Content uploaded by Adriaan Perrels
Author content
All content in this area was uploaded by Adriaan Perrels on Nov 07, 2022
Content may be subject to copyright.
Project URCLIM is part of ERA4CS, an ERA-NET initiated by JPI Climate and is co-funded by the European Union (Grant 690462)
Deliverables 5.3 and 5.4 combined:Impact simulation for combined urban and
climate change scenarios
Responsible partner: Finnish Meteorological Institute (FMI)
Authors: Zenaida Chitu (Meteo Romania), Roxana Bojariu (Meteo Romania), Sorin Dascalu (Meteo
Romania), Madalina Gothard (Meteo Romania), Liliana Velea (Meteo Romania), Athanasios Votsis (FMI),
Carl Fortelius (FMI), Rafiq Hamdi (RMI), Valéry Masson (Météo France), Reija Ruuhela (FMI), Olli Saranko
(FMI), Pilvi Siljamo (FMI), Bert van Schaeybroek (RMI), Adriaan Perrels (FMI)
Due Date: 01.12.2019 – delivered 30.06.2021
PU
Public
X
PP
Restricted to other programme participants (including the Commission services)
RE
Restricted to a group specified by the consortium (including the Commission
services)
CO
Confidential, only for members of the consortium (including the Commission
services)
Contents
Deliverables 5.3 and 5.4 combined:Impact simulation for combined urban and climate change
scenarios ................................................................................................................................................. 1
DISSEMINATION LEVEL ..................................................................................................................... 1
1. Introduction ................................................................................................................................... 2
2. Belgium – Brussels ....................................................................................................................... 4
2.1 Urban climate change and intervention scenarios .................................................................. 4
2.2 Highlights of the results for focus variables ............................................................................ 6
2.3 Obstacles in implementation and quality assurance............................................................. 10
3. Finland – Helsinki metropolitan area ........................................................................................... 11
3.1 Calibration of soft-coupled SLEUTH and SURFEX models .................................................. 11
3.1.1 Introduction ...................................................................................................................... 11
3.1.2 SLEUTH ........................................................................................................................... 11
3.1.3 SURFEX .......................................................................................................................... 12
3.2 Urbanisation scenario for greater Helsinki ........................................................................... 13
3.3 Scenario for Local Climate Zones ........................................................................................ 14
3.4 Exploration of urban heat occurrence in current and future climate in current and future built-
up area ........................................................................................................................................... 17
3.5 Sensitivity of UTCI to localized morphological characteristics and implications for adaptation
measures ....................................................................................................................................... 20
2
3.6 Exploration of difficult traffic and pedestrian conditions in current and future climate in current
and future built-up area................................................................................................................... 22
3.7 Services based on the results ................................................................................................... 24
4. France – Paris ............................................................................................................................ 25
4.1 climate change impacts and effects of adaptation with respect to UHI / heat waves ................. 25
4.2 Interaction with stakeholders to evaluate the potential of the indicators for future climate ......... 27
5. Romania - Bucharest .................................................................................................................. 29
5.1 Urban climate change .......................................................................................................... 29
5.2 Highlights of the results for focus variables .......................................................................... 31
References ........................................................................................................................................ 36
Impact simulation for combined urban and climate change scenarios
The URCLIM research plan states for the entire Work Package 5 and for Task 5.2 in particular the
following:
“The objective of this task (in fact work package) is to illustrate the usefulness of an urban model platform
by producing several scenarios of city evolution in conjunction with a changing urban climate, and by
evaluating the consequences and adaptation options for a suite of societal impacts in collaboration with
stakeholders (city expansion, density, architectural evolution, urban greening, etc.)”
“Deliverable D5.3 Impact simulation for combined urban and climate change scenarios
For at least Toulouse and Helsinki Metropolitan areas coordinated simulations of SURFEX-TEB and the
urban expansion model SLEUTH will be generated (by Météo-France and FMI) for each selected climate
scenario – urban development scenario combination, without the inclusion of adaptation measures.
Furthermore, transferability from Helsinki and Toulouse to other cities (Brussels, Bucharest) is tested. In
addition to physical indicators produced in Task 5.2 this task also explores heat related health effects in
terms of number of affected persons per year, as well as real estate valuation effects owing to both quality
changes in the ambient environment and changed risk levels.
The technical features and prerequisites of coordinated model use will be reviewed, with the aim to write
guidelines so as to ease replication of the approach in other cities. The results will be assessed in terms
of effects of climate on the urban spatial and socio-economic development, whereas also feedback effects
of urban form and size on urban climatic conditions will be assessed.
The suite of indicators encompasses: intense precipitation; snowfall; snow cover; floods; heat wave
occurrence and severity; human thermal comfort; building energy demand; air quality; health relevant
climate indicators”
“Deliverable 5.4 Report on the scientific analysis of adaptation strategies.
Based on premeditated adaptation measures belonging to one or more of the urban development
scenarios (Task 5.1), the scenario combinations of Task 5.2 and 5.3 are rerun while including the
adaptation measures. The effects of selected policy interventions will be explored, notably in terms of
3
changing building density, amount of vegetation, and building's physical characteristics. Evaluation will
encompass (1) the local ambient conditions and the possible surpassing of threshold”
At the 5th project plenary meeting was decided to merge D5.3 and D5.4 as for some applications it
appeared to be difficult to separately present climate change effects, other scenario effects and
adaptation effects, and if not difficult the stepwise approach could be easily confusing when presented in
separate consecutive reports. This Deliverable (5.3) builds on the current climate simulations of D5.2 in
which approaches and results were presented for current (recent) climate and current city for selected
regions and phenomena. Here the results for future climate and – if feasible – for the future city are
first presented based on the same models and approaches as presented in D5.2, after which the effect
of potential adaptation measures is explored. The presentation of results encompasses:
● Highlights of the results for focus variables*, including guidance on interpretation, e.g. in relation
to critical thresholds and regarding its basis for comparison of climate change impact in adaptation
scenarios;
● Discussion of obstacles in implementation of simulations and quality assurance (data
availability/processing, resolutions, quality issues; ease of reproducing such results)
● Discussion of service readiness of the baseline output and fitness for linking with other types of
(urban) data
The focus variables per country are as follows:
● Belgium: UHI
● Finland: UHI, UTCI, street surface slipperiness
● France: UHI, PET/UTCI
● Romania: UTCI UHI
4
2.1 URBAN CLIMATE CHANGE A ND INTERVENTIO N SCENARIOS
Two types of model experiments were performed using the land-surface model SURFEX over Brussels:
one vegetation and one albedo experiment. Their aim is to investigate how adaptation measures for
Brussels could impact the urban climate in the future. In practice a change of albedo in urban areas can
be established by a change the building colors. The vegetation scenario, on the other hand is implemented
by the widespread introduction of new parks and plantation of trees. Since urban areas suffer from a wide
range of impacts during warm and especially heat wave periods, we focus mostly on the summer period
and perform a separate investigation for the impact during heat waves. Our experiments are very simple
in nature since they involve spatially homogeneous changes in land-use characteristics to values that may
be unrealistic to obtain in practice. Therefore, the experiments presented here should rather be interpreted
as sensitivity experiments i.e. we estimate the sensitivity of the UHI with respect to changes in the land-
use characteristics.
Climate data generation
In order to obtain the 1-km climate simulations over Brussels, the SURFEX model was run in off-line mode
using model output of the ALARO-0 model as atmospheric forcing data (Termonia et al, 2018a). More
specifically, the output of the 4-km simulations over Brussels was used that follows the IPCC scenario
RCP8.5 for greenhouse concentrations until 2100 (Termonia et al., 2018b). This simulation was obtained
by dynamically downscaling the ALARO-0 simulations over the European (EURO-CORDEX) domain, as
validated in Giot et al (2016). Comparable downscaling efforts for Brussels were performed and validated
in Hamdi et al. (2015) with the difference that here no daily restarts of the ALARO model are used and
SURFEX was not (inline) coupled to the (4-km) ALARO model in the configuration at hand. As shown in
Hamdi et al. (2014) the lack of this coupling results in an underestimation of the UHI effect.
