论文信息

Francisco Roco / Pamela Smith

Discover Cities

论文解读

AI 辅助整理
方法
2013–2021年Landsat影像与Mavic、Phantom 4无人机热红
发现
紧凑住宅区、工业区及裸地最热;温度与天空视域因子正相关,与NDVI和建筑高宽比负相关。
机会点
可借鉴其多尺度遥感与形态因子耦合方法,研究城乡绿地与建筑布局对微气候的调控,支撑气候敏感型规划。

论文摘要

城市化进程伴随着城市气候行为的改变,而这又表现为与城市及其环境特征相关的温度时空变化。在此背景下,城市形态与植被是高温时期缓解微气候影响的一种策略。本研究旨在探究圣地亚哥市西北部伦卡区住宅区夏季地表温度与城市设计特征之间的关系。利用夏季LANDSAT卫星影像(2013—2021年)以及热红外与多光谱无人机(分别为Mavic Enterprise Dual和Phantom 4),获取了局地与微尺度地表温度的详细信息,其中Landsat刻画了局地气候区的平均热行为,无人机观测则解析了代表性住宅场景内部的精细尺度变异,从而使两套数据集对同一现象提供了互补的时空视角。在局地尺度上,LANDSAT结果表明气候区与地表温度相关。更为紧凑的住宅区、工业区以及铺装或裸露土壤区域温度最高。在微尺度上,通过与环境地表温度数据相关联的城市与环境描述指标,对一系列城市场景进行了评估。所得结果有助于识别解释微尺度气候行为的因素。这些因素包括温度与天空视域因子之间的正相关关系,以及与植被(NDVI)和高宽比之间的负相关关系。整个公社的地表温度范围为13 °C至47 °C,且无人机反演的平均温度(40.3 °C)与Landsat反演的平均温度(33.6 °C)之间相差约7 °C。这些发现为气候敏感型城市设计与规划导则提供了可操作的输入依据。

全文

Discover Cities

Discover Cities
RESEARCH Open Access
Roco and Smith Discover Cities (2026) 3:234
https://doi.org/10.1007/s44327-026-00423-9
Study of micro-scale urban climatology
integrating remote sensing as an opportunity
for sensitive planning in the Chilean
Mediterranean area
Francisco Roco1 and Pamela Smith2*

  • C o r r e s p o n d e n c e :
    Pamela Smith
    pamelasmit@uchilefau.cl
    1Geographer, Santiago, Chile
    2Department of Geography, Centre
    for Climate Science and Resilience
    (CR)2, University of Chile, Santiago,
    Chile
    © The Author(s) 2026. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
    License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate
    credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material.
    You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party
    material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material
    is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted
    use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit h t t p : / / c r e a t i v e c o m m o n s . o r g / l i c e n s
    e s / b y - n c - n d / 4 . 0 / .
    Abstract
    The process of urbanization is accompanied by alterations in the urban climate
    behavior, and this in turn is characterized by spatial-temporal variations in
    temperature that are related to the urban and environmental characteristics of the
    city. In this context, urban morphology and vegetation are a strategy for reducing
    the impact of microclimate during periods of high temperature. This study aimed
    to investigate the relationship between summer surface temperature and urban
    design characteristics in residential areas of the Renca district, which is in the
    northwest of the city of Santiago. The use of summer LANDSAT satellite imagery
    (2013–2021) and thermal and multispectral UAVs (Mavic enterprise dual and
    Phantom 4, respectively) provided detailed information about surface temperature
    at local and micro scale, with Landsat characterising the average thermal behaviour
    of local climate zones and UAV observations resolving fine-scale variability within
    representative residential scenarios, so that the two datasets offer complementary
    spatial and temporal perspectives on the same phenomenon. At local scale, the
    LANDSAT results show that climate zones are related to surface temperature. More
    compact residential areas, industrial zones and areas with paved or bare soils are
    the warmest. At micro-scale a series of urban scenarios were evaluated by means
    of urban and environmental descriptors that were related to surface temperature
    data. The results obtained have contributed to the identification of factors that
    explain climate behavior at the micro-scale. These factors include a direct relationship
    between temperature and sky view factor, and an inverse relationship with respect to
    vegetation (NDVI) and height-to-width ratio. Surface temperatures ranged from 13 °C
    to 47 °C across the commune, and a difference of about 7 °C was observed between
    UAV-derived (40.3 °C) and Landsat-derived (33.6 °C) average temperatures. These
    findings provide actionable inputs for climate-sensitive urban design and planning
    guidelines.
    Keywords Microclimate, Urban morphology, Urban vegetation, Residential areas

Page 2 of 23
Roco and Smith Discover Cities (2026) 3:234
1 Introduction
The study of the city becomes relevant when considering the growing process of global
urbanization and its associated phenomena. According to the United Nations (UN, [ 1]),
more than 81% of the population in Latin America and the Caribbean lived in urban
areas. In parallel, the Chilean cities are home to 87.8% of the national population (INE,
[2]) and the Metropolitan Region concentrates 60% of the total.
The process of urbanization exerts an influence on climate behavior. Within urban
areas, altered climatic conditions are expressed in changes to the spatial variability of
parameters such as temperature, which manifest as vertical and horizontal differences
at the surface and atmospheric levels [ 3], due to surface contrasts and the interaction of
radiative and advective exchanges generated by elements with different energy balances
[4]. As a result of this phenomenon, the urban climate is characterized by the occur -
rence of urban heat islands (UHI), which result in a greater degree of warming in urban
areas relative to their non-urban surroundings. The UHI phenomenon is of particular
concern because of the increase in global temperatures. In their study, Milner et al., [ 5]
examined the projected increase in temperature in temperate and mid- and high-lati -
tude cities.
The study of the urban climate encompasses various scales of spatial analysis. At each
scale, the behaviour of meteorological parameters, such as atmospheric temperature or
wind, is explained by different environmental and urban factors. The micro-scale cor -
responds to the immediate neighbourhood scale, covering a few hundred metres and
developing in the UCL (Urban Canopy Layer), between the ground and the upper limit
of buildings [3], that is, the space where people live, act and suffer the grievances of the
city’s problems (Gómez and Ferrer, 2010). The importance of this scale, from the point of
view of geographical climatology, is that it allows us to evaluate the relationship between
urban morphology and microclimatic conditions and, in turn, to understand the rela -
tionship between these and society, through the evaluation of pedestrian thermal com -
fort, providing input for urban planning [6].
In this context, studying micro-scale allows for more precise and detailed research.
Surface temperature is one of the parameters involved in microclimate studies and influ-
ences air temperature in the lower atmosphere, surface energy balance, indoor climates,
and energy exchanges important to inhabitants [7].
Surface temperature information can be obtained from medium- or low-resolution
satellite imagery [ 8], using Landsat or Sentinel products, among other LST-monitoring
satellite platforms currently available. However, tools capable of acquiring more detailed
information, such as unmanned aerial vehicles (UAVs), are important to complement the
study of surface temperature. In this context, Guzmán , [ 9] highlights the versatility of
UAVs and thermal cameras in scientific studies due to their autonomy, while Song et
al., [10] recognize their usefulness in analyzing and comparing the urban physical and
thermal environments in high resolution. Various studies involving the use of UAVs have
been conducted, revealing the advantages and disadvantages of collecting and analyzing
microclimate or surface features [8, 10–17].
In terms of urban design strategies, proposals include managing geometry, using dif -
ferent materials and incorporating urban vegetation to mitigate high temperatures. For
instance, Zölch et al., [ 18] identified the potential of strategically planting trees and
installing green façades to reduce heat stress at pedestrian level. Meanwhile, Alchapar

