Validation and climate projections of theALARO-0 model on the EURO-CORDEX domain
O. Giot1,2, P. Termonia1,3, D. Degrauwe1, R. De Troch1,3, S. Caluwaerts3,G. Smet1, J. Berckmans1,2, A. Deckmyn1, L. De Cruz1, P. De Meutter1,3,F. Duchene1, A. Duerinckx1,3, L. Gerard1, R. Hamdi1, J. Van den Bergh1,
M. Van Ginderachter1,3, B. Van Schaeybroeck1,
1 Royal Meteorological Institute of Belgium2 Centre of Excellence PLECO (Plant and Vegetation Ecology), University of Antwerp
3 Department of Physics and Astronomy, Ghent University
Joint 26th ALADIN Workshop & HIRLAM All Staff Meeting 2016Lisbon, April 4-8, 2016
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 1 / 26
1 ALARO-0 climate runs at RMIB: status
2 Validation of ALARO-0 for climate
3 Climate projections
4 Subdaily precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 2 / 26
1 ALARO-0 climate runs at RMIB: status
2 Validation of ALARO-0 for climate
3 Climate projections
4 Subdaily precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 3 / 26
A little bit of history
Extended downscaling experiment by De Troch et al., JoC, 2013:
Evaluation of ALADIN and ALARO-0 cy36t1 at 40, 10 and 4 km.
Initial and lateral boundary conditions: ERA-40 or model at40km resolution (one-way nesting)
40-year run with daily reinitializations
Reference: to station observations 1961-1990
Thanks to the 3MT physics parameterization scheme, ALARO-0generates consistent results across scales and correctly representsextreme daily precipitation, even at high resolutions.
Results indicate that ALARO-0 is a good candidate for regionalclimate modelling.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 4 / 26
ALARO-0 climate runs at RMIBParticipation in the Coordinated Regional Climate DownscalingExperiment (CORDEX):
Runs are performed with ALARO-0 cy36t1Boundary conditions: ERA-Interim (evaluation)or CMIP5 GCM: CNRM-CM5 (historical and future)Run continuously (one month at a time) for a 31-year period.Domain and resolutions: EUR-44 (0.44◦ ≈ 50 km) and EUR-11(0.11◦ ≈ 12.5 km)
BI
IP
FR
ME
SC
AL
MD
EA
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 5 / 26
ValidationValidated using state-of-the-art performance metrics.
Geosci. Model Dev., 9, 1143–1152, 2016
www.geosci-model-dev.net/9/1143/2016/
doi:10.5194/gmd-9-1143-2016
© Author(s) 2016. CC Attribution 3.0 License.
Validation of the ALARO-0 model within the EURO-CORDEX
framework
Olivier Giot1,2, Piet Termonia1,3, Daan Degrauwe1, Rozemien De Troch1,3, Steven Caluwaerts3, Geert Smet1,
Julie Berckmans1,2, Alex Deckmyn1, Lesley De Cruz1, Pieter De Meutter1,3, Annelies Duerinckx1,3, Luc Gerard1,
Rafiq Hamdi1, Joris Van den Bergh1, Michiel Van Ginderachter1,3, and Bert Van Schaeybroeck1
1Royal Meteorological Institute, Brussels, Belgium2Centre of Excellence PLECO (Plant and Vegetation Ecology), Department of Biology, University of Antwerp,
Wilrijk, Belgium3Department of Physics and Astronomy, Ghent University, Ghent, Belgium
Correspondence to: Olivier Giot ([email protected])
Received: 29 July 2015 – Published in Geosci. Model Dev. Discuss.: 1 October 2015
Revised: 3 March 2016 – Accepted: 4 March 2016 – Published: 30 March 2016
Abstract. Using the regional climate model ALARO-0, the
Royal Meteorological Institute of Belgium and Ghent Uni-
versity have performed two simulations of the past observed
climate within the framework of the Coordinated Regional
Climate Downscaling Experiment (CORDEX). The ERA-
Interim reanalysis was used to drive the model for the pe-
riod 1979–2010 on the EURO-CORDEX domain with two
horizontal resolutions, 0.11 and 0.44◦. ALARO-0 is char-
acterised by the new microphysics scheme 3MT, which al-
lows for a better representation of convective precipitation.