The climate simulations span the time period 1976-2100 but for the analysis only the period 2010-2100
will be used to compare the impact of the different land-use scenarios. The area covered is a region of 30
km by 30 km over Brussels with a resolution of 1km and the land-use characteristics taken from the
ECOCLIMAP (see Figure 1).
Figure 1: Land use characteristics at 1-km scale as used for the climate runs within SURFEX in the
urban scenarios.
Land-use scenarios
5
Prior to introducing the land-use scenarios it is useful to discuss the land use characteristics over Brussels
as shown in Figure 1. The boundaries of the Brussels Capital Region (BCR) are given in Figure 1 and it
is seen that the dense urban region is concentrated in the very center. The sub-urban areas, on the other
hand, reach far beyond the BCR. It is also seen that the north-east part of Brussels has a large fraction of
impermeable surface. Whereas different land-use types are used within the SURFEX model, we introduce
here three new categories: dense urban, sub-urban and rural tiles. The first two correspond to the 17 dark-
red (“dense urban”) and 250 light-pink (“temperate sub-urban”) tiles in Figure 1, respectively. The rural
tiles comprise the last six tile types listed in the legend of Figure 1.
The scenarios presented further only target the urban and sub-urban tiles while all results are averages
over the concerned tiles. Their main characteristics (impermeable fraction, vegetation fraction and building
heights) are tabulated in Table 1. The first experiment involves a change in albedo in the urban and
suburban regions while the second vegetation scenario introduces a modified fraction of vegetation in the
dense urban environment only. Finally, note that these changes are introduced in the land-use
characteristics of the (last) downscaling step using SURFEX only.
Table 1: Land-use characteristics of the dense urban and sub-urban tiles.
Land-use feature
Dense Urban
tiles
Suburbain tiles
Impermeable
surfaces
90 %
50 %
Vegetation
10 %
50 %
Building height
30 m
10 m
Albedo scenario
The albedo scenario involves a change in albedo of streets, walls and roofs, that are used in the
impermeable surfaces for both the urban and suburban regions. The default or “Initial” albedo values are
8%, 25% and 15% for streets, walls and roofs, respectively. In Table 2 we tabulate the values used for the
three scenarios: Albedo min, Albedo av and Albedo max. In practice, the maximal values (around 80%)
are almost impossible to obtain for a large urban area but, as aforementioned, these extreme values are
chosen to probe the sensitivity of the UHI with respect to changes in the land-use characteristics.
Note that, although an increase in the albedo of the urban environment will generally reduce the urban
heat island effect, it may have detrimental effects on human comfort during day-time. Indeed upon
increase of the albedo a person walking outside will be subject to increased shortwave radiation that
strongly affects his/her comfort level. The investigation of such effects is beyond the scope of this
sensitivity study.
Table 2: The albedo in the SURFEX model for the street, walls and roofs for the different albedo
scenarios.
Albedo
experiment
streets
walls
Roofs
6
Initial
8 %
25 %
15 %
Albedo min
20 %
40 %
25 %
Albedo av
50 %
62 %
55 %
Albedo max
80 %
85 %
85 %
Vegetation scenario
While the vegetation fraction of the dense urban tiles is by default 10%, two vegetation scenarios are
proposed here that increase this. More specifically the FractVeg 0.3 and FractVeg 0.5 scenarios increase
the fraction to 30% and 50%, respectively as tabulated in Table 3. Apart from the expected reduction of
the UHI upon increase of the vegetation fraction, one can expect the improvement of human comfort due
to shading. Again, this will not be investigated here.
Table 3: The vegetation fraction of the dense urban tiles for the default configuration and two
vegetation scenarios.
Vegetation experiment
Vegetation fraction dense urban
tiles
Initial
10 %
FractVeg 0.3
30 %
FractVeg 0.5
50 %
2.2 HIGHLIGHT S OF THE RESULTS FOR FOCUS VARIABLES
Results of the reference runs
Figure 2 shows increase in the average yearly temperature for the urban, sur-urban and rural parts of
Brussels for the years 2010-2100 following the RCP 8.5 scenario. Temperatures in the dense urban
environment are systematically higher than those in the sub-urban areas which, in turn, systematically
exceed those in the rural areas. This clearly shows the existence in the model of the observed urban heat
island effect over Brussels (Hamdi et al., 2014).
7
Figure 2: Average yearly temperature for rural (grey), suburban (orange) and dense urban (blue)
locations over Brussels for the years 2010-2100 following the greenhouse gas scenario RCP8.5.
The summer temperature difference of the dense urban and sub-urban environment with the rural area is
shown in Figure 3 for three periods: 2010-2040, 2040-2070 and 2070-2100. It is clear that the differences
among the different periods are very small and similar findings recur when studying the scenarios. The
invariance of UHI with respect to the time period might, however, not be realistic and most probably arises
due to the absence of explicit urban-atmosphere coupling in the SURFEX forcing data (Hamdi et al., 2014).
There is, on the other hand a marked difference in the UHI between the dense urban and the sub-urban
region, especially at night. This is in line with the expected behavior and confirms earlier findings for
Brussels.
Figure 3: Diurnal cycle of the summer Urban Heat Island (UHI). The UHI is obtained by subtracting
the temperature of the dense urban (blue line) or sub-urban (orange line) environment with the
temperature of the rural environment.
8
For the heat wave in this work, we use the definition from the Belgian Public Health authorities. According
to this definition, a heat wave is a period of at least three consecutive days, with an average daily minimum
and maximum temperatures which exceed 18°C and 30°C, respectively. The constraint on the minimal
temperature in this (health-related) definition stems from the fact that people suffer most from heat waves
when there are high night-time temperatures.
In line with the overall increase in temperature until the end of the century as shown in Figure 2, there is
also a stark increase in the number of heat waves, especially in the period 2070-2100. This is shown in
Figure 4 for the dense urban, sub-urban and rural areas.
Figure 4: The number of heat waves for the three land-use types and for the periods 2010-2040,
2040-2070 and 2070-2100 following the RCP8.5 scenario.
Results for the Albedo scenarios
Figure 5 shows the impact of the albedo increase of the dense urban and sub-urban environment on the
diurnal cycle of the summer UHI for the urban and sub-urban environment for the period 2010-2040. Note
that results for the periods 2040-2070 and 2070-2100 were practically identical. As expected, the largest
impact of the albedo increase concerns the day-time UHI as a consequence of the increased incoming
radiation. While the maximal UHI reduction in the dense urban environment is 0.46°C (dashed orange
line), the reduction in the sub-urban areas is 0.55°C. The difference can be attributed to the dense urban
environment where, compared with the sub-urban areas, 1) there is an enhanced heat trapping in the
urban canopy and 2) there is more conversion to sensible as opposed to latent heat of the (remaining)
incoming radiation.
9
Figure 5: Diurnal cycle of the summer Urban Heat Island (UHI) for different albedo scenarios for
the period 2010-2040. Note that the results obtained for the periods 2040-2070 and 2070-2100 are
almost identical.
Results for the Vegetation scenario
While a change in albedo mostly affects the day-time UHI, the inverse is true for the increase in vegetation
fraction. As seen in Figure 6, there is a strong reduction in the night-time UHI of 0.47°C following the
VegFraction 0.5 scenario and only a slight UHI increase during day time.
Figure 6: Diurnal cycle of the summer Urban Heat Island (UHI) in the dense urban environment for
different Vegetation scenarios for the period 2010-2040. Note that the results obtained for the
periods 2040-2070 and 2070-2100 are almost identical.
