Proceeding Paper

Urban Heat Island Intensity Prediction in the Context of Heat

Waves: An Evaluation of Model Performance †
Aner Martinez-Soto 1,2, * , Johannes Fürle 2 * and Alexander Zipf 2,3 *

1 Department of Civil Engineering, Faculty of Engineering and Science, Universidad de La Frontera,

Temuco 4780000, Chile
2 GIScience, Institute of Geography, University of Heidelberg, 69120 Heidelberg, Germany;
[email protected] (J.F.); [email protected] (A.Z.)
3 Heidelberg Institute for Geoinformation Technology gGmbH, 69118 Heidelberg, Germany
* Correspondence: [email protected]; Tel.: +56-45-2596816
† Presented at the 9th International Conference on Time Series and Forecasting, Gran Canaria, Spain,
12–14 July 2023.

Abstract: Urban heat islands, characterized by higher temperatures in cities compared to surrounding
areas, have been studied using various techniques. However, during heat waves, existing models
often underestimate the intensity of these heat islands compared to empirical measurements. To
address this, an hourly time-series-based model for predicting heat island intensity during heat
wave conditions is proposed. The model was developed and validated using empirical data from
the National Monitoring Network in Temuco, Chile. Results indicate a strong correlation (r > 0.98)
between the model’s predictions and actual monitoring data. Additionally, the study emphasizes
the importance of considering the unique microclimatic characteristics and built environment of
each city when modelling urban heat islands. Factors such as urban morphology, land cover, and
anthropogenic heat emissions interact in complex ways, necessitating tailored modelling approaches
for the accurate representation of heat island phenomena.

Keywords: urban heat islands; heat waves; prediction model

Citation: Martinez-Soto, A.; Fürle, J.;

Zipf, A. Urban Heat Island Intensity
1. Introduction
Prediction in the Context of Heat
Waves: An Evaluation of Model
An urban heat island (UHI) is defined as the temperature difference observed between
Performance. Eng. Proc. 2023, 39, 80. urban areas and the surrounding rural regions [1]. UHIs can occur in any season of the
https://doi.org/10.3390/ year and any time of day [2]. However, their effects are more noticeable during periods
engproc2023039080 of temperature increase (e.g., heatwaves in summer). The increase in global temperatures
due to global warming intensifies the effect of urban heat islands [3]. As the ambient
Academic Editors: Ignacio Rojas,
temperature rises, urban areas experience even higher temperatures [4].
Hector Pomares, Luis Javier Herrera,
High temperatures in heat islands lead to the need for air conditioning and cooling
Fernando Rojas and Olga Valenzuela
in buildings, which increases energy consumption and results in higher greenhouse gas
emissions, further contributing to global warming [5,6]. Furthermore, high temperatures
Published: 12 July 2023 can have adverse health effects on individuals, such as heat strokes, dehydration, and
respiratory problems. For example, during the summer of 2003 in Europe, more than 70,000
additional deaths were attributed to heat waves [7]. Elderly individuals and households
without access to air conditioning systems are identified as the first at-risk group. However,
Copyright: © 2023 by the authors.
this risk level increases in urban heat islands, making the identification of these areas crucial
Licensee MDPI, Basel, Switzerland.
This article is an open access article
for the development of mitigation measures (e.g., incorporating green spaces or planning
distributed under the terms and
open spaces that promote air circulation and shade), as well as for the protection of people.
conditions of the Creative Commons
Various techniques are employed to map urban heat islands in cities [8–10]. These
Attribution (CC BY) license (https:// include satellite remote sensing for large-scale temperature assessment, ground-based
creativecommons.org/licenses/by/ sensors and weather stations for real-time and precise data collection, aerial thermography
4.0/). using infrared cameras mounted on aircraft or drones to obtain detailed thermal images,

Eng. Proc. 2023, 39, 80. https://doi.org/10.3390/engproc2023039080 https://www.mdpi.com/journal/engproc

Eng. Proc. 2023, 39, 80 2 of 6

on-site temperature measurements using portable thermometers or thermographic devices,

and simulation models that incorporate urban geometry, land use, vegetation, and solar
radiation to predict and map heat islands [11–19]. A combination of these techniques and
data sources is crucial to gain a comprehensive understanding of heat islands, enabling
informed decision-making in urban planning, mitigation strategies, and the informed
safeguarding of residents’ health [13,20].
This study presents a combined technique for locating heat islands in the city of
Temuco, Chile, as a case study. Using data from 23 monitoring stations and utilizing
QGIS, areas with higher temperatures were mapped. Subsequently, a methodology for
predicting heat islands was proposed for days when the external temperature exceeded
30 degrees Celsius for 3 consecutive days (heatwaves). The results are validated by com-
paring modeled values for specific heat island sectors in the city with actual measurements
taken during heatwave days in the summer of 2019. Due to the accuracy of the results
(r > 0.98), it is concluded that it is possible to predict the location of heat islands during
heatwave events using the proposed methodology.

