Theoretical and Applied Climatology (2019) 136:971–984

https://doi.org/10.1007/s00704-018-2534-1

ORIGINAL PAPER

Green or blue spaces? Assessment of the effectiveness and costs

to mitigate the urban heat island in a Latin American city
Admir Créso Targino 1 & Guilherme Conor Coraiola 2 & Patricia Krecl 1

Received: 20 October 2017 / Accepted: 1 June 2018 / Published online: 11 June 2018
# Springer-Verlag GmbH Austria, part of Springer Nature 2018

Abstract
We measured air temperature at 14 sites with different land cover composition within the urban canopy layer of a mid-sized
Brazilian city. The intensity (ΔT) of the urban heat island (UHI) was calculated using data collected above a lake and at an urban
park as references. We investigated the spatio-temporal variability of ΔT during four contiguous days with varying weather. The
first day was overcast and rainy, giving rise to a moderate UHI. The second day was sunny, which caused the diurnal ΔT fields to
become heterogeneous, due to larger heating rates at sites with more man-made surfaces compared to natural surfaces. A high-
pressure system observed on the last days brought cloudless skies, causing smaller ΔT during the day and greater at night. We
hypothesise that the effect was due to the reduction of cooling via evapotranspiration caused by closing of the stomata as the soil
dried out, which reduced the daytime temperature differences among the sites. The night-time effect was caused by stronger
radiative cooling due to clear skies. The temperature within the park was always lower than over the lake, confirming that urban
forestry is a more effective mechanism to combat the UHI. Introducing a park would be about sevenfold cheaper than building a
city pond. Hence, green spaces are not only more efficient to combat the UHI but it is also a cheaper strategy compared to blue
spaces. Moreover, vegetation delivers other benefits, such as removal of air pollutants, attenuation of urban noise, improvement
of city aesthetic and their use as recreational spaces.

Keywords Urban climate . Urban greenery . Air temperature field . Surface energy balance

1 Introduction rural areas (Oke 1988). The UHI arises from the energy bal-
ance within the urban canopy layer (UCL)—the part of the
Urban surfaces usually consist of manufactured materials (e.g. atmosphere below the height of the buildings—as a result of
cement, asphalt and roof tiles) with lower albedo and greater the incoming and outgoing energy fluxes between the atmo-
specific and volume heat capacities than natural surfaces. Due sphere and the urban system. The energy exchange is mediat-
to their high thermal inertia, they retain significantly more ed by radiation, advection, convection and conduction pro-
thermal energy than green (e.g. urban forests, parks, tree cesses. Strictly speaking, the phenomenon must be referred
patches) or blue spaces (e.g. lakes, ponds), raising the temper- to as canopy layer urban heat island, since it represents a
ature of the air aloft much more efficiently (Ramamurthy and difference in temperature between the air contained in the
Bou-Zeid 2016). This behaviour gives rise to the phenomenon UCL and the air in the near-surface layers of the countryside
dubbed urban heat island (UHI), whereby the air temperature (Oke et al. 2017). However, in this work, we will use the
in the urban core is higher than in the surrounding suburbs and acronym UHI for the sake of simplicity.
The energy balance within the UCL is composed by the
anthropogenic energy flux (mainly due to vehicular emis-
sions, stationary sources and the metabolism of people), the
* Admir Créso Targino
[email protected] turbulent flux of sensible heat, the turbulent flux of latent heat
(associated with the evaporation of water or deposition of
1 water vapour onto surfaces), the heat flux into the substrate
Graduate Program in Environmental Engineering, Federal University
of Technology, Av. Pioneiros 3131, Londrina 86036-370, Brazil and the heat added to, or subtracted from, the UCL by advec-
2 tion (Oke 2017, 1988; Ferreira et al. 2010). These factors are
Department of Environmental Engineering, Federal University of
Technology, Av. Pioneiros 3131, Londrina 86036-370, Brazil well correlated with the city population, built-up density and