Scenario impacts on the heat waves
While, at least in the current model setup, the UHI is not affected by the time period considered, the
background temperatures will be steadily increasing with time following the RCP8.5 scenario and reach
3°C to 4°C at the end of the century. Therefore, a local reduction of temperature with 0.5°C will become a
minor contribution. This effect can be quantified by the use of the fraction of avoided heat waves upon
10
implementation of a certain adaptation measure. These are shown in Figure 7 for the Albedo scenarios
(left) and the vegetation scenarios (right).
While the fraction of avoided heat waves can be up to 35% for the period 2010-2040 and even up to
45% for 2040-2070 (both upon following the Albedo max scenario), there seems to be a strong decline
for the period 2070-2100, except for the Albedo max scenario in the sub-urban environment. The effect
of the vegetation adaptation on the (fractional) heat wave reduction becomes negligible for the period
2070-2100. Note, however, that, apart from the reduction in the heat wave number, all scenarios
systematically reduce the heat wave durationa and intensities.
Figure 7: Fraction of avoided heat waves (%) different time periods and for different albedo
scenarios (left) and vegetation scenarios (right), all following the RCP8.5.
2.3 OBSTACLES IN IMPLEMENTATION AND QUALITY ASSURANCE
As expected, an albedo increase strongly reduces the day-time UHI while an increase of the vegetation
fraction reduces the night-time UHI. Both scenarios had a maximal effect of reducing the UHI by 0.5°C.
The vegetation scenario can therefore be considered to have the largest impact on health as it allows
people in an urban environment to cool down at night. Moreover, the day-time outdoor human comfort is
generally improved by shading when trees are introduced while it is deteriorated by the increase of albedo.
These human comfort effects were not taken into account in this study.
The adaptation scenarios have a strong and systematic effect on the reduction of heat wave number,
length and intensity with respect to the case where no adaptation is introduced. However, following the
RCP8.5, the background temperature will also tend to strongly increase the number, length and intensity
of the heat waves. Therefore the fraction of avoided heat waves will strongly be reduced by the end of
the century (except for the strongest albedo change in sub-urban areas)
.
11
3.1 CALIBRATION OF SOFT -COUPLED SLEUTH AND SURFEX MODELS
3.1.1 Introduction
In this chapter we present the implementation of our approach for the Finnish Capital Region (see
Deliverables 3.x for the characteristics and current climate of the region). Figure 3.1 provides a general
overview of the modelling steps, which are described in more detail in the subsequent sections.
Figure 3 1: Main steps of the coupled urban climate and urbanization modelling.
We define soft-coupling of the SLEUTH and SURFEX models as the execution of input-output chains,
where output of one model is used as input to the other model (with pre-processing when necessary)
without modification of the source code of either model. This implies that, while the SURFEX and
SLEUTH models are prepared and/or calibrated to reproduce, respectively, urban climate and
urbanization dynamics independently of each other, the main point of attention is to prepare and/or
calibrate them with as many common data layers as possible, so that their scenarios have common
denominators.
3.1.2 SLEUTH
The SLEUTH-3r (Jantz et al. 2010) cellular automaton model is utilized for the forecasting of two-
dimensional urbanization dynamics of the Finnish Capital region. The model was calibrated and
validated to reproduce urbanization at a horizontal spatial resolution of 50 × 50 meters with historical
data of the transport network, topography, zoning, and urban growth. Figure 3.2 shows the data used
in the calibration. SLEUTH reproduces four spatial growth behaviors (diffusive, new spreading center,
edge, and road-influenced growth) by using five underlying growth coefficients that drive those
behaviors (diffusion, breed, spread, slope resistance, and road gravity). The aim of calibration is to
identify the specific parameter mix for the five growth coefficients that reproduces observed behavior
as closely as possible. For this calibration, the corresponding mix was determined as [diffusion: 1,
breed: 29, spread: 56, slope resistance: 42, and road gravity: 61], which indicates that urbanization in
Model
calibrations
Independent calibrations of urban climate (SURFEX) and urbanization (SLEUTH) models.
Common geographical domain (Helsinki Capital Region).
Common Local Climate Zone (LCZ) layers using the 2018 ECOCLIMAP-SG dataset.
Urbanization
simulation
Forecast of urban built-up growth till year 2040, based on historical data since year 2000.
Forecasts at annual timesteps and at 50 meters spatial resolution.
Local Cimate
Zone scenarios
Selection of cells that are forecasted to be built-up by year 2035 with ≥ 0.95 probability.
Conversion into two distinct LCZ scenarios for year 2035 at SURFEX spatial resolution.
Urban climate
scenarios
Forecast of urban meteorological parameters with SURFEX.
Use of SLEUTH-generated LCZ layers.
Focus on thermal environment, winter road conditions, pedestrian slipperiness.
12
Finnish Capital Region is realized mainly as road-oriented growth with expansion of existing urban
areas and filling-in of unbuilt areas. More details can be found in Votsis (2017).
Figure 3.2: Calibration data for the SLEUTH implementation in Finnish Capital Region.
3.1.3 SURFEX
The procedure for creating localized urban climate scenarios with the aid of the surface-atmosphere
interaction model SURFEX is described in Saranko et al. (2020) and in the deliverable 5.2 of the
present project. The recent past climate is mimicked by an artificially constructed reference year
(TRY2012, Kalamees et al., 2012). Following the EN ISO 15927-4 standard, the year has been
constructed out of 12 historical months selected from the period 1980–2009 in such way that the
monthly cumulative frequency distributions of daily mean air temperature, relative humidity, solar
radiation and wind speed were as close as possible to their respective climatological, i.e., 30-year
average, cumulative frequency distributions. As a modification to the standard, in the selection
procedure the four meteorological variables were weighted unequally, in order to take into account their
importance for building energy demand in Finland.
Gridded time series of meteorological variables are needed as prescribed forcing data to execute
SURFEX in stand-alone mode. The required variables included air temperature, relative humidity, wind
speed, rain, snowfall, as well as downwelling solar and thermal radiation. Data on a resolution of 2.5 km
were generated for the reference-year by the HARMONIE-AROME high-resolution numerical weather
prediction system (Bengtsson et al., 2017). Short-range hind-casts were produced four times for each
day of the respective months of the test-year, and hourly time series were extracted for the SURFEX
domain.
13
As detailed in Saranko et al. (2020), projections of climate change were derived from output of global
climate models (GCMs) that participated in the latest phase of the Coupled Model Intercomparison
Project, CMIP5 (Taylor et al., 2012). The CMIP5 global climate models were run under the
Representative Concentration Pathway (RCP) scenario RCP8.5 for global greenhouse gases (GHGs)
and aerosols (Taylor et al., 2012; van Vuuren et al., 2011). The CMIP climate change projections were
used to modify the hourly meteorological data so as to represent future climate conditions. Our method
falls into the category of morphing, time series adjustment or delta change methods (Belcher et al.,
2005; Räisänen and Räty, 2013). T
In a set of simulations, SURFEX was executed subject to forcing from present and future climates,
accounting for changes in the urban properties as obtained from SLEUTH. The sensitivity of urban
thermal climate to changes in climate and urban morphology were analysed based on output from
SURFEX. Wintertime conditions for road traffic were further analysed based on results given by the
RoadSurf model (Kangas et al., 2015) when coupled to output from SURFEX. An overview of the
simulations is given in Figure 3.3.
Figure 3.3: Main steps in producing the simulations by SURFEX and RoadSurf.