In Kotlarski et al. (2014) several metrics assessing the per-
formance in representing seasonal mean near-surface air tem-
perature and precipitation are defined and the corresponding
scores are calculated for an ensemble of models for differ-
ent regions and seasons for the period 1989–2008. Of special
interest within this ensemble is the ARPEGE model by the
Centre National de Recherches Météorologiques (CNRM),
which shares a large amount of core code with ALARO-0.
Results show that ALARO-0 is capable of representing
the European climate in an acceptable way as most of the
ALARO-0 scores lie within the existing ensemble. However,
for near-surface air temperature, some large biases, which
are often also found in the ARPEGE results, persist. For pre-
cipitation, on the other hand, the ALARO-0 model produces
some of the best scores within the ensemble and no clear re-
semblance to ARPEGE is found, which is attributed to the
inclusion of 3MT.
Additionally, a jackknife procedure is applied to the
ALARO-0 results in order to test whether the scores are ro-
bust, meaning independent of the period used to calculate
them. Periods of 20 years are sampled from the 32-year sim-
ulation and used to construct the 95 % confidence interval for
each score. For most scores, these intervals are very small
compared to the total ensemble spread, implying that model
differences in the scores are significant.
1 Introduction
The climate projections used in the Fifth Assessment Re-
port (AR5) of the Intergovernmental Panel on Climate
Change (IPCC, 2013) are based on the set of global climate
model (GCM) simulations performed within the fifth Cou-
pled Model Intercomparison Project (CMIP5; Taylor et al.,
2011). The horizontal resolution of the contributing GCMs
is limited to typically 1–2◦ by computational constraints. For
many local climate impact studies, regional climate models
(RCMs; Giorgi and Mearns, 1999) are needed to reveal the
fine-scale details of potential climate change (Teutschbein
and Seibert, 2010). In addition, specific downstream models
which simulate processes such as vegetation interactions, ur-
ban effects (e.g. Hamdi et al., 2015) or extreme hydrological
events in river catchments often require high-resolution (both
in time and space) forcing data from atmospheric models.
Published by Copernicus Publications on behalf of the European Geosciences Union.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 6 / 26
Current status
Runs are ongoing on the Tier-1 supercomputer at Ghent University.The checked runs are finished or ongoing, green ones are next.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 7 / 26
National project
CORDEX.be
Dynamical downscaling of EURO-CORDEX 12.5km or 50kmruns on a high-resolution O(4km) domain over Belgium
In addition to our contribution with ALARO-0, partner institutesuse e.g. COSMO-CLM, MAR
This provides an ensemble of high-resolution climate runs forlocal impact modellers.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 8 / 26
National project: CORDEX.be
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 9 / 26
Technical challenges
Creating netCDF files that conform to the CORDEX archivespecifications
Processing 100s of TBs of historical files to extract TBs of data
... in R: new R package CordextractR (flexibility required!)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 10 / 26
Technical challenges
Creating netCDF files that conform to the CORDEX archivespecifications
Processing 100s of TBs of historical files to extract TBs of data
... in R: new R package CordextractR (flexibility required!)
variable 1
variable n
FA file 1 FA file 2 FA file 3 FA file 4
variable 1
variable n
separate spacetime netCDF files for each variable
All variables in one FAfile per model output time interval
CordextractR
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 10 / 26
Technical challenges
Creating netCDF files that conform to the CORDEX archivespecifications
Processing 100s of TBs of historical files to extract TBs of data
... in R: new R package CordextractR (flexibility required!)
Specify your in/output variables of choice, including functionssuch as sum, modulus, mask, threshold, max... in a declarativewaySet-up of a data conversion pipeline usingproducer-filter-consumer pattern (avoids(computation-intensive) logic/branches during the conversion)Fast conversion (in spite of R): the bottleneck is mainly IOUnit tests for all functions
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 10 / 26
Technical challenges
Creating netCDF files that conform to the CORDEX archivespecifications
Processing 100s of TBs of historical files to extract TBs of data
... in R: new R package CordextractR (flexibility required!)