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

972 A. C. Targino et al.

fraction of green and blue spaces (Hung et al. 2006). The of urbanisation on the air temperature, when Luke Howard
anthropogenic energy flux represents a positive contribution recorded differences of 0.2 and 2.0 °C during daytime and
to the energy balance, whilst the urban features (such as the night, respectively, between readings in central and suburban
built environment and the presence of vegetation and water London in the early nineteenth century. Since then, a vast
bodies) control the signal of the sensible and latent fluxes number of research has identified the UHI using paired air
which, in turn, determine the intensity of the UHI. temperature sensors, array of sensors and mobile transects. It
Impervious materials divert the energy into heating up the is beyond the scope of this work to report the findings of
surface and adjacent air, whereas vegetation uses it to evapo- individual studies. However, it is important to highlight that
rate water off of the leaves and soil to cool the surrounding air. these rather straightforward techniques have proved
Off-site factors, especially the large-scale atmospheric circu- successful in identifying the UHI in small, medium and
lation, have a substantial effect on the intensity of the UHI, as large cities. For example, Cardoso et al. (2017) performed
the surface-mediated heating and cooling differ considerably air temperature transects across Rancharia (a small city in
depending on the meteorological setting (Targino et al. 2014; Brazil) and found maximum UHI intensity of 2.9 °C.
Kim and Baik 2005; Hart and Sailor 2009). Aflaki et al. Targino et al. (2014) reported the largest UHI intensity
(2017), Ramamurthy and Bou-Zeid (2016) and Kleerekoper (6.0 °C) under high-pressure conditions in the city centre of
et al. (2012) provide excellent discussions on different off- Londrina, and Kolokotroni and Giridharan (2008) found max-
and on-site drivers of the UHI. imum nocturnal UHI of 8.6 °C in London when comparing air
In the first part of this study, Targino et al. (2014) measured temperature from 80 urban sites with a rural station.
the air temperature in urban and suburban areas of Londrina— In regard to remote sensing, Gamarra et al. (2014) correlat-
a mid-sized Brazilian city—and reported that a pronounced ed surface albedo and surface temperature data from the
hot core emerged at night in the city centre, especially under MODIS instrument aboard the Terra and Aqua satellites
lingering high-pressure settings. They also observed an up- with in situ air temperature data. They reported a negative
trend in the number of days with daily maximum and mini- relationship between albedo and air temperature and a
mum air temperatures exceeding their historical 90th percen- positive relationship between surface and air temperatures.
tile counterparts (from 6 and 17 days, respectively, in 1976 to The authors discussed that the approach is a robust proxy to
42 and 57 in 2012). These numbers, albeit representative of quantify the UHI in the absence of in situ air temperature data.
one city, must be regarded as an eye-opener due to the impacts Lucena et al. (2013) mapped the thermal field over Greater
of increased air temperature on the local population. Rio de Janeiro using imagery from Landsat 5 Thematic
By 2050, 90% of Latin America and the Caribbean popu- Mapper (TM) for specific dates in 1987, 1998 and 2007.
lation will be urban inhabitants (United Nations, 2014) subject They found that the surface temperature over areas dominated
to the effects of the UHI and heat surges triggered by global by impervious materials was consistently higher than subur-
warming. Latin America’s fast shift to a highly urbanised re- ban areas and, hence, prone to the development of the UHI.
gion puts pressure on already strained natural resources and Combining the abovementioned remote sensing approaches,
poses challenges to address issues, such as access to fresh Hu and Jia (2010) analysed the influence of land use change
water (Shiklomanov and Rodda 2003), food security (Crush on UHI in greater Guangzhou (China) from 1980 to 2007 and
and Frayne 2011) and good air quality (Duh et al. 2008). The reported an increase in land surface temperature of 2.48 °C as
risks of exposure to unsafe air quality may be magnified when the green vegetation fraction decreased by 0.16 over the
the episodes occur in conjunction with air temperature ex- period.
tremes (Breiter et al. 2014a). Whilst observational studies can map the spatiotemporal
The UHI is associated with adverse health aspects, such as variability of the UHI and pinpoint the on- and off-site drivers
thermal discomfort (e.g. Nastos and Matzarakis 2012), in- of the phenomenon, approaches that use modelling tools have
crease in deaths due to cardiovascular diseases (Breitner et the advantage of isolating the factors that contribute most to
al. 2014b) and deterioration of the air quality. For example, the UHI and of assessing their relative importance. Ryu and
Fiore et al. (2015) observed a positive correlation between Baik (2011) used the Advanced Research Weather Research
surface ozone (O3) and increase in air temperature. Walcek and Forecasting (WRF) model and investigate how the anthro-
and Yuan (1995) showed that the production rate of O3 was pogenic heat, impermeability and the 3D urban geometry af-
6.98 ppb h−1 in cities and 3.08 ppb h−1 in cooler suburban fected the UHI in Seoul. They found that the impervious sur-
areas in the USA, and Rosenfeld et al. (1998) reported that O3 faces had the greatest effect on the daytime UHI intensity
level of 120 ppbv at 22 °C could double with a 10 °C increase (98%), whilst the anthropogenic heat contributed 86% to the
in air temperature. nocturnal UHI. Ashie et al. (1999) created different scenarios
The UHI has been assessed by different methods, such as in of vegetation around and on top of a building and reported air
situ observations, remote sensing data and modelling. In situ temperature reductions between 0.4 and 1.3 °C, with building
observations was the first technique used to identify the effect cooling energy savings of as much as 25%.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 973

Although the conditions that favour the appearance of the geographic regions and types of water reservoirs. The major-
UHI are well-established and understood, the implementation ity of the studies reported the positive influence of water bod-
of the best strategies to combat this phenomenon within cities ies on lowering the air temperature in urban regions. However,
that have already experienced quick and unplanned urbanisa- specific aspects, such as the water body geometry, location
tion remains controversial. Among the solutions proposed are and the fraction of surrounding built-up area, were key factors
reroofing and repaving in lighter colours; installing cool as- controlling the magnitude of the effect. For example, the effi-
phaltic and concrete photocatalytic pavements; decrease of ciency of the cooling effect in Beijing was positively correlat-
the anthropogenic heat; building envelope modifications using ed with the water body mass with a mean value of 1.76 °C/hm/
natural materials; harnessing natural wind; covering of roofs, ha (°C per one hectometre around the water body and per 1 ha
walls and façades with vegetation and introducing green infra- of water) (Sun and Chen 2012). Sugawara et al. (2009)
structure and urban water surfaces (Kandya and Mohan 2018; showed that the temperature of a shallow water channel (a
Gunawardena et al. 2017; Kyriakodisa and Santamouris 2018; few centimetres deep) in Tokyo exceeded the air temperature
Kleerekoper et al. 2012; Rosenfeld et al. 1998). Of special during daytime, but the air temperature over deeper water
interest to this study is the effect of green and blue spaces in channels (about 5 m) was cooler than the surrounding air,
cities, whose implementation is usually reliant upon the avail- which helped combat the UHI. Zhu et al. (2011) reported that
ability of open areas for trees or water bodies, the amount of rivers with width > 40 m promoted significant air temperature
financial resources and the climate. Whatever the strategy, the decrease in surrounding urban areas. Finally, urban features
basic principle consists in modifying the surface energy bal- that obstruct wind flows also reduce atmospheric advection
ance to divert the energy surplus into latent heat flux through and surface mixing. The Bwind path effect^ was discussed by
evapotranspiration and vaporisation. Suzuki (1999) who found that the cooling was more efficient
Most research outputs have converged to the conclusion that for rivers surrounded by open spaces than in densely built-up
the introduction of urban forestry is not only aesthetically pleas- areas. A drawback identified in many studies was that water
ant, but it is also very effective to cool the air and create a bodies have the ability to maintain water temperatures at night
phenomenon referred to as ‘park cool island’ through different time due to its thermal inertia and high heat capacity, which
mechanisms: (i) evaporation and evapotranspiration (incoming could lead to relative warming effect at night.
solar energy is used to transform liquid water into water va- The present study extends the research by Targino et al.
pour), (ii) reflectance (higher albedo compared to urban mate- (2014) and addresses the effectiveness of two urban redevel-
rials) and (iii) shading (the ground moisture is preserved and the opment scenarios to mitigate the UHI within one of the city’s
surface heating is diminished). For example, Ca et al. (1998) neighbourhoods: the introduction of a park or a lake. We also
reported that planting a 0.6-km2 park could reduce air temper- estimate the economic costs to implement such measures.
atures by 1.5 °C in Tama New Town (west of the Tokyo) with
the additional benefit of savings of 4000 kWh in an hour of a
summer day. Shashua-Bar and Hoffman (2000) reported that a 2 Methodology
0.15-ha park in Tel Aviv caused a cooling effect between 1.5
and 3.0 °C, and Upmanis et al. (1998) found nocturnal temper- We conducted this research in Londrina—a city of approxi-
atures about 6.0 °C cooler in a city park in Gothenburg com- mately 550,000 inhabitants located in the southern Brazilian
pared to the urban surroundings. A broad survey of outcomes in state of Paraná (lat. 23°18′36″S, lon. 51°09′46″W, alt. 630 m
regard to the benefits of green spaces for the urban climate above mean sea level). Following its foundation in 1934, the
would necessitate more room to be described here. However, semi-deciduous native forest that occupied the region was
it is relevant to cite the work by Peng et al. (2011) who analysed eradicated by slash-and-burn techniques to clear space for
the surface urban heat island across 419 global big cities and coffee plantations. The urban area grew rapidly in the 1960s
made linkages with several biophysical and socio-economic and 1970s, which resulted in a compact city centre with tall
drivers. They found that the surface temperatures correlated buildings (up to 20 storeys) and reduced green areas. Native
negatively with the vegetation cover and emphasised the key vegetation was further sacrificed over the years due to an
role of vegetation feedbacks in attenuating the phenomenon. increasing demand for urban infrastructure (sidewalks, com-
Water affects the air temperature aloft by two pathways: (i) mercial centres and parking lots), reducing the city’s urban
cooling from the evaporation and (ii) through the heat buffer green areas to only 16% (ca. 3.60 km2) (Polidoro et al.
effect that keeps the water temperature relatively unchanged 2011). Londrina has a humid subtropical climate (Cfa in the
due to its larger heat capacity (4184 against 920 J kg−1 K−1 of Köppen–Geiger classification) with an annual mean tempera-
asphalt, for example). Research on the effects of urban water ture of 21.0 °C. In winter, the mean temperature ranges from
bodies on the air temperature abounds, and a thorough review 11.6 to 25.8 °C and from 19.0 to 29.7 °C in summer. The city
on the subject is given by Manteghi et al. (2015) who has relatively dry winters and rainy summers, with annual
discussed 26 articles featuring different spatial scales, data, mean precipitation of 1630 mm. Insolation is greatest during

