Course: Climate Responsive Design (3672) Semester: Spring, 2020

ASSIGNMENT - 1

Q.3 How can we separate Urban and rural climates? Define the major impacts of
urbanization on the climate of the cities and the neighboring sub-urban areas.

1. Urban Climate
Towns and cities are the most densely populated areas on Earth and will continue to be the
artificial landscapes most widely used by the greater part of the Earth’s population in the
future. In 2030 more than 60% of humans will live in cities. Changes in urban conditions
have often caused deterioration in environmental quality and may result in damage to the
health of city-dwellers. The differences between the climate of a city and the climate of its
surroundings are referred to as the “urban climate”.

Figure 1: Urban Climate

The most important features of urban climate include higher air and surface temperatures,
changes in radiation balances, lower humidity, and restricted atmospheric exchange that
causes accumulations of pollutants from a variety of sources. Although these changes mainly
affect local or regional conditions, persistent substances released into the atmosphere may
also affect larger areas or even the global climate.
Causes of Urban Climate
 The four main causes of urban climate, which result from different uses of built-up
areas, are: replacement of natural soil by sealed surfaces, mostly artificial and having
a strong 3-D structure;
 reduction of the surface area covered by vegetation;
CLIMATE RESPONSIVE DESIGN 1
  reduction of long-wave emission of the surface by street canyons
 release of gaseous, solid and liquid atmospheric pollutants, and waste heat.
These factors all have severe impacts on radiation and thermal properties such as
evapotranspiration, water storage, and atmospheric exchange in near-surface layers.
Properties of urban climates throughout the world are generally comparable. However,
the regional and local situation of an urban area, the infrastructure available and local
economic structures all modify the local anthropogenic climate (Wienert 2002). Here, we
will discuss only the most important characteristics of urban climate, and only for mid-
latitude settlements.
Thermal and Hydrological Properties of Urban Surfaces
The thermal behavior of sealed surfaces is largely determined by the density, heat capacity,
thermal conductivity, thermal diffusivity and thermal admittance coefficients of the materials
used. The behavior of sealed surfaces with respect to water drainage and seepage is highly
hetero-generous, because porosity and water-bearing properties may fluctuate severely as a
function of capillarity and pore volume. The large-scale use of impermeable materials for the
almost complete sealing of urban surfaces normally means that precipitation is drained
rapidly through underground sewers which are protected against evaporation, and that
exposed surfaces are only wetted for every short time. As a result of reduced evaporation,
more energy is available for long-wave emission, sensible heat flux and conduction to the
subsurface, and latent heat flux of evaporation is severely reduced.

2. Rural Climate
There is a lack of clear definition of what constitutes rural areas, and definitions that do exist
depend on definitions of the urban. Across the world, the importance of peri-urban areas and
new forms of rural-urban interactions are increasing (limited evidence, high agreement).
Rural areas, viewed as a dynamic, spatial category, remain important for assessing the
impacts of climate change and the prospects for adaptation.
In different regions, absolute rural populations have peaked or will peak in the next few
decades. The proportion of the rural population depending on agriculture is extremely varied
across regions but declining everywhere. Poverty rates in rural areas are higher than overall
poverty rates, but also falling more sharply, and the proportions of population in extreme
poverty accounted for by rural people are also falling in both cases except for sub-Saharan
Africa, where these rates are rising. Accelerating globalization, through migration, labor
linkages, regional and international trade, and new information and communication
technologies, is bringing about economic transformation in rural areas of both developing and
developed countries.
In developing countries, the levels and distribution of rural poverty are affected in complex
and interacting ways by processes of commercialization and diversification, food policies,
and policies on land tenure. In developed countries, there are important shifts toward multiple
uses of rural areas, especially leisure uses, and new rural policies based on the collaboration
of multiple stakeholders, the targeting of multiple sectors, and a change from subsidy-based
to investment-based policy.
There is low agreement on some of the key factors associated with vulnerability or resilience
in rural areas, including rainfed as opposed to irrigated agriculture, small-scale and family-
managed farms, and integration into world markets. There is high agreement on the
importance for resilience of access to land and natural resources, flexible local institutions

CLIMATE RESPONSIVE DESIGN 2

and knowledge and information and on the association of gender inequalities with
vulnerability. Specific livelihood niches such as pastoralism, mountain farming systems and
artisanal fisheries are vulnerable and at high risk of adverse impacts (high confidence), partly
owing to neglect, misunderstanding, or inappropriate policy toward them on the part of
governments.