Specify your in/output variables of choice, including functionssuch as sum, modulus, mask, threshold, max... in a declarativewaySet-up of a data conversion pipeline usingproducer-filter-consumer pattern (avoids(computation-intensive) logic/branches during the conversion)Fast conversion (in spite of R): the bottleneck is mainly IOUnit tests for all functions
Submitting data to the ESGF nodes (many of which have beendown for a while...)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 10 / 26
1 ALARO-0 climate runs at RMIB: status
2 Validation of ALARO-0 for climate
3 Climate projections
4 Subdaily precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 11 / 26
Validation of ALARO-0 for climate
Evaluation run:
Lateral boundary conditions from the ERA-Interim reanalysis
Continuous 31-year run (1979-2010)
Reference: E-OBS 7 data set
Can ALARO-0 represent the most important features of theEuropean climate?
In practice:
1 Is ALARO-0 competitive with other EURO-CORDEX ensemblemembers, using the standardized performance metrics as inKotlarski et al., 2014?
2 Are these metrics robust?
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 12 / 26
Performance metrics
Scores are based on seasonal mean values of near-surface airtemperature and precipitation.
BIAS: mean bias
95%-P: 95th percentile of the absolute grid point differences
RSV: ratio of spatial variability
PACO: pattern correlation
RIAV: ratio of interannual variability
TCOIAV: temporal correlation of interannual variability
All scores except TCOIAV should be similar for reanalysis- andGCM-driven runs (if GCMs represent the climate well)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 13 / 26
Performance metrics
Scores are based on seasonal mean values of near-surface airtemperature and precipitation.
BIAS: mean bias
95%-P: 95th percentile of the absolute grid point differences
RSV: ratio of spatial variability
PACO: pattern correlation
RIAV: ratio of interannual variability
TCOIAV: temporal correlation of interannual variability
All scores except TCOIAV should be similar for reanalysis- andGCM-driven runs (if GCMs represent the climate well)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 13 / 26
Robustness
Are the scores robust, i.e. independent of the period used?⇒ Jackknife procedure:
Calculate all scores for 1000 random 20-year samples out of the32-year period
Construct 95% confidence intervals
Compare interval width to the ensemble spread.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 14 / 26
111
domain
1
BI
11111
IP
11111
FR
11111
ME
11111
SC
11111
AL
11111
MD
11111
EA
111
1
1
1
1 season
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1 YEAR
11
11
1
1 BIAS [K]● ●●●●● ●● ●
● ●●● ●● ●●●
●●●● ●●●●●
● ●●●●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●● ● ●● ●●●
●●●● ●●●●●
●●●● ●● ●●●
● ●●● ●● ●●●
● ●●●●● ●● ●
● ●●● ●● ●●●