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

974 A. C. Targino et al.

winter time (mean value of 225 h) when cloud coverage is at 3 Results

its lowest level (Targino et al. 2014).
3.1 Land cover analysis
2.1 Study area
Figure 3 shows the results of the land cover analysis around
We focused our study on a relatively new urbanised, densely the monitoring sites. The sites presented a large variability in
populated area in the southern part of the city. This area is urban, water and greenery density, with asphalt and roof cover
ideal to investigate the influence of green and blue spaces on as large as 91% around Lessence Building (LES) to null
the air temperature, due to the presence of a lake and a remnant around Igapó Lake (LAG). The white areas are due to errors
patch of native trees. The Igapó Lake covers an area of ap- during the classification, caused by wrong interpretation of the
proximately 168,100 m2 and has an average depth of 2.0 m. spectral properties. However, these errors represent typically
The lake is non-swimmable but the margins have a recreation- less than 5% of the analysed area and do not affect the overall
al function sought by outdoor enthusiasts and for social cohe- interpretation. Table 1 shows the land cover percentage for
sion. The patch of native trees (hereinafter BValley^) covers an each category around the sites.
area of 31,966 m2 and is located about 480 m north of Igapó
Lake. The valley is a non-walkable, densely vegetated area, 3.2 General overview
consisting mostly of evergreen trees with large trunks and
well-developed crowns. During the campaign, mean daily air temperatures ranged be-
tween 15.3 °C at Igapó Residence (RMI)—an area with about
2.2 Data collection and treatment 60% of vegetation—and 21.2 °C at LES—an area with about
90% of urban features. Due to instrumental malfunction, the
We prospected the streets around Igapó Lake and Valley and data at VAL was collected until 5 June, 2013. All statistical
identified 14 suitable sites to deploy air temperature sensors. calculations were performed considering data up to this point.
The choices were based on the land cover, easy access to Figure 4 shows a statistical summary of the air temperature at
install and routinely check the instruments, as well as protec- each site as boxplots, organised in descending order of vege-
tion against theft and vandalism (Fig. 1). We measured air tation cover fraction. Sites with larger fraction of vegetation
temperature from 23 May to 19 June 2013 using data loggers (e.g. VAL, RMI and HAR) showed lower median and mean
with built-in sensors capable of measuring in the range be- temperatures and relatively lower extreme values (percentile
tween − 40 and 70 °C, with ± 0.2 °C accuracy and 0.02 °C 95), whilst sites with large fraction of man-made surfaces (e.g.
resolution. The data acquisition interval was set to 2 min. All LES, STU and ENS) presented greater median, mean and
data loggers were nested inside radiation shields to reduce minimum air temperatures. The different behaviour of air tem-
interference from direct radiation and wind and deployed at perature over man-made and natural surfaces is related to the
heights between 1.1 and 4.8 m above the ground. The data interaction of incoming solar radiation with the materials.
from the sensors installed within the Valley canopy (VAL) and High-absorbing materials use the energy to heat up the surface
mounted on a post in the centre of Igapó Lake (LAG), about and adjacent air whilst vegetation uses most of the energy in
1.8 m above the water (Fig. 2), were used as reference. All photosynthesis and evapotranspiration processes rather than
times in this study are reported in local winter time (UTC-3) to heat up the leaves and adjacent air. As consequence, the
and as 24-h notation. air temperatures in built-up areas reached higher values during
Following Targino et al. (2014), we performed an image the day compared to vegetated areas. This process is con-
segmentation and unsupervised classification using the trolled by the availability of moisture in the soil, and the day-
Georeferenced Information Processing System (SPRING) time differences tend to smooth out as the soil becomes drier.
software (Camara et al. 1996) around each site. We used Besides being within a highly urbanised area, LES, STU and
high-spatial resolution Google Earth images from the ENS are surrounded by high-rise buildings. After sunset, the
QuickBird satellite to evaluate the land cover in a circular area energy released by the urban canopy may be absorbed and re-
of radius 150 m centred at each site. The image classification emitted by the surrounding buildings, slowing down the cooling
was performed with the following steps: (a) imagery rate compared to rural sites or sites with higher sky-view factor.
importing and geo-referencing with aid of a high-resolution For example, Targino et al. (2014) reported cooling rates twice as
image of the Shuttle Radar Topography Mission (SRTM); (b) large at a vegetated site compared to an urban canyon site in
performing supervised segmentation (identification of regions Londrina at sunset (− 2.63 and − 1.25 °C h−1, respectively).
made up of pixels that have similar spectral properties); (c) Figure 5 shows the mean daily air temperature for all sites
training by selecting regions (ground truth) of known land during the sampling period. In general, the curves followed one
cover categories: trees, grass, asphalt, roof, water and bare soil another closely, with a few outstanding features. Over some
and (d) classifying into the six categories. periods, large daily intra-site air temperature differences were