Figure 2: Comparison Between Urban and Rural Effects

There is a need to understand how implementation of these policies will impact on rural
livelihoods. These secondary impacts, and trade-offs between mitigation and adaptation in
rural areas, have implications for governance, including the need to promote participation of
rural stakeholders.
Valuation of climate impacts needs to draw on both monetary and non-monetary indicators.
The valuation of non-marketed ecosystem services and the limitations of economic valuation
models that aggregate across multiple contexts pose challenges for valuing impacts in rural
areas (high confidence).
Public policies supporting decision making for adaptation exist in developed and,
increasingly, in developing countries, and there are also examples of private adaptations led
by individuals, companies, and non-governmental organizations (high confidence).
Constraints on adaptation come from lack of access to credit, land, water, technology,
markets, knowledge and information, and perceptions of the need to change; and are
particularly pronounced in developing countries (high confidence). Gender and institutions
affect access to adaptation options and the presence of barriers to adaptation (very high
confidence)

3. Impact of urbanization patterns on the local climate of a tropical city,

Singapore: An ensemble study
 Introduction

CLIMATE RESPONSIVE DESIGN 3

As one of the extreme examples of human beings' changes to the natural environment,
urbanization is continuously impacting our life in various aspects. Urbanization is not simply
a microscale to local‐scale effect, and it also demonstrates a regional‐scale signature [Y. Zhou
et al., 2015]. A notable result of microclimatic changes caused by anthropogenic alterations
to the Earth's surface is the urban heat island (UHI) effect [Landsberg, 1981], a phenomenon
that the urban area is hotter than its surrounding rural area. Studies have shown that urban
thermal conditions vary not only from those in rural surroundings but also within the urban
area due to intraurban differences in land use and surface characteristics. The footprint of
UHI effect can reach about 4 times of the urban size, with large spatiotemporal
heterogeneities [D. Zhou et al., 2015]. Urbanization also affects local winds and water cycles
through the modification of natural surfaces and atmospheric conditions and further changes
the local weather and climate system. While global‐scale climate forcing are superimposed
on the effects of the built environment on local transfers of heat and moisture, the impacts of
cities on weather and climate may extend to regional and global scales by changing
atmospheric composition, impacting components of the water cycle, and modifying the
carbon cycle and ecosystems [Shepherd, 2005]. It has been shown that the global forcing and
the forcing of the built environment are of similar order of impact, and some recent studies
[Georgescu et al., 2013, 2014] have explicitly made this important distinction between them.
Therefore, a rising urgency exists to incorporate the scale‐dependent built environment when
local and strategical deployed measures (e. g., green roofs and cool roofs) are implemented to
ameliorate the negative consequences on urban climate [Georgescu et al., 2015].