Page 3 of 23
Roco and Smith Discover Cities (2026) 3:234
and Correa , [ 19] focus on selecting materials that achieve passive cooling, highlight -
ing the factors that influence their reflective and thermal behavior. Thus, it is evident
that climatic phenomena are more complex than the urban heat island, which is a broad
expression of the alterations associated with urbanization. There are also other scales at
which the microclimate responds to different factors present in the urban fabric, which
is of interest to decision-makers in urban design and planning. Several studies support
the importance of urban morphology parameters, with varying degrees of correlation
between geometry and microclimate [15, 20–24].
Regarding urban design strategies aimed at improving the microclimate, Mills et al.,
[25] note that the effects of decisions made at different scales are mutually dependent.
In other words, planning or urban design measures can either reinforce or negate each
other’s effects. The scales are also integrated with each other: the scale of properties
can determine characteristics at the whole level, while blocks determine characteristics
at the neighborhood scale. Ideally, this should be considered when implementing the
respective strategies.
On the other hand, it is advisable to ensure that measures aimed at dealing with warm
and cold seasons are compatible, as well as prioritizing strategies attentive to the local
climate. For example, in the temperate climate context of Santiago, Palme and Carrasco
[26] recommend greening measures and cool materials for building façades. The recom -
mended measures also depend on the analyzed microclimatic parameter, i.e. strategies
aimed at reducing air temperature will not necessarily be the same as those aimed at
reducing surface temperature.
This study was designed to analyze the behavior of summer surface temperature
within representative residential areas of the two most important local climate zones in
the study area. The temperature data were related to different urban and environmental
parameters like the vegetation and the morphoclimatic conditions of each urban canyon.
For this purpose, the study uses high-resolution data, enabling a level of detail consistent
with the micro-scale to be achieved in climatic and urban analysis. This can inform plan-
ning and climate-sensitive design guidelines, with the commune of Renca serving as a
case study representing an inner mediterranean urban area located in a Latin American
metropolitan region.
2 Materials and methods
The study adopted a multi-scale approach to urban climatology, progressing from the
local to the micro-scale. For the local-scale analysis, the entire district of Renca was con-
sidered. Located northwest of Santiago in the Metropolitan Region, it is situated at coor-
dinates 33°24′00″ South and 70°44′00″ West. It covers an area of 24 km² (BCN, [27]) and
is inhabited by 143,622 people (INE, [ 28]). Sarricolea et al., [ 29] point out that Renca is
one of the warmest districts of Santiago in terms of the average spatial distribution of
surface heat islands during the day and night. For the micro-scale analysis, two represen-
tative residential areas were selected. Two flights were carried out using a UAV, which
allowed the surface and its temperature to be recorded on a sub-metric scale. This also
enabled seven urban scenarios to be recognized and delimited at the level of the whole
and of individual properties (Fig. 1). Figure 2 presents a schematic of the methods used
at the local and microscale, which are described below.

Page 4 of 23
Roco and Smith Discover Cities (2026) 3:234
Fig. 1 Flight zones and urban scenarios

Fig. 2 Methodological workflow summarizing the local-scale (Landsat) and micro-scale (UAV) data collection and
processing steps, and their integration into the comparative analysis

2.1 Analysis at local level
To characterize the local climate of the district, we considered three criteria, two of
them related to urban morphology, the urban grid and local climatic zones, and the third
to characterize the local urban climate, the average summer surface temperature. The
urban grid typology was based on the proposal of Giraldo-Ospina and Vásquez-Varela

Page 5 of 23
Roco and Smith Discover Cities (2026) 3:234
Table 1 Typology of the urban fabric for the commune of Renca
Reticular Weft Regular reticular With a layout that configures grids and presents uniformity in the blocks.
Irregular
reticular
With a layout that presents continuity through orientations adapted to the
conditions of the environment, configured by blocks of irregular morphology.
Non-grid
weft
Irregular With a discontinuous layout that does not form a grid and presents diversity
in the morphology of its blocks.
Organic With an asymmetrical and sinuous layout, without order and apparently with
a spontaneous development. It has shapes adapted to the topography.
Linear It develops along a road without being configured around a center.
Dead-end streets It has dendritic or branched characteristics, with no interconnection between
streets and no block configuration.
Source: Own elaboration, based on Giraldo-Ospina and Vásquez-Varela [ 30]
Fig. 3 Local climatic zones for the commune of Renca.Source: Smith et al. [31], based on Stewart and Oke [32]

[30] and adapted for the communal level according to the road layout and the shape of
the blocks (see Table 1).
For local climate zones, we used the layer developed by Smith et al., [ 31], based on
the Stewart and Oke [ 32] classification, which recognizes 17 categories: 10 urban and 7
natural (Fig. 3).
The diurnal surface temperature data for the summer season (December-February)
average were calculated from the Landsat 8 satellite between 2013 and 2021 in the
Google Earth Engine software. This satellite passes over the study area around midday
every 16 days, so an average of six images per year were used to calculate the average.
The thermal band has a spatial resolution of 100 m but was rescaled to 30 m (resampling
only interpolates pixel values to a finer grid and does not increase the original 100  m
thermal information content).
2.2 Microscale analysis
The local-level analysis previously allowed representative residential areas of the com -
mune to be selected, considering the following criteria:
a. These areas should represent regular and irregular grid patterns (Table 1). Morphologies
without a road grid, associated with the topography of the hills, were excluded to
minimize possible anomalies and topographic effects of slope on temperature; and.