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 975

Fig. 1 Sites where the air

temperature sensors were
deployed. The sites are identified
with three-letter acronyms de-
fined in Table 1

observed, whilst the difference was modest during other periods. Our study period was characterised by cloudy and rainy
For example, on 08 June, the average daily difference between days associated with cold front passages and clear sky days
sites with different land cover was up to 3.1 °C (LES and ATE), caused either by anticyclones that occurred in the wake of
whilst on 16 June, the difference between the same sites was cold fronts or by lingering high-pressure conditions. For
0.04 °C. Targino et al. (2014) also observed a similar behaviour example, the maximum solar radiation was only 70 W m−2
and argued that since the city’s fabric makeup was unaltered on 28 May and 715 W m−2 on 03 June (Fig. 6). Rain
over the study period, the differences in the air temperature were occurred in 17 days and amounted to 204.5 mm. 23 and
controlled not only by local but also by external variables, such 28 May and 02 June together accounted for more than half
as cloud cover and increased amount of incoming solar radiation of the total rainfall over the period, with 29.0, 68.8 and
(usually observed under anticyclonic conditions). 45.5 mm, respectively.

Fig. 2 a Deployment of the air

temperature sensor and radiation
shield on a post at the centre of
Igapó Lake. b Radiation shield in
place

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

976 A. C. Targino et al.

Fig. 3 Land cover analysis Valley (VAL) Hotel Harbour (HAR) Igapó residence (RMI)
around the monitoring sites

IATE club (IAT) Landfill (ATE) Police station (DEL)

Energy gym (GYM) Tijuca street (TIJ) Mobile shop (MOB)

Studio building (STU) Palhano sports (PAL) Lessence building (LES)

Enseadas building (ENS) Igapó lake (LAG)

Trees
Grass
Asphalt
Roof
Water
Bare soil

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 977

Table 1 Land cover percentage

around each site obtained using Sites Land cover (%)
an image segmentation and
unsupervised classification Trees Grass Asphalt Roof Water Bare soil

Valley VAL 76.7 0.0 8.5 12.7 0.0 2.1

Hotel harbour HAR 45.6 25.8 8.6 18.4 0.2 1.3
Igapó residence RMI 38.6 19.9 14.5 17.5 9.7 0.0
Iate Clube IAT 27.8 11.8 18.3 1.3 40.7 0.0
Landfill ATE 22.4 22.8 31.2 20.5 0.1 2.9
Police station DEL 22.4 15.1 30.4 30.2 0.1 1.7
Energy gym GYM 22.1 35.8 24.5 15.2 0.4 2.0
Tijuca street TIJ 22.1 10.0 28.8 27.9 5.3 5.9
Mobile shop MOB 12.4 10.0 31.2 45.7 0.7 0.0
Studio building STU 12.3 0.0 34.2 53.0 0.5 0.0
Palhano sports PAL 7.8 20.0 12.3 53.6 0.0 6.2
Lessence building LES 7.5 0.0 9.8 81.6 1.1 0.0
Enseadas building ENS 7.5 7.5 17.1 62.1 1.7 4.1
Igapó lake LAG 1.2 3.5 0.0 0.0 95.3 0.0

3.3 Spatial distribution of air temperature and maximum air temperatures for the three radiative condi-
tions are shown in Fig. 7. The smallest differences across the
We segregated the dataset according to the amount of integrat- region were observed for dose 1. This radiative regime was
ed daily incoming solar radiation into three categories: dose 1 associated with cold fronts and accompanied by thick cloud-
(< 6 MJ m−2), dose 2 (between 6 and 12 MJ m−2) and dose 3 iness and rainfall. The reduced incoming solar radiation
(> 12 MJ m−2). The spatial distributions of the mean minimum disrupted the heating discrepancies among sites with different

Fig. 4 Statistical summary of the air temperature at the surveyed sites. The green circles indicate the mean values

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

978 A. C. Targino et al.

22
VAL HAR RMI IAT ATE DEL GYM TIJ MOB STU PAL LES ENS LAG

21

20
Air temperature [°C]

19

18

17

16

15
23 May 26 May 29 May 01 Jun 04 Jun 07 Jun 10 Jun 13 Jun 16 Jun
Date [2013]
Fig. 5 Time series of daily air temperature at the monitored sites

land covers, kept the maximum air temperatures from rising high-pressure systems. Both settings brought clear skies and
and the UHI intensity moderate (Fig. 7a). Despite the modest advection of cold air. During the day, the sites responded to
amount of incoming solar radiation, it was enough to cause a the solar heating and the surfaces made of manufactured mate-
difference of about 2.0 °C across the area with a cool core rials heated the air aloft more efficiently than the natural sur-
around VAL, TIJ and LAG. The cloud cover trapped the out- faces, favouring the development of stronger spatial gradients
going nocturnal long-wave radiation, reducing radiative in air temperature. The maximum air temperature fields showed
cooling and contributing to keep the minimum air tempera- heterogeneous features with a pronounced cool interfacial re-
tures relatively high (Fig. 7d). gion in the central part of the neighbourhood and warmer areas
Doses 2 and 3 were associated with cold anticyclones in the southeast and northeast corners (Fig. 7b, c). These warm
(which occurred in the wake of cold fronts) and lingering areas were strongly influenced by the large amount of solar