There are various types of UHI [Oke,

1976], depending on the temperature
used to define it. Here we focus
on the urban canopy layer UHI, which
measures the thermal environment below
the average building height
and with immediate impact on human
comfort on a daily basis. One major
effect of urbanization is the gen-
eral reduction of the surface albedo, the
ratio of outgoing shortwave radiation to
that of incident shortwave
There are various types of UHI [Oke, 1976], depending on the temperature used to define it.
Here we focus on the urban canopy layer UHI, which measures the thermal environment
CLIMATE RESPONSIVE DESIGN 4
below the average building height and with immediate impact on human comfort on a daily
basis. One major effect of urbanization is the general reduction of the surface albedo, the
ratio of outgoing shortwave radiation to that of incident shortwave radiation in a three‐
dimensional environment. Lower albedo is due in part to darker surface materials making up
the urban mosaic and also to the effects of trapping shortwave radiation by the vertical walls
and the urban, canyon‐like morphology [Heisler and Brazel, 2010]. The relative lack of
vegetation in urban areas also contributes to the warmer environment than the surrounding
rural areas. However, a recent report [Zhao et al., 2014] argued that instead of the reduction
in evaporative cooling in urban areas, it is the lower efficiency of convicting heat to the lower
atmosphere (which depends on local background climate) in North American cities that
contributes strongly to UHI. Anthropogenic heat (AH) is the energy released from human
sources such as vehicles, commercial and residential buildings, industry, power plants, and
human metabolism [Quah and Roth, 2012]. AH can greatly affect the urban environment by
directly changing the surface air temperature and by indirectly modifying urban boundary
layer structure, precipitation, and other conditions [e.g., Ichinose et al., 1999; Bornstein and
Lin, 2000; Lin et al., 2008; D. Li et al., 2013; Bohnenstengel et al., 2014].
Extreme precipitation events (drought and flooding) in urban areas are the major hazard that
motivates research work in urban‐induced precipitation. Urban‐enhanced precipitation events
in conjunction with the increased extent of impervious surface have resulted in heavy
runoff/urban flooding events [Shepherd et al., 2011]. Therefore, the possible implications of
urban‐induced precipitation on the design of urban drainage systems should be taken into
account [Burian et al., 2004].
Model Configuration and Experiment Design
 The Study Area
Singapore is an island state located between 1°09  N to 1°29  N and 103°36  E to 104°25 
E. Its climate is a typical wet equatorial type (Köppen classification: Af) [Essenwanger, 2001]
with uniformly high monthly mean temperature (26–27.7°C) and annual rainfall (≈2300 mm).
Singapore is located at the southern tip of the Malay Peninsula and is affected by monsoons.
The Northeast monsoon season from December to March, which is associated with the
highest monthly rainfall and weaker winds, and the Southwest monsoon season from June to
September, which corresponds to a relatively drier period, are separated by two brief inter-
monsoon or pre-monsoon periods [Chia and Foong, 1991]. There is no large and pronounced
topography. The highest natural point is Bukit Timah at 164 m above sea level, which is
exceeded in height by the tallest buildings in the central business district (CBD). This makes
the study almost free of topographic effects, and our focus will be solely on the urbanization
effect. There are ongoing land reclamation projects (mostly along the southern coastline),
which have increased the land area to the current 704 km2.
With a population reaching 5 million in 2011, Singapore has undergone dramatic
urbanization during the past 45 years. The built‐up area of Singapore has increased from 28%
in 1955 to 50% in 1998, while the farm area and forest have decreased correspondingly
[Chow and Roth, 2006]. Most of the urbanization is associated with the building of high‐rise
housing estates around the island, high‐rise offices in the CBD in the south and industrial
factories in the west.
 The WRF Model
Numerical simulations in this study were performed with the WRF model [Skamarock et
al., 2008] version 3.2.1 with Advanced Research WRF dynamics core developed by the
CLIMATE RESPONSIVE DESIGN 5
National Center for Atmospheric Research. The WRF model is a fully compressible,
nonhydrostatic model with a terrain‐following mass coordinate system and is designed to
simulate or predict regional weather and climate. The WRF model is applied to the 5 one‐way
nested domains shown in Figure 1a. The horizontal resolutions (grid sizes) of these five
domains are 24.3 km (76×76), 8.1 km (79 × 91), 2.7 km (112 × 112), 0.9 km (112 × 112), and
0.3 km (211 × 130). In the vertical direction, there are 38 full sigma levels from the surface to
50 hPa, with the lowest 14 levels resolving the planetary boundary layer (PBL).
Several physical parameterization schemes are employed in the WRF system to account for
different physical processes. The Noah land surface model provides surface sensible and
latent heat fluxes and surface skin temperature as lower boundary conditions to WRF. It can
be coupled to various urban canopy models (UCM) through the urban fraction parameter
(Furb) that represents the proportion of impervious surfaces in the WRF subgrid scale. The
UCM coupled to the WRF/Noah system in this study is the single‐layer urban canopy model
(SLUCM) developed by Kusaka and Kimura [2004]. This UCM has an intermediate level of
complexity among the available UCMs for WRF. It assumes infinitely long street canyons
parameterized to represent urban geometry but recognizes the three‐dimensional nature of
urban surfaces [Chen et al., 2011]. More details of the UCM parameters used in this study
can be found in X.‐X. Li et al. [2013].
Other physical parameterizations employed in this study are: rapid radiative transfer model
longwave radiation scheme, Dudhia shortwave radiation scheme, Mellor‐Yamada‐Janjić
turbulence kinetic energy PBL scheme, Monin‐Obukhov surface layer scheme, Goddard
microphysics scheme, and Kain‐Fritsch cumulus scheme (for d01 and d02 only).
The same urban parameters used in the CONTROL experiment [X.‐X. Li et al., 2013, Table
1] are used for all other experiments, with the exception of the peak AH. The diurnal profile
and peak value of AH used in the CONTROL experiment are documented in detail in X.‐X.
Li et al. [2013, Figure 3]. The same diurnal profile of AH is used in the above mentioned
experiments. In order to remove the unrealistic excessive direct warming that would have
occurred due to AH from converting all urban land use types to the commercial/industrial
land use type, we have to roughly preserve the same domain‐averaged 1 day integrated
sensible heat flux as the CONTROL experiment in domain d05. Therefore, with some
sensitivity tests of different peak AH, the peak AH values for the SURB, NURB, and HURB
experiment are modified from 113 W m−2 in the CONTROL experiment to 39.0, 97.5, and
58.0 W m−2, respectively. However, due to the important role played by AH in altering the
urban environment (as demonstrated in X.‐X. Li et al. [2013]), two more experiments using
the same peak AH value and diurnal profile as in the SURB experiment, i.e., NURBS and
HURBS, are added in order to compare more consistently the effect of AH.
Results and Discussions
In this section, the effect of urbanization patterns (as described in the previous section) on the
local climate of Singapore will be examined based on the WRF‐SLUCM simulation results.
Unless otherwise specified, the numerical results presented in this section are for domain d05.
 Urban Heat Island