3.2 URBANISATION SCENARIO FOR GREATER HELSI NKI
Following the calibration process overviewed in Section 3.1, the model was asked to forecast
urbanization until year 2040 at annual timesteps and at 50 × 50 meters spatial resolution. This forecast
can be regarded as a baseline urbanization scenario in the sense that it reproduces recent (2000
onwards) spatial growth patterns without introducing any shifts in policy or ad-hoc localized policies. It
should be noted that the scenario does account for standard urban policy approaches, since it has
been calibrated to recognize and reproduce their year-to-year impacts on urbanization, albeit in an
14
aggregate manner. What the scenario does not include is radical departures from the recent planning
paradigm in the Finnish Capital Region. Figure 3.4 shows the expected urbanization at 0.99 probability
by 2040, following the calibrated growth parameters discussed in 3.1.
Figure 3.4: Initialization (left) and cumulative urbanization in 2040 (right) in Finnish Capital Region.
From this urbanization forecast, year 2035 was extracted, retaining only the urban pixels with
urbanization probability ≥ 0.95, that is, a more relaxed assumption than shown in Figure 3.4. Figure
3.5 shows the raw forecast (left) with the various probabilities of urbanization alongside the filtered
urban pixels (right) for year 2035.
Figure 3.5: Raw forecast (left) and filtered urban pixels (right, in grey) for year 2035.
3.3 SCENARIO FOR LOCA L CLIMATE ZONES
The coupling of SURFEX and SLEUTH is in practice achieved by the use of common ECOCLIMAP-SG
layers. ECOCLIMAP (Champeuax et al. 2006) is a global land use/cover dataset that has been developed
to cater for the needs of urban climate models. The spatial resolution of the current version (ECOCLIMAP-
SG) is 300 300 meters. The dataset is used by SURFEX as a reference about the physiography of the
modeled urban area, that is, about parameters such as building height, density, water areas, and
vegetation types.
ECOCLIMAP-SG contains 33 land use/cover classes (Table 1). Of special interest to urban applications
are classes 24-33, called Local Climate Zones (LCZs). The LCZs aim to provide a typology of the built
environment built around three parameters, density, height of buildings, and green cover, with a latent role
of the use of buildings as well (Stewart and Oke, 2012). Figure 5 shows the 2018 ECOCLIMAP-SG for
parts of the Helsinki Capital Region with all the natural classes aggregated to one color in order to highlight
15
the LCZ morphology. Figure 5 also displays a stylized visual snapshot of the kind of built-up morphology
implied by each of the 10 LCZs.
Table 3.1: ECOCLIMAP-SG land use/cover classes; Local Climate Zones (24-33) in bold.
1. sea and oceans
12. boreal needleleaf
evergreen
23. flooded grassland
2. lakes
13. temperate needleleaf
evergreen
24. LCZ1 compact high-rise
3. rivers
14. boreal needleleaf
deciduous
25. LCZ2 compact midrise
4. bare land
15. shrubs
26. LCZ3 compact low-rise
5. bare rock
16. boreal grassland
27. LCZ4 open high-rise
6. permanent snow
17. temperate grassland
28. LCZ5 open midrise
7. boreal broadleaf deciduous
18. tropical grassland
29: LCZ6 open low-rise
8. temperate broadleaf
deciduous
19. winter C3 crops
30: LCZ7 lightweight low-rise
9. tropical broadleaf deciduous
20. summer C3 crops
31: LCZ8 large low-rise
10. temperate broadleaf
evergreen
21. C4 crops
32: LCZ9 sparsely built
11. tropical broadleaf evergreen
22. flooded trees
33: LCZ10 heavy industry
Figure 3.6: ECOCLIMAP-SG representation of parts of the Finnish Capital Region.
In the present implementation of SLEUTH, forecasts provide information about the urban or non-urban
state of every 50-meter pixel for every year till 2040, including the selected forecast for year 2035 (Figures
3-4 in the preceding sections). In other words, the forecasts are “two-dimensional” without indication about
16
the density, height, use, or green cover inside a given urbanized pixel. In order to provide this information,
which is necessary to construct a future LCZ map, we have adopted the following methodology. First, we
modify the 2018 ECOCLIMAP-SG dataset, by tagging as ‘urban’ all of its natural pixels that are forecasted
by SLEUTH as likely to be urbanized (probability ≥ 0.95). We subsequently assign an LCZ class to these
newly assigned urban pixels by applying elements of agglomeration theory and of spatial econometrics.
In particular, we construct a spatial connectivity matrix (spatial weights matrix) in order to derive
information from the near neighbors about the prevailing density and height. With this information at hand,
we then refer to the definitions of the various LCZs, in order to correspond the resulting density and height
conditions to a particular LCZ class. In other words, we assume that newly urbanized pixels will follow—
but with various degrees of flexibility—the built-environmental profile of the local agglomeration in which
they are located. Depending on the specific assumptions about the strength and geographical extent of
influence of the local agglomeration, different LCZ scenarios can be constructed that still comply with the
built-up characteristics of the study domain. Figure 3.7 provides an overview of the aforementioned
workflow.
Figure 3.7: Main steps in producing LCZ forecasts from SLEUTH forecasts.
We have applied the aforementioned workflow in the construction of two scenarios: “Suburban
densification 1” and “Suburban densification 2” (see Figures 3.8-3.9). The total number of urbanized pixels
is the same in the two scenarios, but the type of growth is different. Suburban densification 2 assumes a
more continuous/connected suburbanization than Suburban densification 1, realized as a larger growth of
open low-rise developments around suburbanization centers (nodes) and smaller growth in sparse
development elsewhere; the intensity of growth in these two classes is swapped in Suburban densification
1 (see Figure 8). The scenarios assume no densification in already built-up areas (which should be
addressed in future research). Lastly, no changes between natural areas are considered, except of natural
to urban. As noted, these suburbanization scenarios comply both with the type of urbanization forecasted
by SLEUTH’s underlying drivers of growth based on observed growth behavior in the study domain as
well as with the morphology of the local agglomeration in which the newly urbanized pixel is located.
Radical departures from observed growth behavior has not been modelled in the present study but should
be taken into account in the development of additional scenarios as there are such indications in the
Finnish Capital Region.
Urban pixel
assignment
RECLASSIFY ECOCLIMAP-SG 2018 natural pixels as URBAN, IF SLEUTH forecasts
indicates that they are likely to be urbanized (probability ≥ 0.95).
Spatial
weights
Operationalize local agglomeration by constructing spatial connectivity matrices.
Parameteriza
tion
Calculate prevailing HEIGHT and DENSITY parameters for local agglomeration.
LCZ
assignment
Assign LCZ class [1-10] based on HEIGHT and DENSITY profile of each pixel.
17
Figure 3.8: The two LCZ scenarios.
Figure 3.9:: Differences between the two LCZ scenariosService readiness and fitness for linking with other
data
3.4 EXPLORATION OF URBAN HEAT OCCURRENCE IN CURRENT AND FUTURE
CLIMATE IN CURRENT AND FUT URE BUILT-UP AREA
The UHI effect may magnify heat-related mortality, especially during heatwaves. We modelled
temperature-related mortality in the city of Helsinki and in the surrounding Helsinki-Uusimaa hospital
district (excluding Helsinki) in 2000‒2018 using the distributed lag non-linear model. The results showed
that the heat-related mortality risk was substantially higher in Helsinki than in surrounding, more rural
areas. The mortality rates attributable to four intensive heatwaves (2003, 2010, 2014 and 2018) were
about 2.5 times higher in Helsinki than in the surrounding hospital district. Among the elderly, heat-
related risks were also higher in Helsinki, while cold-related risks were higher in the surrounding region
(Ruuhela et al., 2021). In that study we used FMI operational gridded temperature data set, thus, without
SURFEX output and therefore probable that the modelling underestimated the actual exposure to the
heat stress in Helsinki. In following paragraphs we describe how UHI effect modifies experienced
thermal stress based on SURFEX model runs.l Figure 3.10 shows the spatial distribution of moderate
heat stress (h) in June, July, and August of the test year. July is usually the hottest month of the year in
Finland. In June and especially in July, the coastal line is cooler than the rest of the area. In August, the
gap has already clearly leveled off. Heat stress is stronger than any other area in the densely built city
18
center, the densely populated trackside and some large suburbs also stand out. On the other hand, in low-
rise building areas, the thermal stress is lower.