Fig. 6 Incoming solar radiation 19

over the study period 700

0 100
10
17 100

0
10
100

600 300
100
100

300

300
300

100
300 00

500
300

0 50
15
300

500 50 0
100

5
500

0
30
500
300

500
100

300
500

100
[W m ]
-2

500

400
Hour

13
100

300
500
500

300
500
500

300

0
500

300
300

50

300
300

11
30

100
100
0

300
100
100

500
200
300
100

9
0
100 10 100
100

0
7
23 May 26 May 29 May 01 Jun 04 Jun 07 Jun 10 Jun 13 Jun 16 Jun
Date [2013]

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 979

Temp. [°C]
a b c

d e f

Fig. 7 Spatial distributions of maximum air temperature for a dose 1, b dose 2 and c dose 3 and minimum air temperature for d dose 1, e dose 2 and f dose 3

radiation absorbed by buildings, pavements and anthropogenic the reference microenvironments; the larger and more positive
activities around the sites STU, MOB and LES. The warmest ΔT, the cooler the air temperature around the reference site.
areas were around MOB, reaching about 26.0 and 28.1 °C The inspection of the hourly UHI over a contiguous period of
during dose 2 and 3, respectively. varying radiation doses and rainfall is a finer approach than
At night, the pool of cold air advected by the anticyclones using bulk data over an extended period. Figure 8 shows the
combined with clear skies led to a rapid cooling and to lower spatiotemporal variability of ΔTL and ΔTV.
minimum air temperatures than during dose 1. The diverging On 29 May, ΔTL and ΔTV were modest and ranged be-
nocturnal cooling rates rendered a heterogeneous minimum air tween 0.8 and − 0.8 °C at most between 06 and 18 h. The
temperature field, with the greatest values around STU and LES homogeneous air temperature field was caused mainly by
(Figs. 7e, f). Note that dose 3 represented conditions with clear- the thick cloud cover that brought rainfall and yielded a small
er skies than dose 2, why the former caused larger maximum air solar radiation dose throughout the day. These conditions not
temperatures and lower minimum temperatures than the latter. only decreased the air temperature, but they also suppressed
A hallmark in the three radiative regimes is a core of cold air the spatial variability across areas with different land cover.
around VAL and LAG in the maximum air temperature fields. Despite the rainfall on 30 May (0.8 mm around 01 h), the
increased radiation dose caused larger temperature differences
3.4 Effect of the valley and lake on the UHI across the area. The ΔT fields started to become heteroge-
within the study area neous at 10 h, and at 13 h the Valley was 4.5 °C cooler than
the southern sector, whilst the Lake was only 1.8 °C cooler.
To illustrate the influence of urban forestry and water on the air Sites with more man-made surfaces, such as MOB and LES,
temperature field, we focus on the UHI during four contiguous had heating rates of 2.5 °C h−1 between 10 and 11 h whilst
days (29 May through 01 June) with varying weather condi- RMI and VAL had rates of 0.5 °C h−1, which explains the
tions. The first day received rainfall and little incoming solar higher air temperatures at sites covered with impervious sur-
radiation, whilst the last days were dry and cloudless (Table 2). faces. This feature prevailed throughout the day until about
We used the air temperatures at the Lake (TL) and at the 19 h, thereafter, the ΔT field structure started to dissolve. It
Valley (TV) as references and compared the air temperature at persisted until about 07 h of 31 May when the northeast and
each site (Ti) against them on an hourly basis for the period 29 southwest corners of the study area were still about 2.5 °C
May through 01 June. The air temperature differences (ΔTL = warmer than the Valley. Lakshmi et al. (2003) and Ahmad
Ti − TL and ΔTV = Ti − TV) indicate the cooling potential of and Hashim (2007) showed that the soil water content plays

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

980 A. C. Targino et al.

Table 2 Ranges of
meteorological conditions from 29 May 30 May 31 May 01 June
29 May through 01 June, 2013
across the study area Maximum air temperature (°C) 19.7–20.7 20.0–24.2 22.2–24.4 24.9–26.8
Minimum air temperature (°C) 15.6–16.1 13.0–16.4 11.1–14.0 14.1–16.5
Solar radiation dose (MJ m−2) 5.4 8.4 11.3 15.0
Accumulated rainfall (mm) 8.4 0.8 0.0 0.0