The coupled WRF/UCM model's performance in reproducing the surface air temperature and
surface energy balance has been evaluated by X.‐X. Li et al. [2013]. The 2 m air temperature
and specific humidity calculated from the coupled WRF/UCM model were validated against
those from a sensor network operated across the Singapore island. A mean bias of −0.27°C
CLIMATE RESPONSIVE DESIGN 6
for temperature was achieved, while a worse agreement of specific humility (about 10%
underprediction) was reported. In addition, the surface energy balance components from the
model were compared favourably with the observed values from a flux tower located within
an urban neighbourhood [X.‐X. Li et al., 2013]. In this section, the model will be used to
investigate the UHI characteristics of Singapore under different urbanization scenarios.

 Calculation Method of UHI Intensity

The canopy‐layer UHI intensity due to urbanization can be calculated in two ways. The first
way is the traditional difference in the 2 m air temperature, T2, between urban and nearby
undeveloped rural areas (evergreen broad leaf forest in this case). The second is the
difference in the 2 m temperatures between the CONTROL and FOREST simulations, which
is termed as “urban increment.” This method largely removes impacts of clouds or
topography that might alter the surface temperatures [Bohnenstengel et al., 2011]. In this
section, the second method is adopted in order to facilitate the comparison between different
experiments. The calculated UHI intensity is further averaged over the same land use type
(commercial/industrial) in domain d05.
The local circulations (sea/land breezes) can play a significant role in modulating UHI
intensity in a coastal urban environment [Joseph et al., 2008; X.‐X. Li et al., 2013] and will
create an ambient temperature gradient in Singapore. To standardize the geographical
sampling and thus minimize the impact of local circulations when comparing UHI intensity
in different experiments, the UHI intensities of the HURB/HURBS experiments are split into
two parts, i.e., southern and northern part (hereafter denoted as HURB/HURBS‐south and
HURB/HURBS‐north, respectively) in accordance with the SURB and NURB experiments.
 Effect of Urbanization Patterns
The symmetric and asymmetric urbanization patterns in Singapore can affect the distribution
of UHI intensity. With the contiguous urban areas in the asymmetric urbanization
experiments (SURB and NURBS, Figure 3), the urban canopy will not only increase the total
amount of AH release but also make the released heat not easily removed by horizontal
advection as neighboring areas are as warm. As a result, the magnitudes of the UHI intensity
in the asymmetric urbanization settings are consistently higher than the corresponding part of
the symmetric urbanization experiment, HURBS (Figures 3a and 3c). A further check by
utilizing a paired, unequaled t test reveals that the above observation is statistically significant
during early morning (2:00 to 7:00 LT) with p value <0.01, but not significant during
daytime. A notable phenomenon in Figures 3a and 3b is that the UHI discrepancy between
asymmetric/symmetric urbanization experiments is larger during nighttime (late evening to
early morning) than during daytime. To elaborate this, it is necessary to introduce here
surface energy balance, which is the key to understanding the surface air temperature
evolution. The surface energy balance for urban area reads

CLIMATE RESPONSIVE DESIGN 7

 Figure 3: UHI intensity and energy balance change comparison between asymmetric and symmetric urbanization
patterns.