Because UTCI (Universal Thermal Climate Index) and heat stress are the highest in July, we will focus
on July hereafter.
Figure 3.10: Cumulative moderate heat stress in the sun (h) in June, July and August in the test year.
The climate will become warmer in the coming decades. This raises the UTCI (Universal Thermal
Climate Index) value even if the urban structure does not change at all (Figure 3.11). Urban densification
raises UTCI even more in the Helsinki metropolitan area, but this is less than the impact of climate
change (Figure 3.11).
19
Figure 3.11: Average UTCI (℃) in July in present city and climate (upper left corner), present city and
2035 climate (lower left corner), densification 1 city scenario and 2035 climate (lower middle), and
densification 2 city scenario and 2035 climate (lower right corner)
At the practical level, this reflects in an increase in strong heat stress (Figure 3.12). Also in this case,
climate change has a greater role to play than urban structure. However, this underlines the importance
of urban planning to minimize the adverse effects of climate change.
20
Figure 3.12: Strong or worse heat stress (days) in July in present city and climate (upper left corner),
present city and 2035 climate (lower left corner), densification 1 city scenario and 2035 climate (lower
middle), and densification 2 city scenario and 2035 climate (lower right corner)
3.5 SENSITIVITY OF UTCI TO LOCALIZED MORPHOLOGICAL
CHARACTERISTICS AND IM PLI CATIONS FOR ADAPTATION MEASURES
Figure 3.13 shows the fraction of buildings in the Helsinki metropolitan area, the height of the buildings,
the fraction of gardens and parks, and the heat flux of industry. The fraction and height of buildings
correlate with each other as well as they negatively correlate with fraction of gardens. In the same dense
urban areas, there is a higher UTCI and heat stress in the Figures 3.11 and 3.12.
Figure 3.13: Fraction of buildings (a), building height (b), fraction of gardens and parks c) and industry
heat flux (d) in Helsinki metropolitan area.
Figure 3.14 shows the effect of building height on moderate and strong heat stress in July. As the height
of the buildings increases, so does the heat stress. When buildings are higher than about 20 m, heat
stress begins to decrease or no more grows. This same phenomenon is evident in both sun and shade
and in both moderate and strong heat stress. We think this is because higher buildings overshadow
more. However, it should be noted that only a small part of the buildings in the Helsinki metropolitan
area are more than 20 m high (Figure 3.13 b).
21
Figure 3.14: Building height (m) vs. moderate heat stress (a) or strong heat stress (b). Red in sun and
blue in shade.
Parks and green areas are known to have a cooling effect during summer. This can also be seen in the
model we used, but only with quite a large (0.5-0.6) fraction of gardens and parks (Figure 3.15).
Comparing the heat stress and the fraction of gardens, it is noticed that the scatter is very large. This is
due to the fact that there are green areas in very different types of environments: by the sea, in detached
house areas, in suburbs and even in the city center, and in larger park areas. Effect6 of large green
areas also seems to be spread over a wide area. . On the other hand, the Surfex-TEB model does not
take into account the effect of individual trees on heat stress, but they nevertheless have a great local
significance.
Figure 3.15: Fraction of gardens vs. moderate heat stress (a) or strong heat stress (b). Red in sun and
blue in shade.
22
3.6 EXPLORATION OF DI FFICULT TRAFFIC A ND PEDESTRIAN CONDITIONS IN
CURRENT AND FUT URE CLIMATE IN CURRENT AND FUTURE BUILT -UP AREA
The fractions of different surface conditions in the Helsinki metropolitan area in the test year January
predicted by FMI’s RoadSurf model are presented in the Figures 3.16 and 3.17. In the Figure 3.16, the
influence of changing climate is visible, as the scenarios of present climate and 2035 climate with
unmodified physiography are shown side by side. The number of situations related to snow and ice
(‘frost’, ‘partly icy’, ‘icy’ and ‘dry snow’) is decreasing, while wet-related conditions (‘wet’, ‘damp’) and
‘dry’ conditions are increasing. The already scarce ‘wet snow’ conditions decrease slightly according to
the simulations.
The impact of urban densification is shown in Figure 3.17, and it can be seen that it is negligible
compared to the effect of climate change. The impact is most visible in ‘dry’ and ‘frost’ cases, as the
number of ‘dry’ cases is slightly decreased in the lighter densification scenario while the number of ‘frost’
cases is slightly increased.
Figure 3.16: January surface conditions in present climate (a) and 2035 climate (b). The city
physiography in these scenarios is the present city.
23
Figure 3.17: January surface conditions of 2035 in densification 1 city scenario (a), and densification 2
city scenario (b).
Figure 3.18: The number of hours that difficult or very difficult traffic conditions occur in January in the
Helsinki metropolitan area. The maps are organized by the climate (upper row: present climate; lower
row: 2035 climate) and by city physiography (from left to right: present city, densification 1 city scenario
and densification 2 city scenario).
The Figure 3.18 shows difficult and very difficult traffic conditions in January in different climates and
different city scenarios.The areas with large low-rise buildings and coastal areas tend to have fewer
cases of difficult traffic conditions. The effect of city densification is quite subtle, but it can be seen it the
24
areas where the city is growing. In these areas, it can be seen that according to the simulations, the
densification 2 scenario has slightly less cases of difficult traffic conditions.
Figure 3.19 The number of hours that any slippery conditions occur in January in the Helsinki
metropolitan area. These conditions contain the slipperiness classes used by RoadSurf: slipperiness,
slipperiness caused by ploughing or packing, water on ice, and snow on ice. The maps are organized by
the climate (upper row: present climate; lower row: 2035 climate) and by city physiography (from left to
right: present city, densification 1 city scenario and densification 2 city scenario).
Similar to the number of difficult traffic conditions, the number of difficult pedestrian conditions decrease
with the climate change according to these simulations, as seen in Figure 3.19. The physiography
seems to have a significant effect on pedestrian conditions and the effect of the sea is visible, but the
differences between the two densification scenarios are again quite subtle.
3.7 SERVICES BASED ON THE RESULT S
Results for temperature, heat stress, and slipperiness in the greater Helsinki area presented above, will
be made available in an open multipurpose map application of the Helsinki Metropolitan Environmental
Service (HSY; https://www.hsy.fi/en/air-quality-and-climate/geographic-information/open-map-service/).
Planned to be functional in September 2021.
25
4.1 CLIMATE CHANGE I MPACTS AND EFFECTS OF ADAPTATION WIT H
RESPECT TO UHI / HEAT WAVES
A statistical–dynamical downscaling methodology is developed to quantify the UHI of the city of Paris
(France), based on a Local Weather Types (LWTs) classification combined with short-term high-
resolution (1-km) urban climate simulations. The daily near-surface temperature amplitude, specific
humidity, precipitation, wind speed and direction simulated by the RCMs are used for the LWTs
attribution. The LWTs time series is associated to randomly selected days simulated with the mesoscale
atmospheric model Meso-NH coupled to the urban canopy model Town Energy Balance to calculate the
UHI corresponding to the successive LWTs (Le Roy et al 2021).
The downscaling methodology is applied to the EURO-CORDEX ensemble driven by the ERA-Interim
reanalysis, and evaluated for the 2000–2008 period against station observations and a 2.5-km
reanalysis. The short-term dynamical simulations slightly underestimate and overestimate near-surface
minimum and maximum air temperature respectively, but capture the UHI intensity with biases in the
order of a tenth of a degree. RCMs show significant differences in the variables used for the LWTs
attribution, but the seasonal LWT frequencies are captured. Consequently, the reconstructed
temperature fields maintain the small biases of the Meso-NH simulations and the statistical–dynamical
downscaling greatly improves the UHI compared to the raw data of RCMs.