an important role in the configuration of the air temperature cover. This effect was also observed by Targino et al. (2014)
field. Following the rain, impervious surfaces lose water rap- when comparing the air temperature gradients across
idly due to run-off whilst natural surfaces keep the soil water Londrina under lingering high-pressure conditions.
content for a longer period. In such conditions, the incoming The air temperature around the Valley was always lower
solar radiation is used to heat the built-up surfaces whilst nat- than around the Lake, indicating that the former has a larger
ural surfaces divert the energy for evaporative processes. This cooling potential than the latter. These results compare
results in larger spatial variability and milder air temperature favourably with the findings reported in the literature. For
over the latter. example, Morris et al. (2016) found that vegetation was more
As the high-pressure system established itself over the area efficient than water body in reducing the daily air temperature
on 31 May and 01 June producing sustained dry weather, in Putrajaya (Malaysia) (− 0.14 and − 0.39 °C, respectively).
cloudless skies and larger radiation dose, the ΔT differences Steeneveld et al. (2011) quantified the UHI in the Netherlands
became smaller during the day and greater at night. We ob- using in situ observations and reported a negative trend be-
served that the heating rates at built-up areas were very similar tween the vegetation fraction and the maximum daily intensity
to the rates at sites with larger vegetation fraction. For exam- of the UHI. The median of maximum intensity ranged from
ple, the rate at MOB and LES was 2.3 °C h−1 between 10 and 3.0 to 1.4 °C as the green fraction increased from 5 to 55%,
11 h on 31 May, whilst at RMI, LAG and VAL the rate was whilst no clear relation between UHI and water cover fraction
1.7 °C h−1. We hypothesise that as the green areas dried out was found.
during the day, the vegetation started to regulate the water loss
by closing the stomata and inhibiting the cooling via 3.5 Cost analysis
evapotranspiration; this reduced the air temperature
differences between vegetation and impervious surfaces. We assessed the costs to install an urban park and to build a
Kotthaus and Grimmond (2014) showed that evaporation is pond, both with an area of 5000 m2 (50 m × 100 m). We
important immediately following precipitation. However, the considered the following expenditures for the park: acquisi-
latent heat flux decreases rapidly in the first 12 h and the tion of nursery stocks, earth-moving work, digging, planting
sensible heat flux quickly regains its dominance. When the and maintenance activities (pruning, pest control and irriga-
incoming solar radiation is low, like on 29 and 30 May, water tion) during the first 3 years. The nursery stocks (about 2-m
remains on the surface for a longer period. Approximately high) consist of 16 native species which can grow medium to
between 12 and 18 h after rainfall and under large solar radi- large trunks, yield well-developed crowns and have an aver-
ation doses, most precipitation events cease exerting control age growth rate of 1.0 m year−1 (e.g. Pires et al. 2015; Alves et
over the partitioning of the latent and sensible heat fluxes, al. 2011): Luehea divaricata, Schinus molle, Peltophorum
especially in an urban park like ours surrounded by paved dubium, Croton floribundus, Cedrela fissilis, Parapiptadenia
roads and sidewalks which facilitate quick surface runoff. At rigida, Inga edulis, Tabebuia roseoalba, Guazuma ulmifolia,
other sites with more bare soil or vegetation fractions, rain Bauhinia forficata, Gallesia integrifolia, Aspidosperma
events can influence the energy balance over longer periods. cylindrocarpon, Tibouchina granulosa, Machaerium
A hallmark between 11 and 18 h on 31 May was that the stipitatum, Colubrina glandulosa and Eugenia pyriformis.
Lake was about 3.2 °C warmer than the Valley and the areas A pond is made by digging a pit and by filling it with water.
around HAR and RMI, characterised by a large fraction of For a humid subtropical climate, the pit should be at least 2.1-
vegetation cover. Between 08 and 20 h of 01 June, both ΔT m deep to offset probable seepage and evaporation losses (Soil
fields were rather homogeneous with modest values of − Conservation Service 1982). We considered the following ex-
0.8 °C around the lake. This was a clear and dry day with a penditures to build a 3-m deep pond on a level terrain: dig-
large solar radiation dose. The stronger nocturnal gradients ging, removal of waste material, soil compaction to make the
were attributed to the large amount of energy stored in the bottom impervious, landscaping, installation of pipelines and
impervious materials during the day and released after sunset spillway to pass excess storm runoff.
at a slower cooling rate due to the greater thermal inertia (the Tables 3 and 4 show the costs in US$ for both projects
hysteresis lag effect) compared to the sites with larger green (considering that US$1 exchanged at a rate of 3.44 Brazilian

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 981

Fig. 8 Spatio-temporal
distribution of ΔT using Valley
(left column) and Lake (right
column) as reference in the period
29 May to 01 June. We have
omitted the sites’ acronyms due to
space limitation

Real as of 24 April 2018). We did not factor in the costs of The largest expenditure to introduce a green space in
acquiring the property considering that the area belongs to the Londrina would be for maintenance activities. We considered
municipality. The prices were commissioned to specialised maintenance in the first 3 years because it is a critical period
engineering companies in Londrina and were given in local for young trees, as their growth relies on the availability of soil
currency. moisture and nutrients and protection from pests. Hence,

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

982 A. C. Targino et al.

Table 3 Cost estimates to build a

5000-m2 urban park Expenditure Price (US$/m2 or US$/unit) Amount (m2 or unit) Total (US$)

Nursery stocks 1.45 555 804.75

Earth-moving work 0.46 5000 2300
Digging and planting 0.58 555 321.90
Maintenance (3 years) 10.46 555 5805.30
Grand total 9231.95

watering, mulching (spreading of organic matter on the soil pond. The introduction of urban forestry is not only more
surface) and weed control are the main follow-up maintenance effective and cheaper to mitigate the UHI, but it also delivers
for tree survival (Kuhns and Rupp 2000). Newly planted trees a range of benefits such as reduction of runoff and erosion
only have the roots that grew in their containers, why keeping from storms, CO2 sequestration and reduction of air pollution
the roots moist is critical. Mulch applied to the area around the (e.g. Janhäll 2015). On that front, trees are also cost compet-
base of the tree acts as a covering that holds moisture in the itive with digging a pond.
soil and can improve soil quality as it decomposes. Finally,
combating weeds is crucial since they use water and mineral
nutrients that would otherwise be available for the newly 4 Conclusions
planted trees.
The overall costs would be substantially lower than in the We observed large air temperature variability across a
USA or Italy. McPherson et al. (2005) showed that the annual neighbourhood on a scale of ten to hundreds of metres, espe-
expenditures associated with planting and maintaining trees cially under the influence of anticyclonic circulation. During
ranged from US$15 per tree in the US Southwest region to the passage of cold fronts, the UHI was relatively modest
US$65 in Berkeley, California. Like in our study, maintenance irrespective of the land cover. An hour-to-hour analysis during
was the greatest expenditure (from US$4 to US$20 per tree), four contiguous days revealed a large spatio-temporal vari-
whilst tree planting accounted for the lowest expenditures ability in the intensity (ΔT) of the UHI. Overcast and rainy
(from US$0.50 to US$4 per tree). A study conducted in 40 conditions on the first day smoothed out the ΔT fields. On the
municipalities in the Veneto Region (Italy) showed that the second day, as the solar radiation dose became greater, the
maintenance cost was on average US$1.30 per m2 per year diurnal ΔT fields started to become heterogeneous due to
(Tempesta 2015), which would amount to US$6500 per year larger heating rates at sites with more man-made surfaces
against US$1935 in our case. (2.5 °C h−1) compared to natural surfaces (0.5 °C h−1). On
Comparing the costs of building a pond with other studies the third and fourth days, as the high-pressure system
is more difficult due to the different variables involved in the established itself over the area, bringing clear skies and larger
project, such as the terrain slope, whether the foundation must radiation doses, ΔT became smaller during the day and great-
be blanketed by a layer of impervious material, the require- er at night. We hypothesise that the effect was due to the
ments for the inlet and spillway, among other features. reduction of cooling via evapotranspiration caused by closing
However, as one of the objectives of this study is to compare of the stomata as the soil dried out, which reduced the daytime
costs for introducing urban forestry or building a pond, we temperature differences among the sites. The night-time effect
focus on this aspect only. From Tables 3 and 4, it becomes was caused by stronger radiative cooling due to clear skies.
clear that the costs to plant and maintain trees in an urban The temperature at the Valley was always lower than the tem-
setting are about sevenfold lower than the costs to build a perature over the Lake, confirming that urban forestry is a