Page 6 of 23
Roco and Smith Discover Cities (2026) 3:234
Fig. 4 Satellite average summer surface temperature (°C) 2013–2021.Source: Own elaboration, based on Smith
et al. [38]

Table 2 Urban-residential scenarios
Flying zone Urban residential scenario Local climate zone
Zone n°1 (local warm zone) A B3 (compact low-rise)
B
Zone n°2 (local cold zone) C
D
E
F B2 (Compact mid-rise)
G
b. These areas should represent the predominant urban residential local climate zones:
compact low-rise (B3) and compact mid-rise (B2).
Considering the above, two zones with a difference of about 2 °C in average surface tem-
perature were selected: the warm zone, located to the south-east and corresponding to
zone n°1; and the cold zone, located to the south-west and corresponding to zone n°2
(Fig. 4).
Within each selected area, seven urban scenarios (A–G) were defined. Each scenario
includes a street with its flanking blocks (Table  2). These correspond to urban designs
representative of the residential fabric. Of the 7 scenarios studied, 2 were in the south -
eastern sector (A, B) and 5 in the southwestern sector of the commune (C, D, E, F, G).
Of the 7 cases studied, 2 belong to the high-density, medium-rise zone (B2) and 5 to the
high-density, low-rise zone (B3).
With the aim of identifying the relationship between surface temperature and differ -
ent types of urban design, the following micro-scale variables were obtained for each
scenario considered: (i) Surface temperature, (ii) normalized difference vegetation index
(NDVI), (iii) Sky-view factor (SVF) and (iv) height-to-width ratio (H/W).

Page 7 of 23
Roco and Smith Discover Cities (2026) 3:234
To collect microscale data, unmanned aerial vehicle (UAV) flights were conducted to
obtain very high-resolution images (sub-meter pixels). The Mavic 2 Enterprise Dual was
used for thermal and visible imaging. According to the manufacturer [ 33], its thermal
camera has an uncooled VOx microbolometer sensor (160 × 120 pixels, 12 μm pixel
pitch, 57° horizontal field of view) sensitive to the 8–14 μm long-wave thermal infra -
red (TIR) band (see also [ 34]), and it is complemented by a 12-megapixel visible-light
camera. The manufacturer reports a typical accuracy of up to ± 5% in high-gain mode
(scene range − 10 to 140 °C) and up to ± 10% in low-gain mode, and does not specify the
thermal sensitivity (NETD) of this model; measurement error also increases with atmo -
spheric temperature and humidity, distance to the target and low surface emissivity [33].
The thermal imagery was processed as colour-coded (RGB) orthomosaics rather than
as per-pixel radiometric files, so surface temperatures were retrieved with the empirical
calibration described below, and the uncertainty of the thermal measurements is char -
acterised by the standard errors of that calibration (2.4 °C and 1.8 °C) rather than by an
independent radiometric assessment. A Phantom 4 drone equipped with a multispectral
camera was used to measure the normalized difference vegetation index (NDVI). The
flights were planned using the DJI Pilot application at an altitude of 130 m: for flight
zone 1, from the Renca Municipal Stadium on 4 January 2022 at 14:45, and for flight
zone 2 from the Jorge Alessandri Rodríguez School on the same date at 12:15. Both
flights were conducted under clear sky conditions, with wind speeds below 3 m/s and
relative humidity below 40%, consistent with typical summer days in Santiago’s Medi -
terranean climate. These conditions minimize atmospheric interference on the thermal
measurements. According to the nearest meteorological station (Pudahuel), air tempera-
ture was around 18–20 °C during the first flight and 23–25 °C during the second, with a
maximum of 30 °C that day.
Regarding the processing of the orthophotos obtained, Agisoft Metashape software
was used to construct, in parallel, a true color orthophoto mosaic, a thermal orthophoto
mosaic and an NDVI orthophoto mosaic. Digital surface models (DSMs) were obtained
from the true color orthophoto mosaic and used to calculate morphoclimatic urban
descriptors (SVF, H/W ratio).
In the case of the thermal orthophoto mosaic, the orthophoto mosaics were trans -
ferred to ArcGIS Pro software for georeferencing the images, establishing control points
with RGB information, and converting qualitative thermal data (RGB) to surface tem -
perature in degrees Celsius (°C). This was done using the methodology applied by Smith
et al. [8], which is based on constructing multiple linear regression equations: Eq. (1)

𝑇𝑇𝑇∘C)= Intersection coefficient

  • 𝑇Red-band ×𝑅𝑅 -coefficient)
  • 𝑇Green-band ×𝐺𝐺 -coefficient)
  • 𝑇Blue-band ×𝐵𝐵 -coefficient)
    (1)
    The RGB values were extracted from the color-coded thermal orthophoto mosaic at
    22 calibration points in zone 1 and 25 in zone 2. These points were established along
    the flight paths and matched with the surface temperature (°C) shown on the drone’s
    live thermal display, which was captured through synchronized screen recordings; the
    RGB values at the same coordinates were tabulated to fit the multiple linear regression
    in Excel. The equation for Zone 1 had an R² value of 0.96 and a standard error of 2.43 °C,

Page 8 of 23
Roco and Smith Discover Cities (2026) 3:234
while the equation for Zone 2 had an R² value of 0.97 and a standard error of 1.8  °C.
No independent validation dataset separate from these calibration points was used; the
reported R² (0.96 / 0.97) and standard error (2.43  °C / 1.8  °C) for Zones 1 and 2 are
therefore in-sample fit statistics rather than out-of-sample RMSE. No separate bias term
was computed beyond the regression’s own intercept and coefficients.
Eq. (2) Equation for conversion to Celsius degrees (°C) for zone 1

𝑇𝑇1𝑇∘C) = 14.910
+𝑇 Red-band × 0.1469)
+𝑇 Green-band × 0.0097)
+𝑇 Blue-band × 0.1105)
(2)
Eq. (3). Equation for conversion to Celsius degrees (°C) for zone 2

𝑇𝑇𝑇∘C) = 17.798

  • 𝑇Red-band × 0.1195)
  • 𝑇Green-band × 0.0028)
  • 𝑇Blue-band × 0.0762)
    (3)
    These equations were then applied in the ArcGIS Pro raster calculator to replace the
    band information and intersection coefficients. Considering the time difference between
    flight 1 and 2 (two and a half hours approximately), the last step was the time correction.
    For this purpose, midday was considered the standard time, and the start and end zones
    were taken twice on each flight, calculating the temperature difference at each point
    and then estimating the change per minute. The surface temperature obtained was then
    associated with each defined urban scenario.
    The normalized difference vegetation index (NDVI) was obtained using the red and
    infrared bands from the multispectral UAV orthophoto mosaic in Agisoft Metashape
    software. NDVI data were also associated with each urban scenario.
    To calculate the urban descriptors corresponding to the Sky View Factor (SVF) and
    the height-to-width ratio (H/W), digital surface models (DSMs) were used. These were
    obtained by processing the true color orthophoto mosaic. A total of 374 control points
    were established in both flight zones (zones 1 and 2), for which the building heights (H)
    and street widths (W) were recorded. These values were then used to make calculations
    according to the formulas proposed by Sarricolea [ 35], based on the method of Bonan
    [36]. This process resulted in the calculation of a value in radians (Eq. 4), a light inci -
    dence angle in degrees (Eq. 5), and the value of the SVF (Eq. 6).
    Radians (rad) = tan−1 𝑇2𝐻𝐻𝐻𝐻𝐻) (4)
    Light incidence angle (
    ∘
    ) = Degrees [tan−1 𝑇2𝐻𝐻𝐻𝐻𝐻)] (5)
    Sky View Factor 𝑇Ψ) = cos [tan−1 𝑇2𝐻𝐻𝐻𝐻𝐻)] (6)
    Next, the control points were interpolated with the sky view factor (SVF) and height/
    width factor (H/W) values using ArcGIS Pro’s IDW tool. Once the seven urban scenar -
    ios had covered the area, the information for each scenario was extracted as a table using
    the zonal statistics tool in ArcGIS Pro.