26
Methodology applied to evaluate impacts of climate change on Paris agglomeration (Le Roy et al 2021)
QQ plots of the 12 EURO-CORDEX RCMs against the AROME reanalysis for Minimum daily
temperature (TN), Maximum daily temperature (TX), dT and precipitations (RR) in winter (DJF) and
summer (JJA). The grey area represents an accuracy range of 2 K for TN and TX and 1 K for dT (Le
Roy et al 2021)
The analysis of these simulations done in the PhD of B. Le Roy showed that the UHI will be more
sensitive to future climate change during the day. This will be due not to changes in the behaviour of the
climate in the city, but to changes in the rural areas, mainly due to an evolution of the soil moisture.
Various impacts were evaluated. The heat stress is evaluated by the UTCI index. In summer, during the
day, the city center of Paris is relatively more comfortable than the near suburbs. This can be explained
by the typical urban form in Paris, very dense and hence with a lot of shadowing, limiting the increase of
the air temperature. In the future, an increase of 30 min are expected by the middle of the Century, and
more than 4h of thermal uncomfort by the end of the century. Indicators on energy consumption were
also produced, with a decrease of 40% of domestic heating, but 6 times more of air-conditioning, without
even taking into account an increase in AC systems installation).
27
Hours of Heat stress (outside) per day in present and future summer conditions (PhD thesis of B. Le
Roy).
Evolution of energy consumption for domestic heating in dense and suburban areas (PhD thesis of B. Le
Roy).
4.2 INTERA CTION W ITH STAKEHO LDERS TO EVALUATE THE POTENTIAL OF
THE INDICATORS FOR FUTURE CLIMATE
At the beginning of the project, it was planned to simulate the impact of urban scenarios to evaluate the
effectiveness of adaptation strategies. However, we realized that many results already found on this
subject, including those done at CNRM, were not fully exploited by the stakeholders. Therefore, we
prefered during the last part of URCLIM to focus on how to improve the knowledge transfer to the
stakeholders. The methodology has been to derive several forms for various indicators provided by the
study on Paris, and to present them during interview to stakeholders in the Paris region. This was done
during the internship of Julie André at CNRM. She interviewed 13 stakeholders, representing
municipalities, urban planning agencies, or state decentralized administrations.
The main conclusions from these exchanges with the stakeholders on how the climate indicators
produced by the downscalling method are perceived are that:
28
- the types of indicators produced are interesting, but they need to be more implemented in the
actor language and work methods. The outputs should not be presented in purely “scientific
form”, but included in their visualization tools. Storytelling can help the acculturation process.
- Heat stress and perceived temperature are very informative, and allow the stakeholders to center
their studies and on the human being. Energy consumption projections also interest the actors.
Still, crossing these indicators with socio-economic information and health data is needed for
further uses. This could be done by local agencies and actors. Meteorological institutes should
provide the downscaled climate data. One should note that such interdisciplinary analyses still
are a research domain.
- the horizontal resolution of 1km is fine enough to match their needs. However, the data should
be included in a geospatial tool, in order for them to be able to zoom on their territory.
- Temporal horizons (e.g. 2035 instead of a time period 2020-2049) are much more clearly
apprehended. Stakeholders are not much interested in 2100 horizons, because their temporal
horizons are in line with public action (typically no later than 2050).
- the local authorities, especially those for all individual cities and villages, rely on a small number
of urban actors to receive analysed climate information, typically the urban planning agencies (of
Paris and of its entire region) and the Parisian Climate Association. Transfer of the information
through these institutes can help the science transfer.
- The actors are often keen to have projections taking into account their urban evolution scenarios
(or even ongoing urbanism projects). Still, the key messages on urban and climate change (link
between UHI and city structure and expansion) are not known. Climate impacts taking into
account a changing climate but a stable city should be enough in a first stage of acculturation.
- Open questions remain on the need of, and the way to represent, the uncertainties for the actors.
29
5.1 URBAN CLIMATE CHANGE
Climate of Bucharest was investigated using observed meteorological data provided by two weather
stations and six urban weather stations equipped with air temperature and relative humidity sensors that
allowed analysis of urban heat island effect for actual climate. Changes of city climate were investigated
using EURO-CORDEX data that were bias corrected with observed data.
Urban heat island effect was identified for both summer and winter with an evident increase trend. The
annual mean difference (1961-2020) between dense urban (Filaret weather station) and sub-urban
(Baneasa weather station) environments during summer is 0.8 °C, while during the winter is 0.73 °C.
Figure 1a: UHI effect during summer
Figure 1b: UHI effect during winter
Investigation of hourly data (2015-2020) provided by urban stations (Cervantes, Paradisul Verde, Sf.
Nicolae, Scoala 30, Mihai Bravu and Teatrul Masca) showed for summer a mean difference between city
and sub-urban areas of about 1 °C during the day and 3 °C during the night. The UHI effect was
obtained by subtracting the temperature of the dense urban environment (urban stations) with the
temperature of the sub-urban environment (Baneasa weather station).
Figure 2: Diurnal cycle of the summer Urban Heat Island (UHI)
30
Figure 2: Diurnal cycle of the winter Urban Heat Island (UHI)
During winter the mean difference between city and sub-urban areas is less than 1°C during the day and
can reach 2.5°C during the night.
Evolution of city air temperature in the future was investigated by means of climate projections data
(EURO-CORDEX) that were bias corrected using observed data provided from Filaret weather station
located in the city centre. The modeled time series was bias corrected using the methodology developed
within Inter-Sectoral Impact Model Intercomparison Project (Hempel at al., 2013) that supposes that the
trend and variability of the original data are preserved by adjusting the cumulative distribution of the
simulated data to the observed one. Figure 3 presents the comparison between the distribution (left
panel) and cumulative distribution (right panel) of the raw and bias-corrected modeled temperature from
a specific RCM and the observed temperature series (Tobs).
Figure 3: Bias correction of the modelled temperature
31
Data provided by five RCMs (Regional Climate Model) for RCPs 4.5 and RCPs 8.5 were calibrated and
were represented graphically in the Figure 4. The red line and band represent the mean, respectively the
variation of air temperature from the five RCMs for RCP 8.5, while the green colour is used for RCP 4.5.
Figure 4: Temperature anomalies derived from calibrated EURO-CORDEX data for RCP 8.5 and RCP 4.5,
Bucharest
5.2 HIGHLIGHTS OF THE RESULTS FO R FOCUS VARI ABLES
Projections of Climate Change Impacts on Health, Bucharest Municipality
The effects of extreme temperatures on health in Bucharest were investigated using daily air
temperature data and daily deaths recorded between 1999 and 2019. The relationship between air
temperature data and the health outcome was analysed using the methodologies proposed by Gasparini
et al., 2015; and Vicendo-Cabrera et al., 2019 for both actual and future climate. The present
relationship between exposure and response was assessed using a time series regression model for
assessing the short-term associations. The non-linear and delayed effects of temperature on each death
are taken into account by using a distributed lag nonlinear model (DLNM) that includes a bi-dimensional
cross-basis function, a flexible natural cubic spline functions with three internal knots in the 10th, 75th and
90th percentiles of the temperature distribution, for modelling both exposure-response and lagged-
response dimensions, accounting for 21 days of lag.
The health impact projection was carried out based on a defined set of assumptions such as the
constant population and mortality rates with the aim to highlight the climate effect. The projected
mortality series was estimated as the average mortality for each daily observed death and reproduced
for the projection period of the modelled temperature series. The extrapolation of exposure-response
curves for future distribution of specific climate parameters such as air temperature beyond the observed
boundaries requires few hypotheses such as exposure-response association determined for observed
data will not change in the future and the extrapolation of the curve correspond to the risk over the
unobserved range that is related to uncertainty for future estimates.