● ●● ● ● ●●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●●● ●●● ●
●●●● ●●●● ●
● ●● ● ● ●●●●
● ●●●●● ●● ●
● ●●● ●●●● ●
● ●●● ●● ●● ●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●● ●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●● ●● ●● ●
● ●●● ●● ●●●
●●●● ●●●●●
● ●●● ●● ●●●
●●●● ●● ●●●
−5 −2 0 2 4
11
1
1
1 95%−P [K]●●● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●● ●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●●●●● ●●● ●
●● ●●● ●●● ●
●● ●●● ●●● ●
●● ● ●● ●●●●
●● ●●● ●●●●
●●●●● ●●●●
●● ●●● ●● ●●
●● ●●● ●● ●●
●●● ●●●●●●
● ● ●●● ●●●●
●● ●●● ●●●●
● ●●●● ●●●●
●●●●● ●●●●
●● ● ●● ●● ●●
●● ●●● ●● ●●
●● ●●● ●● ● ●
●● ● ●● ●● ●●
●● ● ●● ●● ●●
● ● ●●● ●● ●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
● ● ●●● ●●● ●
● ●●● ● ●●●●
● ●●●● ●●●●
● ●●●● ●●●●
● ●● ●● ●● ●●
●● ●●● ●●●●
●●●● ● ●●●●
● ●● ●● ●●●●
● ●● ●● ●●●●
0 2 4 6 8 10
1
1
1
1
1 RSV●●● ●● ●●●●
●●●● ● ●●●●
●●●● ● ●●●●
●●● ●● ●●●●
●●●●● ●●●●
●●● ●● ●●●●
●● ●● ● ●●●●
●● ●● ● ●●●●
●●●● ● ●●●●
●●●● ● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
●●●●● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
● ● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
●● ● ●● ●●●●
● ●●●● ●● ●●
●● ●●● ●● ●●
●● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
● ● ●●● ●●●●
●● ●●● ●●●●
●● ●● ● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
● ●●● ● ● ●●●
●●●● ● ●● ●●
●● ●● ● ●●●●
●●●● ● ●●●●
● ●●● ● ●●●●
0 0.5 1 1.5 2
1
1
1
1
1 PACO●● ●● ●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●●●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●●●● ●● ●●●
●●●●●●●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●● ●
●● ● ●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●● ●● ●●●
●●●● ●● ●●●
●●●● ●●●●●
●●●● ●● ●●●
●●●● ●● ●●●
●●●●●● ●●●
●●●●●●●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●● ●●● ●●●
●●● ●●● ●●●
●●●●●● ●●●
●● ●● ●● ●● ●
●●●●●● ●●●
●●● ●●● ●●●
●● ●●●● ●●●
●● ●●●● ●●●
0.7 0.8 0.9 1
1
1
1
1
1 RIAV●●●●● ●●●●
● ●● ●●●●●●
●●● ●● ●● ●●
●●●●● ●●●●
● ●● ●● ●●●●
●● ●●●●●●●
●●● ●●●●●●
●●●● ●● ●●●
●● ●●● ● ●●●
●● ● ●●● ●●●
●●● ●●● ●●●
● ●● ●●● ●●●
●● ●●● ● ●●●
●●●●● ● ●●●
●● ● ●●● ●● ●
●● ●●●● ●●●
● ●● ●●● ●●●
●● ● ●● ●●● ●
●●●●● ● ●●●
●●● ●●●●● ●
● ●●● ●● ●●●
● ● ●●●● ● ●●
●●● ●● ●●●●
●●●●●●●●●
● ● ●●●● ●●●
●● ●●●● ●●●
● ●● ●●● ●●●
●● ●●●● ●●●
●●● ●●● ●●●
● ●●●●● ●● ●
●● ● ●●●●●●
●●● ●●● ●●●
●●●● ●● ●●●
●●● ●●● ●●●
●●● ●●● ●●●
●● ●●●● ●●●
● ●● ●●● ●●●
●●●●●●●● ●
●●●●●● ●●●
●● ● ●●● ●●●
0 1 2 3 41
1
1
1
1 TCOIAV●●●●● ●●●●
●●● ●● ●● ●●
●●●●● ●●●●
●●● ●● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●● ●●● ●● ●●
●● ● ●● ●● ●●
●●● ●● ●● ●●
●● ●●● ●● ●●
●●●●● ●●●●
●●●●● ●● ●●
●● ●● ● ●●●●
●●●●● ●●●●
●●●●● ●● ●●
●● ●●● ●●●●
●●● ●● ●● ●●
●●●● ● ●●●●
●● ●●● ●●●●
●●●●● ●● ●●
●●●●● ●●●●
●●● ●● ●●●●
●●● ●● ●●●●
●●●●●●●●●
●●● ●● ●●● ●
●●●●● ●●●●
●●● ●● ●● ●●
●●●● ● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●●●● ● ●●●●
●●● ● ● ●● ●●
●●●●● ●● ●●