Table 4 Cost estimates to dig a

5000-m2, 3-m deep pond Price (US$/m3 or US$/ Amount (m3 or unit) Total
unit) (US$)

Excavation and removal of waste material 3.05 15,000 45,750.00

Soil compaction 1.41 6800 9588.00
Pipeline 5436.00 1 5436.00
Spillway 5570.00 1 5570.00
Landscaping (grass planting) 1.60 2000 3d200.00
Grand total 69,544.00

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Green or blue spaces? Assessment of the effectiveness and costs to mitigate the urban heat island in a... 983

more effective mechanism to combat the UHI. The effect was Duh J-D, Shandas V, Chang H, George LA (2008) Rates of urbanisation
and the resiliency of air and water quality. Sci Total Environ 400:
more evident in the first 12 h following precipitation and be-
238–256
came less efficient as the soil dried out and the sensible heat Ferreira MJ, Oliveira AP, Soares J (2010) Anthropogenic heat in the city
flux regained its dominance. of São Paulo, Brazil. Theor Appl Climatol 104:43–56
In view of the results presented, cities must curb the erad- Fiore AM, Naik V, Leibensperger EM (2015) Air quality and climate
ication of vegetation in order to keep surfaces shaded and to connections. J Air Waste Manag 65:645-685
Gamarra NLR, Correa MP, Targino ACL (2014) Use of remote sensing to
retain soil moisture, which could divert solar radiation into retrieve surface albedo and land surface temperature in Londrina
evapotranspiration processes instead of heating. In terms of (Paraná): a contribution to urban heat island studies. Rev Bras
costs to implement mitigation strategies against the UHI, in- Meteorol 29:537–550
troducing a park is about sevenfold cheaper than building a Gunawardena KR, Wells MJ, Kershaw T (2017) Utilising green and
bluespace to mitigate urban heat island intensity. Sci Total Environ
city pond. Vegetation is not only cheaper and more efficient to 584–585:1040–1055
counterattack the UHI, but it also has other benefits, such as Hart MA, Sailor DJ (2009) Quantifying the influence of land use and
the reduction of runoff and erosion from storms, CO2 seques- surface characteristics on spatial variability in the urban heat island.
tration, depletion of air pollution, improvement of the city Theor Appl Climatol 19:975–988
aesthetic and recreational function. Hence, the presence of Hu Y, Jia G (2010) Influence of land use change on urban heat island
derived from multi-sensor data. Int J Climatol 30:1382–1395
green spaces brings about a win-win situation. Hung T, Uchihama D, Ochi S, Yasuoka Y (2006) Assessment with sat-
ellite data of the urban heat island effects in Asian mega cities. Int J
Acknowledgments We are grateful to those individuals and companies Appl Earth Obs Geoinf 8:34–48
that hosted the temperature sensors during the measurement campaign Janhäll S (2015) Review on urban vegetation and particle air pollution—
and the Fire Brigade of Londrina who helped install the air temperature deposition and dispersion. Atmos Environ 105:130–137
sensors. Kandya A, Mohan M (2018) Mitigating the urban Heat Island effect
through building envelope modifications. Energ Buildings 164:
Funding information This work was supported by Fundação Araucária 266–277
of Paraná (grant number 470/2010). Kim YH, Baik JJ (2005) Spatial and temporal structure of the urban heat
island in Seoul. J Appl Meteorol 44:591–605
Kleerekoper L, van Esch M, Salcedo TB (2012) How to make a city
climate-proof, addressing the urban heat island effect. Resour
Conserv Recy 64:30–38
References Kolokotroni M, Giridharan R (2008) Urban heat island intensity in
London: an investigation of the impact of physical characteristics
Aflaki A, Mirnezhad M, Ghaffarianhoseini A, Ghaffarianhoseini A, on changes in outdoor air temperature during summer. Sol Energy
Omrany H, Wang Z-H, Akbari H (2017) Urban heat island mitiga- 82:986–998
tion strategies: a state-of-the-art review on Kuala Lumpur, Singapore Kotthaus S, Grimmond CSB (2014) Energy exchange in a dense urban
and Hong Kong. Cities 62:131–145 environment—part I: temporal variability of long-term observations
Ahmad S, Hashim N (2007) Effects of soil moisture on urban heat island in Central London. Urban Clim 10:261–280
occurrences: case of Selangor, Malaysia. Humanity and Social Sci J Kuhns M, Rupp L (2000) Selecting and planting landscape trees. All
2:132–138 Current Publications. Paper 1220. https://digitalcommons.usu.edu/
extension_curall/1220
Alves EU, Andrade LA, Bruno RLA, Vieira RM, Cardoso EA (2011)
Kyriakodis G-E, Santamouris M (2018) Using reflective pavements to
Emergence and early growth of Peltophorum dubium (Spreng.)
mitigate urban heat island in warm climates—results from a large
Taubert seedlings under different substrata. Rev Ciênc Agron,
scale urban mitigation project. Urban Clim 24:326–339
https://doi.org/10.1590/S1806-66902011000200025
Lakshmi V, Zehrfuhs D, Jackson TJ (2003) Soil moisture–temperature
Ashie Y, Thanh VC, Asaeda T (1999) Building canopy model for the
relationships: results from two field experiments. Hydrol Process 17:
analysis of urban climate. J Wind Eng Ind Aerodynam 81:237–248
3041–3057
Breitner S, Wolf K, Devlin RB, Diaz-Sanchez D, Peters A, Schneider A
Lucena AJ, Rotunno Filho OC, França JRA, Peres LF, Xavier LNR
(2014a) Short-term effects of air temperature on mortality and effect
(2013) Urban climate and clues of heat island events in the metro-
modification by air pollution in three cities of Bavaria, Germany: a
politan area of Rio de Janeiro. Theor Appl Climatol 111:497–511
time-series analysis. Sci Total Environ 485:49–61
Manteghi G, bin Limit, H and Remaz, D (2015) Water bodies an urban
Breitner S, Wolf K, Peters A, Schneider A (2014b) Short-term effects of microclimate: a review. Mod App Sci. doi: https://doi.org/10.5539/
air temperature on cause-specific cardiovascular mortality in mas.v9n6p1
Bavaria, Germany. Heart 100:1272–1280
McPherson EG, Simpson JR, Peper PJ, Maco SE, Xiao Q (2005)
Ca VT, Asaeda T, Abu EM (1998) Reductions in air-conditioning energy Municipal forest benefits and costs in five US cities. J Forest 103:
caused by a nearby park. Energ Buildings 29:83–92 411–416
Camara G, Souza RCM, Freitas UM, Garrido J (1996) SPRING: integrat- Morris KI, Chan A, Ooi MC, Oozeer MY, Abakr YA, Morris KJK (2016)
ing remote sensing and GIS by object-oriented data modelling. Effect of vegetation and waterbody on the garden city concept: an
Comput Graph 20:395–403 evaluation study using a newly developed city, Putrajaya, Malaysia.
Cardoso RS, Dorigon LP, Teixeira DCF, Amorim MCCT (2017) Comput Environ Urban 58:39–51
Assessment of urban heat islands in small- and mid-sized cities in Nastos PT, Matzarakis A (2012) The effect of air temperature and human
Brazil. Climate https://doi.org/10.3390/cli5010014 thermal indices on mortality in Athens. Theor Appl Climatol 108:
Crush J, Frayne B (2011) Urban food insecurity and the new international 591–599
food security agenda. Dev South Afr 28:527–544 Oke TR (1988) The urban energy-balance. Prog Phys Geogr 12:471–508