2. Methods
2.1. Case Study
Temuco is located in a valley surrounded by hills and mountains. The city sits at
an altitude of approximately 350 m above sea level and is crossed by the Cautín River.
The predominant vegetation in the area is the temperate rainforest, characteristic of the
southern zone of Chile. Temuco is a relatively large city with a population of around
300,000 inhabitants. It is an urban center that is constantly growing and developing. The
city is an important commercial, educational, and cultural hub in the region, offering a
wide range of services and activities.

2.2. Measurement of the Temperature and Mapping of the Heat Island

To measure the temperature in different sectors of the city, monitoring stations be-
longing to the National Monitoring Network (ReNaM) of Chile were used. The network
Eng. Proc. 2023, 39, 80 3 of 7
in Temuco consists of 23 weather stations (from Netatmo) represented by black dots in
Figure 1, which are installed in private properties across various zones in Temuco.

Z2
Z4
Z1
Z3

Figure 1. 1.
Figure Mapping of the
Mapping UHIUHI
of the phenomenon in Temuco
phenomenon usingusing
in Temuco fixed fixed
station methodology
station (left) and
methodology (left) and
IDW interpolation in QGIS. Temperatures from 4 December 2019, at 2 pm, are taken into account.
IDW interpolation in QGIS. Temperatures from 4 December 2019, at 2 pm, are taken into account.
2.3. Prediction of Heat Islands during Heatwaves
To predict the intensity of heat islands in the city of Temuco, temperature differences
were observed/analyzed between the weather station at Maquehue Airport and the sta-
tions situated in the four zones within the city that recorded the highest temperatures. The
Eng. Proc. 2023, 39, 80 3 of 6

The Netatmo weather stations comprise two devices (indoor and outdoor) made of UV-
resistant aluminum, capable of measuring temperatures ranging from −40 ◦ C to 65 ◦ C with
an accuracy of ±0.3 ◦ C. The outdoor sensors are shielded from rain and direct sunlight to
prevent deterioration and to ensure better data accuracy. Data are captured at thirty-minute
intervals, following a fixed schedule to maintain consistency (e.g., 8:00–8:30–9:00, etc.).
The sensors underwent calibration and validation by the Ministry of Housing and Urban
Development (MINVU) in collaboration with the Chile Foundation.
For the geolocation of heat islands, individual values from each station were used at
the same hour (e.g., 2 pm). Subsequently, the temperature values along with the station
coordinates were inputted into the QGIS software. The IDW interpolation technique was
employed in QGIS to map the heat islands into a continuous space within the city (Figure 1).
In Figure 1, the spatial distribution of temperatures measured in the city on 4 December
2019, at 2 pm, is shown. Here, there are four zones in the city (Z1–Z4) where the temperature
is higher than the measurement taken in the outer part of the city (Maquehue Airport). In
Z1 and Z2 the central part of the city (characterized by higher building density and low
vegetation), it is observed that the temperature (31 ◦ C) is 7 ◦ C higher than at the Maquehue
Airport station (24 ◦ C).

2.3. Prediction of Heat Islands during Heatwaves

To predict the intensity of heat islands in the city of Temuco, temperature differences
were observed/analyzed between the weather station at Maquehue Airport and the stations
situated in the four zones within the city that recorded the highest temperatures. The
study specifically focused on visualizing the characteristics of heat islands during two
heatwave episodes (referred to as HW1 and HW2) that occurred during the summer of
2020. By recording the temperature differences, Equation (1) has been formulated to
describe the temperature in the 4 zones in the city (with the highest temperatures Z1–Z4)
as a function of the temperature at Maquehue (TR = reference temperature ) for each hour.
Subsequently, these equations have been used to generate a general 24 h model that predicts
the temperature in the four sectors of the city based on the temperature recorded at the
Maquehue Airport station.
Ti,z ( TRi ) = ai,z TRi + bi,z (1)
where:
T = temperature;
TR = reference temperature (measured at the Maquehue airport);
i = time of a day in hours (1, 2 . . . 24);
z = zone in Temuco where heat islands are identified (1, 2, 3, 4).