Page 9 of 23
Roco and Smith Discover Cities (2026) 3:234
To recognise the general relationship between surface temperature and urban and
environmental descriptors (SVF, H/W and NDVI), the raster images were transformed
into points (9,000 in total). Then, the information was extracted for each point, including
surface temperature, in accordance with the methodology used by [37].
3 Results
3.1 Analysis of surface temperature behavior on a local scale
The average summer surface temperature of Renca ranges between 13  °C and 47  °C,
although the central sector of the municipality has a relatively homogeneous distribu -
tion. The eastern and western sectors show a more heterogeneous temperature distri -
bution. At a local scale, a relationship can be observed between the spatial patterns of
summer surface temperature and local climatic zones, particularly between natural and
vegetated areas and built-up areas (see Fig.  4). Thus, high temperatures in the western
sector are associated with zone B8 (large low-rise), reaching 34.8 °C, and with zone N6
(areas of bare soil or sand), reaching 37 °C. In the eastern sector, warm spots are associ -
ated with the industrial area, represented by zone B8 (large low-rise), while in the north-
ern sector, warm temperatures are associated with the hills, reaching 37.7 °C.
The coldest zones are the river at 32.8 °C and zone N4 (low plants), which is associ -
ated with parks and has an average temperature of 31 °C (Fig. 4). Residential zones range
from 33 to 34.5 °C and are associated with zones B2 (compact mid-rise) and B3 (compact
low-rise), respectively. LCZ B3 covers an area of 32.9% at the community level, making it
the largest, and although its variability does not exceed two degrees, it presents atypical
values, i.e., points where the temperature is outside the limits. Given that climate behav -
ior is more homogeneous for the same ZCL, a microscale analysis is used to explore the
phenomenon internally.
3.2 Analysis of surface temperature behavior at the microscale
At the microscale, surface temperature shows higher spatial heterogeneity related to
various urban factors such as urban geometry, pedestrian and vehicle traffic, and the
surface properties of materials. The two sectors selected for the micro-scale work are
located on the north riverside of the Mapocho River and are part of an irregular grid
layout with some dead-end passages. Zone No. 1 has rectangular and elongated blocks,
with some dead-end passages that may affect the circulation of local breezes. Zone No.
2, on the other hand, has heterogeneous and compact polygonal blocks. Table 3 provides
a characterization of seven urban scenarios in terms of housing type, orientation, mate -
riality and the amount of vegetation in public spaces. Five scenarios are associated with
the compact low-rise local climate zone (A, B, C, D and E), and two with the compact
mid-rise local climate zone (F and G).
Regarding the microscale surface temperature captured by the UAV, contrasts were
observed between and within different local climate zones. In terms of the comparison
between the different zones, the scenarios in the compact mid-rise LCZ (B2) (F and G)
averaged a surface temperature of 37.5 °C, while the scenarios in the compact low-rise
LCZ (B3) (A, B, C, D and E) averaged 42 °C, showing a difference of 4.5 °C. To confirm
the internal differences, consider scenarios A (flight zone 1) and C (flight zone 2), which
are both from the compact low rise (B3), yet show a 5 °C difference (Fig. 5).

Page 10 of 23
Roco and Smith Discover Cities (2026) 3:234
Flying
zone
Local
climate
zone
Urban
Scenario
Description Field photography
Flight
zone
N°1
(local
scale
warm
zone)
LCZ B3 A Located in the Placilla passage. It corresponds to a group of
two-store semi-detached and one-store terraced houses; its axis
is oriented approximately in a N-S direction. The predominant
materiality of the dwelling is brick masonry, with an aged zinc roof.
Regarding the roads, there is only a pavement for pedestrians with
no roadway for vehicles, and in the middle line there are some trees.
It is also possible to find trees and shrub vegetation in the interior of
the property.
B Located in Rayen Quitral Street. It is adjacent to Costanera Norte
Highway, being the closest to the Mapocho River, its axis is oriented
approximately in a NW-SE direction. It corresponds to a group of
single-family dwellings and is the scenario with the largest plots.
Its configuration includes an adjacent square. There is no discern-
ible urban design pattern, and the materiality is varied but mainly
composed of stuccoes brick. It is characterized by tree and shrub
vegetation in interior courtyards.
Table 3 Description of urban residential scenarios

Page 11 of 23
Roco and Smith Discover Cities (2026) 3:234
Flying
zone
Local
climate
zone
Urban
Scenario
Description Field photography
Flight
zone
N°2
(local
scale
cold
zone)
C Located in La Rhea Passage. It forms part of a road grid with rectan-
gular blocks, and its axis is oriented in a W-E direction. It corresponds
to one-storey semi-detached single-family dwellings. Its materiality
is composed mainly of brick masonry, light material extensions with
lined partition walls and corrugated zinc roofing. In general, there is
little vegetation.
D Located in the Cariquima passage, it forms part of a rectangular
grid close to the motorway and the Mapocho River. The passage is
oriented in an N-S direction and is made up of two blocks of small
plots with two-store semi-detached single-family dwellings. It has
a brick masonry structure, burgundy-colored facades and a corru-
gated zinc roof. Its configuration includes an adjoining square with
some arboreal vegetation.
Table 3 (continued)

Page 12 of 23
Roco and Smith Discover Cities (2026) 3:234
Flying
zone
Local
climate
zone
Urban
Scenario
Description Field photography
E Located next to Topocalma Street. It corresponds to a closed
complex, located within the block. It is made up of three-store semi-
detached single-family dwellings with interior passages-oriented
N-S and W-E. Its materiality is composed of masonry with brick
blocks and partition walls lined on both sides. It has tree vegetation
concentrated in certain places.
LCZ B2 F Located at the intersection of Esmeralda and Topocalma streets, it
corresponds to a closed complex within the block, consisting of
four burgundy-colored three-store residential blocks, with passages
oriented in an E-W and N-S direction. Its materiality is made up of
masonry with brick blocks and corrugated zinc roofing. It has an
interior garden with trees.
Table 3 (continued)

