (2 runs), MPIMET+DMI. The statistical
downscaling scheme (CCAReg scheme) was applied to each seasons and each predictands (seasonal minimum and maximum temperature). The presence of different
AOGCMs gives the opportunity to construct an Ensemble Mean (EM) of climate
projections. The results obtained by applying the outputs of the AOGCMs to the
CCAReg scheme at Bologna station are presented bellow.
1.3.5.3 Results
The future changes are presented in terms of probability density functions (PDFs)
of seasonal minimum and maximum temperature, which provide a good estimation
of changes not only in the mean but also in the extreme values. As could be noted
from Fig. 1.21, that presents the PDFs of changes in winter minimum temperature
as projected by the CCAReg scheme applied to each AOGCM, all the outputs
emphasizes an increase in the winter Tmin between 0.7 (BCCR and ECHAM5
models) up to 1.8 °C (IPSL and EGMAM run2), over the period 2021–2050 with
respect to 1961–1990.
As concerns the other seasons, the Ensemble Mean of changes in minimum temperature computed taking into account all runs, reveals an increase of temperature
in all seasons (see Fig. 1.22), around 1.5 °C during spring and autumn and 2.5 °C
during summer.
1 Forecasting Models for Urban Warming in Climate Change