3. Results
3.1. Measurement and Recording of Temperatures during Heatwave Conditions
In Figure 2, a temperature comparison appears for two heat wave episodes (HW1 and
HW2) that occurred in February 2020 between the Maquehue station (blue and red line
respectively) and the zones with the highest temperature in Temuco (Z-1 to Z-4 in Figure 1).
Here, it is observed first that the maximum temperature in Temuco occurs between 3 pm
and 5 pm, which represents a behavior that does not follow the common pattern of the UHI
phenomenon. The maximum temperature peak in most of the cases studied was reached
between 7 pm and 8 pm. However, in the case of Temuco at that time, the temperature in
the city dropped, whereas outside the city (reference temperature in Maquehue airport) the
temperature reached its daily maximum.
 Eng. Proc. 2023, 39, 80 4 of 7

the temperature in the city dropped, whereas outside the city (reference temperature in
Eng. Proc. 2023, 39, 80 Maquehue airport) the temperature reached its daily maximum. 4 of 6

reference temperature in HW1

50 reference temperature in HW2
Z1-1
Z2-1
Z3-1
40 Z4-1

Temperature (°C)
Z1-2
Z2-2
30 Z3-2
Z4-2

20

10

0
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72

Time (h)

Figure 2.
Figure 2.Comparison
Comparisonof temperature for two
of temperature forheat
twowave
heatepisodes between between
wave episodes MaquehueMaquehue
station ((blue
station ((blue
and red line respectively)) and the zones with the highest temperature in Temuco (Z1, Z2, Z3, Z4).
and
Datered linewave
for heat respectively)) andFebruary
1 (HW1): 8–10 the zones with
2020. Datethe
forhighest
heat wavetemperature
2 (HW2): 20–22in Temuco (Z1, Z2, Z3, Z4).
February 2020.
Date for heat wave 1 (HW1): 8–10 February 2020. Date for heat wave 2 (HW2): 20–22 February 2020.
Additionally, the maximum temperatures recorded in the city had an average in-
creaseAdditionally,
of 26 °C in a the
shortmaximum temperatures
period of time (between 10 recorded
am and in the city
3 pm). The had
samean average
speed of increase
of ◦
26 C in a change
short period of time
temperature appeared from(between
4 pm where 10 am and 3 pm).
it reached The same
on average speed
5.3 °C of temperature
per hour.
This is higher
change than the
appeared from maximum acceptable
4 pm where temperature
it reached change5.3
on average rate◦ C
(3 per
°C/h), which
hour. is is higher
This
set tothe
than prevent the human
maximum body from
acceptable suddenly feeling
temperature changehot or cold.
rate ◦ C/h),
(3 This suggests
which theisneed
set to prevent
to conduct
the humanmore bodydetailed
from studies on the
suddenly behavior
feeling hotofor
thecold.
phenomenon in Temuco.
This suggests theAnalyz-
need to conduct
ing from the materiality of constructions, green areas, etc., several factors may explain
more detailed studies on the behavior of the phenomenon in Temuco. Analyzing from the
more precisely the behavior of the UHI phenomenon in Temuco.
materiality of constructions, green areas, etc., several factors may explain more precisely
the
3.2. behavior ofTemperatures
Prediction of the UHI phenomenon in Temuco.
during Heatwave Conditions
Using Equation (1), the temperature profiles of the zones with the highest tempera-
3.2. Prediction of Temperatures during Heatwave Conditions
tures in the city (Z1, Z2, Z3, and Z4) have been predicted based on the temperature rec-
ordedUsing Equation(HW3)
at Maquehue (1), the temperature
during profiles of
a new heatwave the
that zones with
occurred the highest
on March temperatures
2020. The
in the citytemperatures
predicted (Z1, Z2, Z3,were
and compared
Z4) have with
beenthe
predicted
actually based ontemperatures
recorded the temperature
during recorded at
the third heatwave
Maquehue (HW3)that took place
during between
a new Marchthat
heatwave 1st and March on
occurred 3rd March
in order2020.
to validate
The predicted
the model and verify
temperatures were the accuracy with
compared of thethe
results (Figurerecorded
actually 3). Figuretemperatures
3 shows a comparison
during the third
between thethat
heatwave temperature
took place modeling
between forMarch
Zone 1 1st
using
andtemperatures
March 3rdfrom the reference
in order sta- the model
to validate
Eng. Proc. 2023, 39, 80 tion and the actual measured temperatures. 5 of 7
and verify the accuracy of the results (Figure 3). Figure 3 shows a comparison between the
temperature modeling for Zone 1 using temperatures from the reference station and the
actual measured temperatures.