Page 13 of 23
Roco and Smith Discover Cities (2026) 3:234
Flying
zone
Local
climate
zone
Urban
Scenario
Description Field photography
G Located on Topocalma Street. This scenario corresponds to a closed
complex within the block and is made up of two three-store hous-
ing blocks of burgundy color, separated by an interior passageway
oriented in an N-S direction. It is made of brick masonry and cor-
rugated zinc roofing. It has little vegetation and a small courtyard.
Table 3 (continued)

Page 14 of 23
Roco and Smith Discover Cities (2026) 3:234
Fig. 5 Surface temperature (°C) in urban-residential scenarios in the commune of Renca, Santiago.Source: Own
elaboration

The roof covering is one of the most decisive factors in determining temperatures, as
demonstrated by the high temperatures in scenarios A (53 °C) and E (52 °C), and the low
temperatures in the residential blocks of scenarios F (33  °C) and G (34  °C). However,
most urban designs are heterogeneous, and changes in the roughness of the residential
fabric lead to variations in the spatial temperature pattern. An example of this is scenario
C, where the roofs have different shades and orientations. Therefore, front extensions,
changes to the building line and roof coverings, and the presence of trees in front gar -
dens and courtyards all influence differences in surface temperature at a micro-scale.
Regarding the comparison between satellite surface temperature and that captured
by drone, the satellite temperature for summer 2013–2021 shows a smoother behavior
and a considerably lower average in most of the scenarios studied, while the microscale
temperature (unmanned aerial vehicle, drone) shows greater variability attributable to
extremes of high and low temperatures captured in the coverage of specific individual
elements that the satellite image does not reflect. The average surface temperature in
urban areas shows a difference of about 7  °C between the average captured with the
UAV (40.3  °C) and the satellite temperature (33.6  °C). Nevertheless, there is a direct
relationship between the two: when the surface temperature is higher in the satellite
image, it is also higher in the UAV image, with coinciding spatial patterns. Quantita -
tively, the scenario-level average temperatures (n = 7 residential scenarios) show a strong
positive correlation between the two datasets (Pearson r = 0.82, p = 0.024) (exploratory,
given the small sample), with a mean difference (UAV minus Landsat) of + 7.1  °C and
a root-mean-square difference of 7.4 °C (UAV warmer than Landsat in every scenario),
consistent with the ~ 7  °C difference reported above. These scenario-level values were
calculated from the seven pairs of average temperatures reported in Roco Castillo (2022)

Page 15 of 23
Roco and Smith Discover Cities (2026) 3:234
Fig. 6 Surface temperature according to SVF and H/W descriptors.Source: Own elaboration

and describe the observed difference between the two datasets under the present design,
not a sensor bias.
3.3 Relationship between surface temperature and urban and environmental descriptors
The sky view factor and the height/width ratio define scenarios involving urban canyons
of different depths. Deeper street profiles tend to be narrower, with a smaller area of
road potentially exposed to high temperatures. In contrast, more extended profiles have
a larger area of road potentially exposed to high temperatures. In this context, the SVF
showed a slight positive correlation, i.e. the higher the SVF, the higher the temperatures
tended to be. Conversely, the height-to-width ratio (H/W) exhibited the opposite behav-
ior: as the H/W ratio increased, the temperature decreased (Fig. 6). In other words, more
extended profiles or greater distances between façades allow for a greater number of
high values.
On the other hand, when the seven scenarios are analyzed individually (Fig.  7), no
direct relationship is observed between urban descriptors and average temperature;
rather, the average thermal values are multifactorial. Geometric aspects of urban mor -
phology and urban surfaces and elements, such as roof cover and the presence of trees,
play an important role. In this case, trees contribute to lowering temperatures, while

Page 16 of 23
Roco and Smith Discover Cities (2026) 3:234
Fig. 7 SVF, H/W and surface temperature per urban scenario.Source: Own elaboration

Fig. 8 Surface temperature according to NDVI.Source: Own elaboration

the material of roofs is also a factor. Urban geometry descriptors were most relevant for
public space surfaces, particularly the area occupied by streets.
Two main morphological patterns connected to the local climate zones were identi -
fied: the urban scenarios A, B, C, D and E (LCZ B3) have a higher SVF and a lower H/W
factor, while the scenarios F and G (LCZ B2) have a lower SVF and a higher H/W fac -
tor. For example, scenarios A and B had SVF values of 0.85 and 0.87 respectively, which
translate into a larger area potentially exposed at road level. At the other extreme, resi -
dential blocks F and G had SVF values of 0.59 and 0.52 respectively, which translates into
a smaller area potentially exposed at road level. Regarding the height-to-width (H/W)
factor, the highest H/W factors were found in areas with the lowest SVF.
In contrast, a negative or inverse relationship was observed between surface tempera -
ture and vegetation (NDVI): as NDVI increased, surface temperature decreased (Fig.  8).
Higher NDVI values at the property level were useful for indicating the presence of veg -
etation in front and backyard areas. High NDVI values in some street areas were associ -
ated with the presence of urban trees in public spaces.

Page 17 of 23
Roco and Smith Discover Cities (2026) 3:234
4 Discussion
Local climate zones were useful in guiding and contextualising climate analyses based
on the identification of urban characteristics. The results regarding their overall ther -
mal behaviour are consistent with previous research conducted in Mediterranean cities,
which found that denser urban typologies, industrial areas and uncovered or paved soils
were the warmest and that natural vegetation cover and scattered residential areas were
cooler [9, 31]. This reaffirms the usefulness of this approach for the climatic characteri -
sation of cities on a zonal scale. At the same time, recent work in Lecce and Milan shows
that, despite sharing the same LCZ classes, the thermal response of corresponding LCZs
can differ between cities depending on urban morphology, land-cover composition and
local climate [ 39]; the zone-level patterns found in Renca should therefore be verified
locally before being transferred to other cities. However, UAVs (unmanned aerial vehi -
cles, or drones) provide more detail, enabling more precise exploration and recognition
of internal contrasts within the same local climate zones.
Based on the multiscale analysis developed in the research, it is possible to demon -
strate the dynamics described by Voogt and Grimmond [ 40], where the heterogeneity of
the urban surface is accompanied by greater variations in surface temperature on a small
scale. In this sense, when the information from the satellite image and the drone was
compared, the results showed a correspondence between the local and micro scales for
both the warmest (zone 1) and coldest (zone 2) zones, these results are consistent with
those obtained in other inland Mediterranean cities in Chile [8].
The microclimatic scale enabled phenomena that the local and mesoscale scales could
not fully address to be explained. This approach revealed various factors influencing
surface temperature behavior, including street geometry, property and complex spatial
configuration, vegetation presence, and the thermal properties of different surfaces and
urban elements.
The spatial configuration of the different scenarios influences surface temperature
directly or indirectly. The SVF and the H/W factor determine whether profiles are more
compact or extended, and whether street fractions are exposed to radiation. This, more -
over, was found to influence the arrangement of urban vegetation and NDVI values, con-
sequently affecting surface temperatures.
Improving outdoor urban spaces can reduce temperatures and moderate microcli -
matic conditions in urban settings. According to [ 41], one such alternative is the addi -
tion of vegetation, which can be implemented at the planning stage and has been proven
to reduce urban temperatures [31, 42]. However, increasing vegetation in Mediterranean
cities presents the challenge of increased water demand [43–45], particularly since urban
green spaces traditionally incorporate introduced species [ 46, 47]. Therefore, improv -
ing outdoor spaces could involve incorporating more versatile species with lower water
requirements or species adapted to the location’s environmental conditions.
Authors such as Zölch et al., [ 18] have demonstrated that improvements in thermal
comfort are mainly achieved through trees and green façades, though they report that
green roofs have a lesser influence. While roofs have a limited impact at pedestrian level,
they are important in analysing the surface temperature of homes. This is evident in
scenarios A and E, where high temperatures were recorded, and in scenarios F and G,
where low temperatures were recorded. Roof coverings become more important in areas
with a larger built environment or higher housing density. Oke et al., [3] discuss different