The estimates of actual and projected impact were expressed both as ratio measure (relative risk) and
excess measures (attributable fraction (AF) and attributable number (AN)) in order to provide valuable
32
information for the planning and evaluation of public health interventions (Gasparini and Leone, 2014).
For each day of the series is computed the number of deaths attributed to a specific air temperature
based on estimated risk and the level of exposure, resulting the daily attributable numbers that are
aggregated for defined intervals of time in the future period.
The actual impact of temperature related mortality
The relationship between mortality and temperature (Fig. 5), estimated based on long term time series
(1999-2019), for all circulatory system diseases (CSDs), including all cardiovascular diseases (CVDs)
such as ischemic heart diseases, cerebrovascular diseases and hypertensive diseases, shows high
relative risk (RR) for elderly people (>75 years). The relative risk is also higher for deaths related to
hypertensive diseases and cerebrovascular diseases than ischemic heart diseases. Analysis on gender
suggests that women are more exposed than the male in Bucharest.
Figure 5: Overall cumulative exposure-response associations with temperature distributions for different
categories of circulatory system diseases, age and gender
The future impact of temperature related mortality
Risk estimates obtained over historical period were applied to future scenarios by performing a log-linear
extrapolation of the exposure-response association beyond the observed boundaries using a set of
33
assumptions: exposure-response association will not change in the future as a result of changes in
population structure and the extrapolation will be done only for unobserved range.
An example of extrapolation of temperature-total deaths related to all circulatory system diseases
association is presented in Figure 6 (adapted after Vicendo-Cabrera et al., 2019). In top panel: the
exposure-response curve presented as mortality RR across the temperature with 95% empirical
confidence interval represented with grey area, the dotted vertical line represents the minimum mortality
temperature which defines the two portions of the curve related to cold and heat; the dashed part of the
curve represents the extrapolation beyond the maxim temperature observed in 2010-2019 (dashed
vertical line); the mid panel is represented distribution of Tmod for the current (2010-2019 – with grey
area) and at the end of the century (2090-2099 -with green area), projected using a specific regional
climate model (MPI-M-MPI-ESM-LR_REMO2009) and scenario (RCP 8.5); bottom panel: the related
distribution of excess mortality, expressed as the fraction of deaths (%) attributed to nonoptimal
temperature.
Fig 6: Temperature and excess mortality for the current and future periods, Bucharest
The results of exposure-response associations were expressed in different manners such as relative
risk, attributable fraction or attributable numbers calculated as separated components for temperature
below (cold) and above (heat) the minimum mortality temperature (MMT) in order to provide useful
information for the planning and evaluation of public health interventions (Gasparini and Leone, 2014).
In the Table 1, are presented the attributable fraction of deaths for cold and hot temperature relative to
2010-2019 decade. In the future (2090-2099) the number of deaths during heat period will increase
(RCP 8.5) with 7.2% (95% eCI 4.0 – 10.1%) for hypertensive diseases and 7.0% (95% eCI 4.2 -9.6%)
for cerebrovascular diseases, relative to 2000-2019. In contrast, during cold period, the number of
34
deaths will decrease with 3.3% (95% eCI 5.5 - 1.2%) in case of hypertensive diseases and 3.1% (95%
eCI 4.7 – 1.7%) in case of ischemic heart diseases. Analysis on group age, with the actual structure of
population, suggests during the heat period that the most affected are elderly people: +84 years: 7.3
(95% eCI 4.5-9.9%) and 75-84 years: 7.1 (95% eCI 4.3-9.7%).
Table 1 Attributable fraction for heat and cold components in Bucharest Municipality relative to 2010-
2019:
RCP 8.5
Urban
AF rel in % (95% CI)
Cold
Heat
Deaths related to all circulatory
system diseases
-3.0 (-4.0- -2.0)
6.6 (4.2-9.0)
Deaths related to ischemic heart
diseases (I20-I25)
-3.1 (-4.7 - -1.7)
6.3 (3.6-8.8)
Deaths related to cerebrovascular
diseases (I60-I69)
-2.6 (-4.2- -1.1)
7.0 (4.2-9.6)
Deaths related to hypertensive
diseases (I10-15)
-3.3 (-5.5--1.2)
7.2 (4.0-10.1)
Deaths among male
-3.0 (-4.4 - -1.8)
7.1 (4.5-9.6)
Deaths among women
-2.9 (-4.2- -1.6)
5.9 (3.4-8.4)
Group age: < 65
-1.9 (-4.2-0.7)
3.9 (0.4-7.1)
Group age: 65-74
-2.6 (-4.3--0.8)
6.3 (3.4-8.9)
Group age: 75-84
-4.2(-5.7--2.8)
7.1 (4.3-9.7)
Group age: => 85
-2.2 (-3.9--0.6)
7.3 (4.5-9.9)
RCP 4.5
Urban
AF rel in % (95% CI)
Cold
Heat
Deaths related to all circulatory system
diseases
-1.1 (-1.6—0.6)
2.3 (1.0-4.3)
Deaths related to ischemic heart
diseases (I20-I25)
-1.2 (-1.9- -0.4)
2.2 (0.9-4.2)
Deaths related to cerebrovascular
diseases (I60-I69)
-1.0 (-1.6- -0.4)
2.5 (1.0-4.7)
Deaths related to hypertensive diseases
(I10-15)
-1.3 (-2.2—0.4)
2.5 (1-5)
35
Deaths among male
-1.2 (-1.7- --0.5)
2.5 (1.1-4.7)
Deaths among women
-1.1 (-1.7- -0.5)
2.0 (0.8-3.9)
Group age: < 65
-0.7 (-1.7- 0.3)
1.3(0.1-3.2)
Group age: 65-74
-0.9 (-1.9-0.2)
2.2 (0.9-4.4)
Group age: 75-84
-1.6 (-2.2--0.9)
2.5 (1.1-4.7)
Group age: => 85
-0.9 (-1.6--0.2)
2.6(1.1-4.9)
Graphically representation (Fig.6) of absolute attributable number of deaths for RCP 8.5 shows during
heat period an increase from 2488 deaths in 2029 to 11001 deaths in 2099, while during cold period
there is a decrease from 17920 deaths in 2029 to 13889 deaths in 2099. In contrast, for RCP 4.5, there
is a slight increase of absolute attributable numbers during heat period, from 3029 deaths to 5033
deaths in 2099 and a slight decrease during cold period from 17580 deaths to 16282 deaths in 2099
(Fig. 7).
Figure 6: Absolute numbers of deaths for RCP 8.5
36
Figure 7: Absolute numbers of deaths for RCP 4.5
Absolute numbers reveal for RCP 8.5 a decrease with 4031 deaths during cold and an increase with
8513 deaths during heat period from 2029 to 2099, resulting a positive balance of 4482 deaths. In
contrast, for RCP 4.5, it was identified a decrease with 2004 deaths during cold period and an increase
with 1298 deaths during heat period, resulting a negative balance estimated at 706 deaths. These
findings reveal the necessity to carefully plan urban development in Bucharest to face the public health
challenges raised by the climate change.
S.E. Belcher, J.N. Hacker, D.S. Powell: Constructing design weather data for future climates, Build.
Serv. Eng. Res. Technol., 26 (2005), pp. 49-61Bengtsson L., Andrae U. Aspelien T.-Batrak Y. Calvo J.
de Rooy W.-Gleeson E. Hansen-Sass B. Homleid M. Hortal M. Ivarsson K.-I. Lenderink G. Niemelä S.
Nielsen K.P. Onvlee J. Rontu L. Samuelsson P. Muñoz D.S. Subias A. Tijm S. Toll V. Yang X. and M.Ø.