50
reference temperature in HW3
Simulated temperature
Measured temperature
40
Temperature (°C)

30

20

10

0
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72

Time(h)

Figure 3.3.Comparison
Figure Comparisonof real andand
of real modeled temperature
modeled profilesprofiles
temperature for zone for
1 in zone
Temuco
1 inbased on the
Temuco based on the
temperatures of the Maquehue station (HW3) located outside the city. Date: 1–3 March 2020.
temperatures of the Maquehue station (HW3) located outside the city. Date: 1–3 March 2020.
Figure 3 shows that the values obtained from the modeling are very close to the real
temperature profiles that occurred during that same heat wave event. Here, it was deter-
mined that the average real and modeled temperature differences in the 72-h period
reached 1 °C and the correlation coefficient was 0.98. This strong correlation suggests it is
possible to determine the temperature in the different zones of the city from the tempera-
Eng. Proc. 2023, 39, 80 5 of 6

Figure 3 shows that the values obtained from the modeling are very close to the
real temperature profiles that occurred during that same heat wave event. Here, it was
determined that the average real and modeled temperature differences in the 72-h period
reached 1 ◦ C and the correlation coefficient was 0.98. This strong correlation suggests it is
possible to determine the temperature in the different zones of the city from the temperature
of the weather station located outside the city (Maquehue). This would also imply that it is
possible to know the past profiles and make a prediction for future heat wave events based
on climate change scenarios. Nevertheless, additional studies are needed to verify these
hypotheses and, in that sense, they represent a continuation of the work presented here.
In this study, different equations were developed to predict temperatures in each zone
of the city and for each hour. Since each equation is specific to its respective zone and
cannot be transferred to other zones, it follows that the modeling of heat islands cannot
be generalized into a single predictive model for heat islands. Instead, each zone must be
studied within its unique microclimate.
The study strongly emphasizes the critical importance of considering the distinct
microclimatic characteristics and built environment of individual cities when modeling
urban heat islands. The complex interplay among various factors, such as urban morphol-
ogy, land cover properties, and anthropogenic heat emissions, requires the adoption of
customized modeling approaches to accurately represent the phenomenon of heat islands.
By acknowledging these factors and employing tailored modeling techniques, a more
precise representation of heat island phenomena can be achieved. This highlights the need
for site-specific analyses and modeling in order to understand and mitigate the impact of
heat islands in urban environments.

4. Conclusions
The results of this study demonstrate the close alignment between the modeled and
actual temperature profiles during the heatwave event, highlighting their significance in
predicting heat islands. Notably, the average temperature differences between the real and
modeled data over the 72 h period were only 1 ◦ C, with a high correlation coefficient of 0.98.
This strong correlation suggests the feasibility of estimating temperatures in various city
zones based on measurements from the weather station located outside the city (Maquehue).
The implications of accurate heat island predictions are substantial, as they directly impact
public health and energy consumption. Furthermore, these findings open the possibility of
retrospectively analyzing past temperature profiles and forecasting future heatwave events
under different climate change scenarios. However, further studies are required to validate
these hypotheses, representing a natural continuation of the research presented here.

Author Contributions: Conceptualization and methodology, A.M.-S.; software, A.M.-S.; formal anal-
ysis, A.M.-S. and J.F.; investigation, A.M.-S.; resources, A.M.-S.; writing—original draft preparation,
A.M.-S. and J.F.; writing—review and editing, A.M.-S. and J.F.; supervision, A.Z.; funding acquisition,
A.M.-S. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are openly available in FigShare at
https://doi.org/10.6084/m9.figshare.20186816.v1.
Conflicts of Interest: The authors declare no conflict of interest.
Eng. Proc. 2023, 39, 80 6 of 6