Page 18 of 23
Roco and Smith Discover Cities (2026) 3:234
coatings for cool roof surfaces, which vary in terms of their emissivity and reflectivity. In
the scenarios studied, it was found that, at the building scale, maximum temperatures
were largely determined by roofs, while minimum temperatures were largely determined
by trees.
In this context, Mills et al., [ 25] caution that the effects of decisions made at different
scales are mutually dependent. In other words, planning measures can either reinforce
or negate each other’s effects. Similarly, measures designed to address conditions in both
warm and cold seasons should be reconciled. Finally, smaller scales contribute to higher-
level scales. In this sense, the characteristics of individual properties can determine those
at the neighborhood or block level. Complementary microclimate simulation tools, such
as ENVI-met, could be used in future work alongside the remote-sensing approach
adopted here to further test and generalise these relationships. Open-source thermal-
comfort tools such as the QSSI (QGIS Summer Simmer Index) calculator [ 48] could
likewise complement this workflow, and similar summer-simmer-index approaches have
also been applied to assess climate suitability for outdoor activities under future climate
scenarios [49].
The study demonstrates that the self-construction of housing and its subsequent
expansion to accommodate new family members is a recurring process in low- and
middle-income neighbourhoods in Latin America. The modifications and improvements
made to homes by residents complicate the morphology of the urban canyon, affecting
ventilation, shade and solar radiation. This alters the original patterns and creates differ -
ences between neighbouring properties, increasing the variety of urban roughness and
roof and wall material composition. Some of these changes can be observed in Scenario
C, where many of the dwellings have been modified, thereby affecting the climate of the
surrounding outdoor space.
A methodologically relevant result of this study is the difference observed between
the average surface temperature captured by the UAV (40.3 °C) and that derived from
Landsat satellite imagery (33.6 °C), representing a gap of approximately 7 °C across the
urban scenarios studied. This value should be read as the observed difference between
the two datasets under the present experimental design, not as a sensor bias, because
it combines several components that cannot be separated with the available data: spa -
tial resolution and representativeness, temporal and meteorological differences, and
uncertainty in the UAV temperature retrieval. Regarding spatial resolution, the Land -
sat thermal band has a native resolution of 100 m (resampled to a 30 m grid), so each
thermal value integrates heterogeneous surface types — including vegetated areas, roof -
tops, streets and bare soil — which smooths the thermal signal and can reduce localised
temperature extremes. Regarding time, the satellite values are a multi-year (2013–2021)
summer climatology of around-midday overpasses, whereas the UAV data were acquired
on a single day (4 January 2022) at 12:15 and 14:45, when air temperature at the near -
est station was about 18–20 °C and 23–25 °C; a single clear-sky summer day need not
coincide with the multi-year average. Finally, the UAV temperatures were retrieved with
an empirical RGB-to-temperature regression whose standard errors (2.4 °C and 1.8 °C)
were estimated in-sample. In contrast, UAV-derived imagery captures surface tempera -
ture at sub-metre resolution, resolving individual urban elements and their distinct ther-
mal properties.

Page 19 of 23
Roco and Smith Discover Cities (2026) 3:234
While both data sources show spatial correspondence in their relative thermal pat -
terns — warmer zones in the satellite image tend to remain warmer at micro-scale — the
magnitude of the difference is not uniform (it ranged from 3.2 °C in scenario F to 9.8 °C
in scenario. It can be hypothesised that the gap is larger over built impervious surfaces,
where UAV imagery records peak temperatures well above the satellite average, than
over vegetated areas, where the two datasets may converge; this was not tested formally
in the present study. This finding has implications for the use of Landsat data in micro-
scale urban climate studies: at a native thermal resolution of 100  m, satellite-derived
temperatures may not capture the localised surface temperatures resolved by UAV imag-
ery, and both refer to surface rather than pedestrian-level thermal conditions. Future
research should quantify this difference by surface type using UAV and satellite obser -
vations acquired on the same day and time and validate the UAV temperature retrieval
against independent reference measurements, before considering correction factors to
improve the applicability of satellite thermal data at the neighbourhood and block scale.
Although the case study focuses on Renca, the multiscale workflow presented here,
combining local climate zone classification, satellite thermal climatology and UAV-based
micro-scale observation, is transferable to other Latin American and Mediterranean-
climate cities with similar residential morphologies; the specific temperature thresh -
olds and regression coefficients, however, would need to be recalibrated for each study
area, as illustrated by comparable microclimate and thermal-comfort studies in other
Mediterranean-climate and semi-arid settings [ 50]. Building-scale geo-information and
machine-learning approaches could further support this kind of transferability assess -
ment across cities [ 51]. Recent multi-city statistical modelling that combines morpho -
logical and biophysical descriptors, such as impervious surface fraction, leaf area index
and albedo [20], offers a complementary way to test which of the relationships identified
here hold across cities.
Some limitations of the study should be considered when interpreting the results:
i. Surface temperature was considered due to the difficulty of obtaining atmospheric
temperature information that would allow analysis at the defined spatial and temporal
scales. Surface temperature is related to air temperature behaviour but does not
reflect the actual microclimates present.
ii. There are differences between surface temperatures captured by satellites and those
calculated from UAV images. The former corresponds to an average, while the latter
corresponds to a summer day. In addition, although this was corrected, the UAV
flights were carried out at different times in both areas, close to the satellite’s passage.
In addition, the UAV temperatures were derived from an empirical calibration
validated only in-sample (Sect.  2.2), so their uncertainty was not independently
quantified.
iii. The present study is based on seven urban residential scenarios, which, while
representative of the predominant local climate zones in the study area, constitutes
a reduced sample that limits the statistical generalizability of the findings. The
relationships identified between surface temperature and urban descriptors (SVF,
H/W and NDVI) should therefore be interpreted as exploratory. Future research
should expand the number of scenarios analysed, ideally encompassing a broader
range of urban morphologies and local climate zones across different districts or