●●●●● ●● ●●
●●●●● ●●●●
●●● ●● ●●● ●
●●● ● ● ●● ●●
●●● ●● ●●●●
●●●●● ●● ●●
0 0.25 0.5 0.75 1
1
1 1 1 1 1 1
K14 models RMIB-UGent (top=.11; bottom=.44)●●
optimal score jackknife 95% confidence interval white background: RMIB-UGent is in K14 green background: RMIB-UGent is not in K14, but better or not the worstyellow background: RMIB-UGent is not in K14 and the worst
Temperature
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 15 / 26
1
EA
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1 YEAR
●● ●●●● ●●●
● ●● ●●● ●●●
●●●●●●●● ●
●●●●●● ●●●
●● ● ●●● ●●●
0 1 2 3 4
1K14 models RMIB-UGent (top=.11; bottom=.44)●●
optimal score jackknife 95% confidence interval
white background: RMIB-UGent is in K14 green background: RMIB-UGent is not in K14, but better or not the worstyellow background: RMIB-UGent is not in K14 and the worst
1
1 RIAV
1
domain
1
1 season
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 15 / 26
111
domain
1
BI
11111
IP
11111
FR
11111
ME
11111
SC
11111
AL
11111
MD
11111
EA
111
1
1
1
1 season
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1
1 YEAR
1
1 DJF
1
1 MAM
1
1 JJA
1
1 SON
1 YEAR
11
1
1
1
1 BIAS [K]● ●●●●● ●● ●
● ●●● ●● ●●●
●●●● ●●●●●
● ●●●●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●● ● ●● ●●●
●●●● ●●●●●
●●●● ●● ●●●
● ●●● ●● ●●●
● ●●●●● ●● ●
● ●●● ●● ●●●
● ●● ● ● ●●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●●● ●●● ●
●●●● ●●●● ●
● ●● ● ● ●●●●
● ●●●●● ●● ●
● ●●● ●●●● ●
● ●●● ●● ●● ●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●● ●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
● ●●● ●● ●●●
●●●● ●● ●● ●
● ●●● ●● ●●●
●●●● ●●●●●
● ●●● ●● ●●●
●●●● ●● ●●●
−5 −2 0 2 4
1
1
1
1
1 95%−P [K]●●● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●● ●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●●●●● ●●● ●
●● ●●● ●●● ●
●● ●●● ●●● ●
●● ● ●● ●●●●
●● ●●● ●●●●
●●●●● ●●●●
●● ●●● ●● ●●
●● ●●● ●● ●●
●●● ●●●●●●
● ● ●●● ●●●●
●● ●●● ●●●●
● ●●●● ●●●●
●●●●● ●●●●
●● ● ●● ●● ●●
●● ●●● ●● ●●
●● ●●● ●● ● ●
●● ● ●● ●● ●●
●● ● ●● ●● ●●
● ● ●●● ●● ●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
● ● ●●● ●●● ●
● ●●● ● ●●●●
● ●●●● ●●●●
● ●●●● ●●●●
● ●● ●● ●● ●●
●● ●●● ●●●●
●●●● ● ●●●●
● ●● ●● ●●●●
● ●● ●● ●●●●
0 2 4 6 8 10
1
1
1
1
1 RSV●●● ●● ●●●●
●●●● ● ●●●●
●●●● ● ●●●●
●●● ●● ●●●●
●●●●● ●●●●
●●● ●● ●●●●
●● ●● ● ●●●●
●● ●● ● ●●●●
●●●● ● ●●●●
●●●● ● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
●●●●● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
● ● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
●● ● ●● ●●●●
● ●●●● ●● ●●
●● ●●● ●● ●●
●● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
● ● ●●● ●●●●
●● ●●● ●●●●
●● ●● ● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●● ● ●● ●●●●
●● ●●● ●●●●
●● ●●● ●●●●
●●● ●● ●●●●
●● ● ●● ●●●●
● ●●● ● ● ●●●
●●●● ● ●● ●●
●● ●● ● ●●●●
●●●● ● ●●●●
● ●●● ● ●●●●
0 0.5 1 1.5 2
1
1
1
1