References
1. García-Cueto, O.R.; Jáuregui-Ostos, E.; Toudert, D.; Tejeda-Martinez, A. Detection of the Urban Heat Island in Mexicali, BC,
México and Its Relationship with Land Use. Atmosfera 2007, 20, 111–131.
2. Kuznetsova, I.N.; Brusova, N.E.; Nakhaev, M.I. Moscow Urban Heat Island: Detection, Boundaries, and Variability. Russ. Meteorol.
Hydrol. 2017, 42, 305–313. [CrossRef]
3. Feinberg, A. Urban Heat Island Amplification Estimates on Global Warming Using an Albedo Model. SN Appl. Sci. 2020, 2, 2178.
[CrossRef]
4. Liu, Z.; Zhan, W.; Bechtel, B.; Voogt, J.; Lai, J.; Chakraborty, T.; Wang, Z.H.; Li, M.; Huang, F.; Lee, X. Surface Warming in Global
Cities Is Substantially More Rapid than in Rural Background Areas. Commun. Earth Environ. 2022, 3, 219. [CrossRef]
5. Narumi, D.; Levinson, R.; Shimoda, Y. Effect of Urban Heat Island and Global Warming Countermeasures on Heat Release and
Carbon Dioxide Emissions from a Detached House. Atmosphere 2021, 12, 572. [CrossRef]
6. Santamouris, M.; Cartalis, C.; Synnefa, A.; Kolokotsa, D. On the Impact of Urban Heat Island and Global Warming on the Power
Demand and Electricity Consumption of Buildings—A Review. Energy Build. 2015, 98, 119–124. [CrossRef]
7. Robine, J.M.; Cheung, S.L.K.; Le Roy, S.; Van Oyen, H.; Griffiths, C.; Michel, J.P.; Herrmann, F.R. Death Toll Exceeded 70,000 in
Europe during the Summer of 2003. Comptes Rendus-Biol. 2008, 331, 171–178. [CrossRef] [PubMed]
8. Hashim, N.M.; Ahmad, A.; Abdullah, M. Mapping Urban Heat Island Phenomenon: Remote Sensing Approach. J.-Inst. Eng.
2007, 68, 25–30.
9. Elmarakby, E.; Khalifa, M.; Elshater, A.; Afifi, S. Tailored Methods for Mapping Urban Heat Islands in Greater Cairo Region. Ain
Shams Eng. J. 2022, 13, 101545. [CrossRef]
10. Abrar, R.; Sarkar, S.K.; Nishtha, K.T.; Talukdar, S.; Shahfahad; Rahman, A.; Islam, A.R.M.T.; Mosavi, A. Assessing the Spatial
Mapping of Heat Vulnerability under Urban Heat Island (UHI) Effect in the Dhaka Metropolitan Area. Sustainability 2022, 14,
4945. [CrossRef]
11. Kopecká, M.; Szatmári, D.; Holec, J.; Feranec, J. Urban Heat Island Modelling Based on MUKLIMO: Examples from Slovakia.
AGILE GIScience Ser. 2021, 2, 5. [CrossRef]
12. Hafner, J.; Kidder, S.Q. Urban Heat Island Modeling in Conjunction with Satellite-Derived Surface/Soil Parameters. J. Appl.
Meteorol. 1999, 38, 448–465. [CrossRef]
13. Voelkel, J.; Shandas, V. Towards Systematic Prediction of Urban Heat Islands: Grounding Measurements, Assessing Modeling
Techniques. Climate 2017, 5, 41. [CrossRef]
14. Wang, K.; Aktas, Y.D.; Stocker, J.; Carruthers, D.; Hunt, J.; Malki-Epshtein, L. Urban Heat Island Modelling of a Tropical City:
Case of Kuala Lumpur. Geosci. Lett. 2019, 6, 4. [CrossRef]
15. Dorigon, L.P.; Amorim, M.C.d.C.T. Spatial Modeling of an Urban Brazilian Heat Island in a Tropical Continental Climate. Urban
Clim. 2019, 28, 100461. [CrossRef]
16. Xu, M.; Bruelisauer, M.; Berger, M. Development of a New Urban Heat Island Modeling Tool: Kent Vale Case Study. Procedia
Comput. Sci. 2017, 108, 225–234. [CrossRef]
17. Kubilay, A.; Allegrini, J.; Strebel, D.; Zhao, Y.; Derome, D.; Carmeliet, J. Advancement in Urban Climate Modelling at Local Scale:
Urban Heat Island Mitigation and Building Cooling Demand. Atmosphere 2020, 11, 1313. [CrossRef]
18. Garzón, J.; Molina, I.; Velasco, J.; Calabia, A. A Remote Sensing Approach for Surface Urban Heat Island Modeling in a Tropical
Colombian City Using Regression Analysis and Machine Learning Algorithms. Remote Sens. 2021, 13, 4256. [CrossRef]
19. Khan, A.; Chatterjee, S.; Wang, Y. Urban Heat Island Modeling for Tropical Climates; Elsevier: Amsterdam, The Netherlands, 2020.
20. Kim, S.W.; Brown, R.D. Urban Heat Island (UHI) Intensity and Magnitude Estimations: A Systematic Literature Review. Sci. Total
Environ. 2021, 779, 146389. [CrossRef] [PubMed]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.