Page 20 of 23
Roco and Smith Discover Cities (2026) 3:234
cities with similar Mediterranean climatic conditions, to validate and strengthen the
identified relationships.
iv. Finally, the detailed analysis of spatial determinants was restricted to a limited
number of parameters. After analysing surface temperatures in residential areas, it
was possible to identify the components responsible for thermal behaviour. These
included the geometry of urban canyons, the spatial configuration of properties and
blocks, and the thermal properties of the surface itself. The study also considered the
number of dwellings and tree cover; however, the results were not included as the
former was not related to surface temperature, and the latter exhibited behaviour like
that observed when analysing the NDVI.
5 Conclusions
Spatial analysis of surface temperature at different scales enables us to identify differ -
ences between and within the various urban and environmental typologies represented
by the residential scenarios evaluated. In this regard, satellite imagery reveals homo -
geneous behaviour at neighbourhood and block levels related to local climate zones,
whereas micro-scale temperature shows a heterogeneous distribution, demonstrating
the advantages of using unmanned aerial vehicles.
The results demonstrate the importance of urban morphology in surface temperature
behaviour. This is evaluated through the sky visibility factor and the height-to-width
ratio, as differences in these parameters explain heat gain and cooling conditions.
Regarding residential scenarios, designs that favour high daytime temperatures in
summer were found at the level of individual properties and/or blocks of flats. Scenario
E is an example of a mixed condition in which high roof temperatures can coexist with
equally unfavourable conditions at block level with respect to the urban fabric. Scenario
E presented high temperatures at property level and is in a network of dead-end interior
passages that hinder surface ventilation.
Characteristic morphological situations were also found at street intersections, open
spaces and avenues, which had a higher sky visibility factor and lower height-to-width
ratio, leading to greater surface areas exposed to radiation during the day.
At the micro-scale, the effects due to the thermal properties of different coverings were
evident. For instance, differences could be observed between roofs and sheds (warmer)
and gardens and interior backyards (cooler), a pattern also evident in the presence of
trees and bodies of water.
It was found that, in addition to being linked by roads, the level between properties
and blocks constitutes an interesting interface. While the property level relates to possi -
ble interventions by residents, the block level is represented by the outdoor public space,
where urban planning has a greater impact and capacity for intervention.
Finally, the variables affecting surface temperature behavior can be considered within
the criteria for climate-sensitive design and planning. This approach contributes to the
study of urban characteristics in residential areas and the development of strategies to
cope with daytime heat, thereby improving environmental conditions for inhabitants.
Abbreviations
UHI Urban heat island
UAV Unmanned aerial vehicles
NDVI Normalized difference vegetation index
SVF Sky view factor

Page 21 of 23
Roco and Smith Discover Cities (2026) 3:234
H/W Height-to-width ratio
LCZ Local climate zone
Acknowledgements
This work was supported by the Center for Climate and Resilience Research (CR2), based at the University of Chile and
funded through the Actividades de Interés Nacional (AIN) program.
Author contributions
Francisco Roco: Conceptualization, methodology, software, formal analysis, investigation, writ-ing—original draft
preparation, visualization.Pamela Smith: Conceptualization, methodology, validation, formal analysis, investigation,
resources, writing—original draft preparation, writing—review and editing, supervision, project administration, funding
acquisition.
Funding
This work was supported by the Center for Climate and Resilience Research (CR2), funded by the Chilean Ministry of
Education through the Actividades de Interés Nacional (AIN) program, hosted at the University of Chile.
Data availability
All data supporting the findings of this study are available within the article. Details regarding microclimatic data and the
analysis of satellite and UAV imagery are available to interested parties upon reasonable request to the corresponding
author.
Declarations
Ethical approval
Not applicable. The study did not involve human participants or the collection of any data from individuals. It used only
remote-sensing data (Landsat and UAV imagery of buildings and public space) and anonymized secondary information
derived from the 2017 Population and Housing Census (INE); ethical approval was therefore not required.
Consent to participate
Not applicable — the study did not involve human research participants; only anonymized census-derived and remote-
sensing data were used.
Consent for publication
Not applicable. The manuscript does not include data, images or videos relating to individual persons.
Competing interests
The authors declare no competing interests.
Received: 6 May 2026 / Accepted: 11 November 2026
References

  1. United Nations, Department of Economic and Social Affairs, Population Division. World urbanization prospects: the 2018
    revision [Internet]. 2018 [cited 2026 Nov 12]. https://population.un.org/wup/DataQuery/.
  2. INE. Results of the 2017 population and housing census [Internet]. 2017. http://resultados.censo2017.cl/.
  3. Oke TR, Mills G, Christen A, Voogt JA. Urban Climates. Cambridge: Cambridge University Press; 2017. h t t p s : / / d o i . o r g / 1 0 . 1 0 1
    7 / 9 7 8 1 1 3 9 0 1 6 4 7 6 .
  4. Arnfield AJ. Two decades of urban climate research: a review of turbulence, exchanges of energy and water, and the
    urban heat island. Int J Climatol. 2003;23:1–26. https://doi.org/10.1002/joc.859.
  5. Milner J, Harpham C, Taylor J, Davies M, Le Quéré C, Haines A, Wilkinson P . The challenge of urban heat exposure under cli-
    mate change: An analysis of cities in the Sustainable Healthy Urban Environments (SHUE) database. Climate. 2017;5(4):93.
    https://doi.org/10.3390/cli5040093.
  6. Smith P , Lamarca C, Henríquez C. A comparative study of thermal comfort in public spaces in the cities of Concepción and
    Chillán, Chile. In: Henríquez C, Romero H, editors. Urban Climate in Latin American Cities. Springer Nature; 2019. Chap. 6.
    https://doi.org/10.1007/978-3-319-97013-4_6.
  7. Voogt JA, Oke TR. Thermal remote sensing of urban climates. Remote Sens Environ. 2003;86(3):370–84. h t t p s : / / d o i . o r g / 1 0 .
    1 0 1 6 / S 0 0 3 4 - 4 2 5 7 ( 0 3 ) 0 0 0 7 9 - 8 .
  8. Smith P , Sarricolea P , Peralta O, Aguila JP , Thomas F. Study of the urban microclimate using thermal UAV. The case of the
    mid-sized cities of Arica (arid) and Curicó (Mediterranean), Chile. Build Environ. 2021;206:108372. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j .
    b u i l d e n v . 2 0 2 1 . 1 0 8 3 7 2 .
  9. Guzmán G. Analysis of the influence of urban design on the meteorology of the Aburrá Valley. National University of
    Colombia; 2018. https://rep ositorio.un al.edu.co/h andle/un al/69423.
  10. Song B, Park K, Kim SH, Park G. Comparison of an unmaned aerial vehicle based physical environment with thermal prop-
    erties from in-situ measurements: Campus of Changwon National University, South Korea. Volume 98. Sustainable Cities
    and Society; 2023. p. 104836. https://doi.org/10.1016/j.scs.2023.104836.
  11. Ahmad J, Eisma JA. Capturing Small-Scale Surface Temperature Variation across Diverse Urban Land Uses with a Small
    Unmanned Aerial Vehicle. Remote Sens. 2023;15(8):2042. https://doi.org/10.3390/rs15082042.
  12. Cho Y-I, Yoon D, Lee M-J. Comparative Analysis of Urban Heat Island Cooling Strategies According to Spatial and Temporal
    Conditions Using Unmanned Aerial Vehicles (UAV) Observation. Appl Sci. 2023;13(18):10052. h t t p s : / / d o i . o r g / 1 0 . 3 3 9 0 / a p p 1
    3 1 8 1 0 0 5 2 .