1 PACO●● ●● ●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●●●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●●●● ●● ●●●
●●●●●●●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●● ●
●● ● ●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●●● ●● ●●●
●●●● ●● ●●●
●●●● ●●●●●
●●●● ●● ●●●
●●●● ●● ●●●
●●●●●● ●●●
●●●●●●●●●
●●●●●● ●●●
●●●●●● ●●●
●●●●●● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●●● ●●● ●●●
●●● ●●● ●●●
●●●●●● ●●●
●● ●● ●● ●● ●
●●●●●● ●●●
●●● ●●● ●●●
●● ●●●● ●●●
●● ●●●● ●●●
0.7 0.8 0.9 1
1
1
1
1
1 RIAV●●●●● ●●●●
● ●● ●●●●●●
●●● ●● ●● ●●
●●●●● ●●●●
● ●● ●● ●●●●
●● ●●●●●●●
●●● ●●●●●●
●●●● ●● ●●●
●● ●●● ● ●●●
●● ● ●●● ●●●
●●● ●●● ●●●
● ●● ●●● ●●●
●● ●●● ● ●●●
●●●●● ● ●●●
●● ● ●●● ●● ●
●● ●●●● ●●●
● ●● ●●● ●●●
●● ● ●● ●●● ●
●●●●● ● ●●●
●●● ●●●●● ●
● ●●● ●● ●●●
● ● ●●●● ● ●●
●●● ●● ●●●●
●●●●●●●●●
● ● ●●●● ●●●
●● ●●●● ●●●
● ●● ●●● ●●●
●● ●●●● ●●●
●●● ●●● ●●●
● ●●●●● ●● ●
●● ● ●●●●●●
●●● ●●● ●●●
●●●● ●● ●●●
●●● ●●● ●●●
●●● ●●● ●●●
●● ●●●● ●●●
● ●● ●●● ●●●
●●●●●●●● ●
●●●●●● ●●●
●● ● ●●● ●●●
0 1 2 3 4
1
1
1
1
1 TCOIAV●●●●● ●●●●
●●● ●● ●● ●●
●●●●● ●●●●
●●● ●● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●● ●●● ●● ●●
●● ● ●● ●● ●●
●●● ●● ●● ●●
●● ●●● ●● ●●
●●●●● ●●●●
●●●●● ●● ●●
●● ●● ● ●●●●
●●●●● ●●●●
●●●●● ●● ●●
●● ●●● ●●●●
●●● ●● ●● ●●
●●●● ● ●●●●
●● ●●● ●●●●
●●●●● ●● ●●
●●●●● ●●●●
●●● ●● ●●●●
●●● ●● ●●●●
●●●●●●●●●
●●● ●● ●●● ●
●●●●● ●●●●
●●● ●● ●● ●●
●●●● ● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●●●●● ●●●●
●●●● ● ●●●●
●●● ● ● ●● ●●
●●●●● ●● ●●
●●●●● ●● ●●
●●●●● ●●●●
●●● ●● ●●● ●
●●● ● ● ●● ●●
●●● ●● ●●●●
●●●●● ●● ●●
0 0.25 0.5 0.75 1
1
1 1 1 1 1 1
K14 models RMIB-UGent (top=.11; bottom=.44)●●
optimal score jackknife 95% confidence interval white background: RMIB-UGent is in K14 green background: RMIB-UGent is not in K14, but better or not the worstyellow background: RMIB-UGent is not in K14 and the worst
Temperature
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 15 / 26
domainBI
IP
FR
ME
SC
AL
MD
EA
seasonDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEARDJFMAMJJASONYEAR
BIAS [%]●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
● ● ●●●● ● ●●
●● ●●●● ● ●●
●● ●●●●● ●●
●● ●●●●● ●●
● ● ●●●● ●● ●
●● ●●●●● ●●
●● ●●●●● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
● ● ●●●● ● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●● ●● ●●●
●● ●●●● ●●●
●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●● ●● ●●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
● ● ●● ●● ● ●●
●● ●● ●● ● ●●
●● ●● ●● ●●●
●● ●● ●● ● ●●
●●●●●● ● ●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
●● ●●●●● ●●
●● ●●●● ● ●●
●●●●● ● ● ● ●
●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●● ●● ●●●
●● ●● ●● ● ●●
● ● ●●●● ● ●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
−60 0 60 120 180
95%−P [%]● ●●●●●● ●●
● ● ●● ●●● ●●
●● ●●●●● ●●
● ●●●●●● ●●
● ● ●●●●● ●●
●● ●● ●●● ●●
● ● ●● ●●● ●●
● ● ●●● ● ●● ●
●● ●● ●●●●●
● ● ●● ●●● ●●