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

984 A. C. Targino et al.

Oke TR, Mills G, Chirsten A, and Voogt JA (2017) Urban climates. Steeneveld GJ, Koopmans S, Heusinkveld BG, van Hove LWA, Holtslag
Cambridge University Press, p. 525 AAM (2011) Quantifying urban heat island effects and human com-
Peng S, Piao S, Ciais P, Friedlingstein P, Ottle C, Bréon F-M, Nan H, fort for cities of variable size and urban morphology in the
Zhou L, Myneni RB (2011) Surface urban heat island across 419 Netherlands. J Geophys Res https://doi.org/10.1029/
global big cities. Environ Sci Technol 46:696–703 2011JD015988
Pires MF, Pereira MP, Castro, EM, Barbosa S, Pereira, FJ (2015) Leaf Sugawara H, Narita K, Kim MS (2009) Cooling effect by urban river. The
micromorphometry of Schinus molle L. (Anacadiaceae) in different 7th International Conference on Urban Climate; 29 June–3 July 3,
canopy heights. Cerne https://doi.org/10.1590/ Yokohama, Japan
01047760201521011530 Sun R, Chen L (2012) How can urban water bodies be designed for
Polidoro M, Lollo J, Barros M (2011) Environmental impacts of urban climate adaptation? Landscape Urban Plan 105:27–33
sprawl in Londrina, Paraná, Brazil. J Urban Environ Eng 5:73–83 Suzuki C (1999) A climatological study of the cooling effect of urban
Ramamurthy P, Bou-Zeid E (2016) Heatwaves and urban heat islands: a rivers on heat island phenomena. Tokyo Metropolitan University,
comparative analysis of multiple cities. J Geophys Res https://doi. Dissertation
org/10.1002/2016JD025357 Targino AC, Krecl P, Coraiola GC (2014) Effects of the large-scale atmo-
spheric circulation on the onset and strength of urban heat islands: a
Rosenfeld AH, Akbari H, Romm JJ (1998) Cool communities: strategies
case study. Theor Appl Climatol 117:73–87
for heat island mitigation and smog reduction. Energ Buildings 28:
Tempesta T (2015) Benefits and costs of urban parks: a review. Aestimum
51–62
67:127–143
Ryu Y-H, Baik J-J (2011) Quantitative analysis of factors contributing to United Nations (2014) World urbanization prospects: the 2014 Revision
urban heat island intensity. J Applied Met Clim 51:842-854 United Nations Population Division
Shashua-Bar L, Hoffman ME (2000) Vegetation as a climatic component Upmanis H, Eliasson I, Lindqvist S (1998) The influence of green areas
in the design of an urban street: an empirical model for predicting the on nocturnal temperatures in a high latitude city (Göteborg,
cooling effect of urban green areas with trees. Energ Buildings 31: Sweden). Int J Climatol 18:681–700
221–235 Walcek CJ, Yuan H-H (1995) Calculated influence of temperature-related
Shiklomanov IA, Rodda JC (2003) World water resources at the begin- factors on ozone formation rates in the lower troposphere. J Appl
ning of the twenty-first century. Cambridge University Press, Meteorol 34:1056–1069
Cambridge Zhu C, Li S, Ji P, Ren B, Li X (2011) Effects of the different width of
Soil Conservation Service (1982) Ponds—planning, design and construc- urban green belts on the temperature and humidity. Acta Ecol Sin
tion agricultural handbook 590, US Department of Agriculture 31:383–394

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]