Køltzow 2017. “The HARMONIE-AROME model configuration in the ALADIN-HIRLAM NWP system”. In:
Monthly Weather Review vol. 145, pp. 1919–1935
Champeaux JL, Masson V, Chauvin F (2006), ECOCLIMAP: a global database of land surface
parameters at 1 km resolution, Meteorological Applications 12(1): 29–32.
https://doi.org/10.1017/S1350482705001519
Demuzere M, Bechtel B, Middel A, Mills G (2019), Mapping Europe into local climate zones, PLOS ONE
| April 24, 2019: 1-27. https://doi.org/10.1371/journal.pone.0214474
Antonio Gasparrini, Yuming Guo, Francesco Sera, Ana Maria Vicedo-Cabrera, Veronika Huber, Shilu
Tong, Micheline de Sousa Zanotti Stagliorio Coelho, Paulo Hilario Nascimento Saldiva, Eric Lavigne,
Patricia Matus Correa, Nicolas Valdes Ortega, Haidong Kan, Samuel Osorio, Jan Kyselý, Aleš Urban,
Jouni J K Jaakkola, Niilo R I Ryti, Mathilde Pascal, Patrick G Goodman, Ariana Zeka, Paola Michelozzi,
Matteo Scortichini, Masahiro Hashizume, Yasushi Honda, Magali Hurtado-Diaz, Julio Cesar Cruz,
Xerxes Seposo, Ho Kim, Aurelio Tobias, Carmen Iñiguez, Bertil Forsberg, Daniel Oudin Åström, Martina
S Ragettli, Yue Leon Guo, Chang-fu Wu, Antonella Zanobetti, Joel Schwartz, Michelle L Bell, Tran Ngoc
37
Dang, Dung Do Van, Clare Heaviside, Sotiris Vardoulakis, Shakoor Hajat, Andy Haines, Ben Armstrong.
2017. Projections of temperature-related excess mortality under climate change scenarios, The Lancet
Planetary Health,Volume 1, Issue 9, Pages e360-e367, ISSN 2542-5196, https://doi.org/10.1016/S2542-
5196(17)30156-0.
Giot, O., Termonia, P., Degrauwe, D., De Troch, R., Caluwaerts, S., Smet, G., Berckmans, J., Deckmyn,
A., De Cruz, L., De Meutter, P., Duerinckx, A., Gerard, L., Hamdi, R., Van den Bergh, J., Van
Ginderachter, M., and Van Schaeybroeck, B.: Validation of the ALARO-0 model within the EURO-
CORDEX framework, Geosci. Model Dev., 9, 1143–1152, https://doi.org/10.5194/gmd-9-1143-2016,
2016
Hamdi, R., Van de Vyver, H.,De Troch, R., Termonia, P., 2014. Assessment of three dynamical urban
climate downscaling methods: Brussels' future urban heat island under an A1B emission scenario. Int. J.
Climatol. 34, 978–999.
Hamdi, R., Giot, O., De Troch, R., Deckmyn, A., Termonia, P., 2015. Future climate of Brussels and
Paris for the 2050s under the A1B scenario. Urban Climate 12, 160–182.
http://dx.doi.org/10.1016/j.uclim.2015.03.003
Jantz, C.A., Goetz, S.J., Donato, D. and Claggett, P., 2010. Designing and implementing a regional urban
modeling system using the SLEUTH cellular urban model. Computers, Environment and Urban Systems, 34(1),
pp.1-16.
T. Kalamees, K. Jylhä, H. Tietäväinen, J. Jokisalo, S. Ilomets, R. Hyvönen, S. Saku: Development of
weighting factors for climate variables for selecting the energy reference year according to the en ISO
15927-4 standard. Energy Build, 47 (2012), pp. 53-60
M. Kangas, M. Heikinheimo, and M. Hippi. 2015 RoadSurf - a modelling system for predicting road
weather and road surface conditions. Meteorological Applications 22(3), pp. 544–553.
Le Roy, B., Lemonsu, A., & Schoetter, R. (2021). A statistical–dynamical downscaling methodology for
the urban heat island applied to the EURO-CORDEX ensemble. Climate Dynamics, 1-22.
Ruuhela R, Votsis A, Kukkonen J, Jylhä K, Kankaanpää S, Perrels A, 2021. Temperature-Related
Mortality in Helsinki Compared to Its Surrounding Region Over Two Decades, with Special Emphasis on
Intensive Heatwaves. Atmosphere, 12, 46, doi: 10.3390/atmos12010046.
J. Räisänen, O. Räty: Projections of daily mean temperature variability in the future: cross-validation
tests with ENSEMBLES regional climate simulations. Clim. Dyn., 41 (2013), pp. 1553-1568
Saranko, Olli & Fortelius, Carl & Jylhä, Kirsti & Ruosteenoja, Kimmo & Brattich, Erika & Di Sabatino,
Silvana & Nurmi, Väinö. (2020). Impacts of town characteristics on the changing urban climate in
Vantaa. Science of The Total Environment. 727. 138471. 10.1016/j.scitotenv.2020.138471.
Stewart, I. D. & Oke, T. R. Local climate zones for urban temperature studies. Bulletin of the American
Meteorological Society 93, 1879–1900 (2012).
K.E. Taylor, R.J. Stouffer, G.A. Meehl: An overview of CMIP5 and the experiment design. Bull. Am.
Meteorol. Soc., 93 (2012), pp. 485-498
38
Termonia et al, 2018a: Termonia, P., Fischer, C., Bazile, E., Bouyssel, F., Brožková, R., Bénard, P.,
Bochenek, B., Degrauwe, D., Derková, M., El Khatib, R., Hamdi, R., Mašek, J., Pottier, P., Pristov, N.,
Seity, Y., Smolíková, P., Španiel, O., Tudor, M., Wang, Y., Wittmann, C., and Joly, A.: The ALADIN
System and its canonical model configurations AROME CY41T1 and ALARO CY40T1, Geosci. Model
Dev., 11, 257–281, https://doi.org/10.5194/gmd-11-257-2018, 2018
Termonia et al, 2018b: Termonia, P., Van Schaeybroeck, B., De Cruz, L., De Troch, R., Caluwaerts, S.,
Giot, O., Hamdi, R., Vannitsem, S., Duchêne, F., Willems, P., Tabari, H., Van Uytven, E.,
Hosseinzadehtalaei, P., Van Lipzig, N., Wouters, H., Vanden Broucke, S., van Ypersele, J.P., Marbaix,
P., Villanueva-Birriel, C., Fettweis, X., Wyard, C., Scholzen, C., Doutreloup, S., De Ridder, K., Gobin, A.,
Lauwaet, D., Stavrakou, T., Bauwens, M., Müller, J.F., Luyten, P., Ponsar, S., Van den Eynde, D.,
Pottiaux, E., 2018. The CORDEX.be initiative as a foundation for climate services in Belgium. Climate
Services, 11, 49–61, doi: 10.1016/j.cliser.2018.05.001.
Vicedo-Cabrera AM, Sera F, Gasparrini A. 2019. Hands-on Tutorial on a Modeling Framework for
Projections of Climate Change Impacts on Health. Epidemiology. 30(3):321-329. PMID: 30829832;
PMCID: PMC6533172. doi: 10.1097/EDE.0000000000000982
D.P. van Vuuren, J. Edmonds, M. Kainuma, K. Riahi, A. Thomson, K. Hibbard, G.C. Hurtt, T. Kram, V.
Krey, J.F. Lamarque, T. Masui, M. Meinshausen, N. Nakicenovic, S.J. Smith, S.K. Rose: The
representative concentration pathways: an overview. Clim. Chang., 109 (2011), pp. 5-31
Votsis, A. (2017), Utilizing a cellular automaton model to explore the influence of coastal flood
adaptation strategies on Helsinki's urbanization patterns, Computers, Environment and Urban Systems,
64: 344-355
39