● ● ●●●●● ●●
●● ●●●● ● ●●
●● ●●● ●● ●●
●● ●●●●● ●●
●● ●●●●● ●●
●● ●● ●● ●●●
●● ●●● ●●●●
●● ●●● ●● ●●
●● ●● ●●● ●●
●● ●●●●● ●●
●●●● ●● ●●●
●●●● ●● ●●●
●● ●● ●●● ●●
●●●● ●●●●●
●● ●● ●● ● ●●
● ● ●● ●●●●●
●● ●● ●●●●●
●● ●● ●● ● ● ●
●● ●● ●●● ●●
●● ●● ●●● ●●
●● ●● ●●● ●●
●● ●● ●●● ●●
●● ●●● ●● ● ●
● ● ●● ●●● ●●
● ● ●● ●● ● ●●
●● ●● ●● ●●●
●● ●● ●● ●●●
●● ●● ● ●● ●●
●● ●● ●●● ●●
●● ●● ●● ● ●●
0 100 200 300 400
RSV● ●●●●●● ●●
● ●●● ●●● ●●
●● ●●● ●● ●●
● ●●●●●● ●●
● ●●●●●● ●●
● ● ●● ●● ● ●●
●● ●● ●●● ●●
● ● ●●●● ● ● ●
● ●●● ●●● ●●
● ● ●● ●●● ●●
● ●●● ●●● ●●
●● ●● ●●●●●
● ● ●● ●● ●●●
●●●● ●●● ●●
● ● ●● ●●● ●●
● ● ●● ●●●●●
●● ●● ● ●●●●
●● ●●● ●●●●
● ● ●●●●● ●●
●● ●● ●●●●●
● ●●● ●●●●●
● ●● ● ●●●●●
●● ●● ●●● ●●
● ●●● ●●●●●
● ●●● ●●●●●
● ● ●● ●●●●●
●● ●● ●●●●●
●●●● ●●●●●
● ● ●● ●●●●●
●● ●● ●●●●●
● ● ●● ●●● ●●
●● ●● ●●● ●●
●● ●● ●●● ●●
● ● ●● ●●● ●●
● ● ●● ●●● ●●
●● ●● ●● ●●●
●● ●● ●●● ●●
●● ●● ●●●●●
●● ●● ●●●●●
●● ●● ●●●●●
0 1 2 3 4
PACO●● ●●●● ●●●
●● ●●●● ●●●
●● ● ●●● ●● ●
●● ●●●● ●● ●
●● ●●●● ●●●
●● ●●● ● ●●●
●● ●●●● ●●●
●●●●●● ●●●
●● ●●●● ● ●●
●●●●●● ●●●
●● ●●● ● ●● ●
●● ● ●●● ●●●
●●● ●●● ●●●
●● ● ●●● ●●●
●● ● ●●● ●● ●
●● ●●●● ●● ●
●● ● ●●● ●● ●
●●● ●●● ●● ●
●● ●●●● ●● ●
●● ●●●● ●● ●
●●●●●●●●●
●●● ●● ●●●●
●●● ●●●●●●
●●●●●●●●●
●●●●●●●●●
●● ●●●● ●●●
●● ●●●● ●●●
●●● ●●● ●●●
●● ● ●●● ●● ●
●● ●●●● ●●●
● ●● ●●●● ●●
●● ●●●● ●●●
●●● ●●● ● ●●
●●● ●●● ●●●
●●● ●●● ●●●
●●●●●●●●●
●●● ●●● ● ●●
●●● ●●● ●●●
●●● ●● ●● ● ●
●● ● ●●● ●●●
0 0.25 0.5 0.75 1
RIAV● ● ●●●● ● ●●
● ● ●●●● ● ●●
●● ●●●●● ●●
●● ●● ●● ● ●●
● ●●● ●●● ●●
●● ●●●●● ●●
● ● ●●●●● ●●
●●● ● ●● ●● ●
● ● ●●●● ● ● ●
●● ●● ●● ● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
●● ●●●● ● ●●
●●● ●●● ● ●●
●● ●●●● ● ●●
●● ●● ●● ●●●
● ●●● ●● ● ●●
●● ●●●● ● ●●
●●●●●● ● ●●
●● ●● ●● ● ●●
● ● ●● ●● ●●●
● ● ●● ●● ● ●●
●● ●● ●●● ●●
● ● ●●●● ●●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
●● ●●●●● ●●
●● ●● ●● ● ●●
●● ●● ●● ● ●●
● ● ●● ●● ● ●●
●●●● ●● ● ● ●
●● ●●●● ● ●●
●●●● ●● ● ●●
● ●●● ●● ●●●
● ● ●● ●● ● ●●
● ● ●● ●● ●●●
● ●●● ●● ● ●●
● ● ●● ●● ● ●●
0 1 2 3 4
TCOIAV●●●●●●●● ●
●●●●● ●●● ●
●●● ●● ●●●●
●●● ●●●●● ●
●●● ●●●●●●
●●●●●●●●●
●●●●● ●●● ●
●●● ●● ●● ●●
●●● ●● ●●● ●
●●● ●●●●●●
●●●●● ●●●●
●●●●● ●●●●
●● ●● ● ●● ●●
●●●●● ●●●●
●● ●● ●●●●●
●●●●● ●●●●
●●●●● ●●● ●
●● ●● ● ●●●●
●●● ● ● ●●●●
●● ●● ● ●●●●
●●● ●●●●●●
●●●●● ●●● ●
●●●● ● ●●●●
●●● ●● ●●●●
● ●● ●●●●● ●
●●●● ● ●● ●●
●●●●● ●●●●
●● ●● ● ●● ● ●
●●● ●● ●●● ●
●● ●●● ●●● ●
●●●●● ●●●●
●●●● ● ●● ●●
●●●●● ●● ● ●
●●● ●● ●●● ●
●●● ●● ●●●●
●●● ●● ●●●●
●●● ●● ●● ●●
●●● ● ● ●●●●
●●● ●● ●● ●●
●●● ● ● ●● ●●
0 0.25 0.5 0.75 1
K14 models RMIB-UGent (top=.11; bottom=.44)●●
optimal score jackknife 95% confidence interval white background: RMIB-UGent is in K14green background: RMIB-UGent is not in K14, but better or not the worstyellow background: RMIB-UGent is not in K14 and the worst
Precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 16 / 26
Temperature bias patterns