 Conclusion
As an extreme case of land cover/land use change, urbanization has strong impacts on local
climate, especially in such tropical coastal cities as Singapore. This paper examines the effect
of urbanization and urbanization patterns on the urban thermal environment and rainfall using
the coupled WRF‐urban canopy model in ensemble experiments. The urbanization pattern
changes are represented by several idealized scenarios: the asymmetric urbanization SURB
(the southern part of Singapore is urban while the northern part is forested), and NURB (the
northern part of Singapore is urban while the southern part is forested), and the symmetric
urbanization HURB (the whole Singapore is covered by alternating urban and forest land use
types. Additional scenarios, NURBS and HURBS, which use the same urbanization pattern
as in NURB and HURB but with the same anthropogenic heat (AH) profile as in SURB, are
also considered to evaluate the effect of AH.
The results from the 28 members of the ensemble for different urbanization scenarios show
that both urbanization patterns and AH have some impact on the urban heat island (UHI)
intensity, although the impact of urbanization patterns is rather weak. In the asymmetric
urbanization patterns (SURB and NURBS), the magnitude of UHI intensity is mildly higher
than that in the symmetric urbanization pattern (HURBS) due to the contiguous distribution
CLIMATE RESPONSIVE DESIGN 8
of urban areas in the former. On the other hand, the changes induced by altering
anthropogenic heat are much more significant. Higher AH will induce higher UHI intensity,
and the effect of AH is more prominent during the night-time than daytime, which is
consistent with the findings from X.‐X. Li et al. [2013]. The climate of urban areas and
downwind of urban areas seems to be sensitive to AH intensity, which can come about more
readily (e.g., by increasing the use of air conditioners at night) than major changes in
urbanization patterns. The local circulations (sea/land breezes) will cool the upwind area, and
air parcels travel downwind with the heat flux accumulated from upwind areas, making the
downwind area hotter.
The rainfall in tropical coastal cities can often be driven by local sea breezes. Sea breezes
bring moist air into Singapore during late morning and afternoon. Our study suggests that sea
breezes have stronger influence on rainfall than the urbanization pattern since the downwind
part (northern Singapore) always has more rainfall than the upwind part (southern Singapore)
even in experiments where the urbanization pattern is entirely reversed. The urbanization and
the associated AH can increase rainfall through increasing buoyancy by AH or decrease
rainfall through reducing evaporation by converting greenery to impervious surfaces. The
ultimate effect is determined by the relative strength of these two influences.
While it is natural to presume that the results of the symmetric urbanization experiment
always lie in between those of the asymmetric urbanization experiments, the UHI intensities
in both asymmetric urbanization experiments (SURB and NURBS) are higher than that from
HURBS. On the other hand, the rainfall amount from the HURBS experiment does indeed lie
between those from the SURB and NURBS experiments most of the time. This different
behaviour indicates that the effect of having uniform urbanization patterns is not always the
same and is, in fact, dependent on the dynamics governing the examined physical variable.
While evidence points strongly to a role for aerosols in urban precipitation modification, the
details of that role still remain highly uncertain [Shepherd, 2005]. Furthermore, the WRF
model itself cannot predict the aerosol distribution without coupling with other
photochemical models such as WRF‐Chem. In addition, our focus is the effect of changing
urbanization patterns, which is supposed to modify little, if any, the aerosol effect between
different urbanization patterns, and hence, the aerosol effect will not affect the results and
conclusions that are based on the difference caused by urbanization pattern change.
Therefore, we have left out the aerosol effect from urbanization in this study regardless of its
importance.
Reference:
1. The Urban Climate – Basic and Applied Aspects by Wilhelm Kuttler
2. https://www.ipcc.ch/site/assets/uploads/2018/02/WGIIAR5-Chap9_FINAL.pdf
3. Impact of urbanization patterns on the local climate of a tropical city, Singapore: An
ensemble study (Xian – Xiang Li, Tieh – Yong Koh, Iagabandhu Panda, Leslie K Norford)

CLIMATE RESPONSIVE DESIGN 9