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 17 / 26
Precipitation bias patterns
ALARO-0
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 18 / 26
Conclusions
A state-of-the-art validation was performed of the ALARO-0evaluation run of RMIB-UGent, following standardized metrics.
ALARO-0 performs well, despite not being tuned for climate:cfr. white/green backgrounds
Temperature biases persist in Scandinavia / Eastern Europe(same spatial pattern as ARPEGE)
For precipitation, ALARO-0 often outperforms all other models!
Robustness test: all scores except RIAV and TCOIAV are robust
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 19 / 26
1 ALARO-0 climate runs at RMIB: status
2 Validation of ALARO-0 for climate
3 Climate projections
4 Subdaily precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 20 / 26
RCP 8.5 vs historical T2m (Uccle, Belgium)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 21 / 26
Change in temperature: RCP 8.5(2070-2100) vs historical (1976-2005)
Winter Summer
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 22 / 26
Change in precipitation: RCP 8.5(2070-2100) vs historical (1976-2005)
Winter Summer
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 23 / 26
1 ALARO-0 climate runs at RMIB: status
2 Validation of ALARO-0 for climate
3 Climate projections
4 Subdaily precipitation
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 24 / 26
Subdaily precipitation
OBS: “centennial” 10-minute precipitation observation series in Uccle
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 25 / 26
References1 De Troch, R., et al.: Multiscale performance of the ALARO-0
model for simulating extreme summer precipitation climatologyin Belgium, Journal of Climate, 26(22), 8895-8915,doi:10.1175/JCLI-D-12-00844.1, 2013.
2 Kotlarski, S., et al.: Regional climate modeling on Europeanscales: a joint standard evaluation of the EURO-CORDEX RCMensemble, Geosci. Model Dev., 7, 1297-1333,doi:10.5194/gmd-7-1297-2014, 2014.
3 Giot, O., et al.: Validation of the ALARO-0 model within theEURO-CORDEX framework, Geosci. Model Dev., 9, 1143-1152,doi:10.5194/gmd-9-1143-2016, 2016.
4 De Troch, R., The application of the ALARO-0 model forregional climate modeling in Belgium: extreme precipitation andunfavorable conditions for the dispersion of air pollutants underpresent and future climate conditions, PhD dissertation, 2016.
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 26 / 26
Backup slides
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 27 / 26
Climate scenarios
(CO2-equivalent vs time)
� Finished: RCP 8.52040-2100
� Ongoing: RCP 4.52040-2100
� Ongoing: RCP 2.62040-2100
� Planned: ∗2005-2040
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 28 / 26
Relative change in yearly precipitation:RCP 8.5 (2070-2100) vs hist (1976-2005)
O. Giot et al. (RMIB) ALARO-0 EURO-CORDEX climate runs 20160407 29 / 26