Page 22 of 23
Roco and Smith Discover Cities (2026) 3:234
13. Fuladlu K. Environmental parameters for campus outdoor space: A microclimate analysis of the Eastern Mediterranean
University (EMU) campus. J Green Building. 2021;16(3):217–36. https://doi.org/10.3992/jgb.16.3.217.
14. Gaitani N, Burud I, Thiis T, Santamouris M. High-resolution spectral mapping of urban thermal properties with Unmanned
Aerial Vehicles. Build Environ. 2017;121:215–24. https://doi .org/10.101 6/j.builden v.2017.0 5.027.
15. Kim J, Lee DK, Brown RD, Kim S, Kim JH, Sung S. The effect of extremely low sky view factor on land surface temperatures
in urban residential areas. Sustainable Cities Soc. 2022;80:103799. https://doi.org/10.1016/j.scs.2022.103799.
16. Malbéteau Y, Parkes S, Aragon B, Rosas J, McCabe MF. Capturing the Diurnal Cycle of Land Surface Temperature Using an
Unmanned Aerial Vehicle. Remote Sens. 2018;10(9):1407. https://doi.org/10.3390/rs10091407.
17. Naughton J, McDonald W. Evaluating the Variability of Urban Land Surface Temperatures Using Drone Observations.
Remote Sens. 2019;11(14):1722. https://doi.org/10.3390/rs11141722.
18. Zölch T, Maderspacher J, Wamsler C, Pauleit S. Using Green infrastructure for urban climate-proofing: an evaluation of heat
mitigation measures at the micro-scale. Volume 20. Urban Forestry & Urban Greening; 2016. pp. 305–16. h t t p s : / / d o i . o r g / 1 0
. 1 0 1 6 / j . u f u g . 2 0 1 6 . 0 9 . 0 1 1 .
19. Alchapar N, Correa E. Solar reflectance of opaque city envelopes and its effect on urban temperatures. Building Rep.
2015;67(540):e112. https://doi.org/10.3989/ic.14.131.
20. Esposito A, Pappaccogli G, Bozzeda F, Buccolieri R. A multi-city statistical modelling of surface urban heat island: Applica-
tion to Italian cities. Urban Clim. 2025;64:102717. https://doi.org/10.1016/j.uclim.2025.102717.
21. Li Y, Ouyang W, Yin S, Tan Z, Ren C. Microclimate and its influencing factors in residential public spaces during heat waves:
An empirical study in Hong Kong. Build Environ. 2023;236:110225. https://doi .org/10.101 6/j.builden v.2023.1 10225.
22. Tsitoura M, Michailidou M, Tsoutsos T. Achieving sustainability through the management of microclimate parameters in
Mediterranean urban environments during summer. Sustainable Cities Soc. 2016;26:48–64. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . s c s . 2 0
1 6 . 0 5 . 0 0 6 .
23. Yang X, Li Y. The impact of building density and building height heterogeneity on average urban albedo and street
surface temperature. Build Environ. 2015;90:146–56. https://doi .org/10.101 6/j.builden v.2015.0 3.037.
24. Yang J, Shi Q, Menenti M, Sing Wong M, Wu Z, Zhao Q, Abbas S, Xu Y. Observing the impact of urban morphology and
building geometry on thermal environment by high spatial resolution thermal images. Volume 39. Urban Climate; 2021. p.
100937. https://doi.org/10.1016/j.uclim.2021.100937.
25. Mills G, Cleugh H, Emmanuel R, Endlicher W, Erell E, Mcgranahan G, Ng E, Nickson A, Rosenthal J, Steemers K. Climate Infor-
mation for Improved Planning and Management of Mega Cities (Needs Perspective). Procedia Environ Sci. 2010;1:228–46.
https://doi.org/10.1016/j.proenv.2010.09.015.
26. Palme M, Carrasco C. Chapter 12 - Urban heat island in Latin American cities: a review of trends, impacts, and mitiga-
tion strategies, Editor(s). Ansar Khan, Hashem Akbari, Francesco Fiorito, Sk Mithun, Dev Niyogi. Global Urban Heat Island
Mitigation. Elsevier; 2022. pp. 251–67. https://doi .org/10.101 6/B978-0-32 3-85539- 6.00014-7.
27. Biblioteca del Congreso Nacional de Chile (BCN). Renca: communal report 2024 [Internet]. 2024 [cited 2026 Nov 12]. Avail-
able from: https://www .bcn.cl/sii t/reportesc omunales /comunas_v.html?anno=2024&idcom=13128.
28. Instituto Nacional de Estadísticas (INE). Censo de población y vivienda 2024: resultados [Internet]. 2024. h t t p s : / / c e n s o 2 0 2 4
. i n e . g o b . c l / r e s u l t a d o s / .
29. Sarricolea P , Smith P , Romero-Aravena H, Serrano-Notivoli R, Fuentealba M, Meseguer-Ruiz O. Socioeconomic inequalities
and the surface heat island distribution in Santiago, Chile. Sci Total Environ Volume. 2022;832:155152. h t t p s : / / d o i . o r g / 1 0 . 1 0
1 6 / j . s c i t o t e n v . 2 0 2 2 . 1 5 5 1 5 2 .
30. Giraldo-Ospina T, Vásquez-Varela LR. Characterization of urban plots in the city of Manizales, Colombia (1849–2017).
Revista de Arquitectura (Bogota). 2020;22(1):30–43. https://doi.org/10.14718/RevArq.2020.2669.
31. Smith P , Peralta O, Sarricolea P , Thomas F, Meseguer-Ruiz O. Climate-sensitive planning. Opportunities through the study of
LCZs in Chile. Build Environ. 2023;242:110444. https://doi .org/10.101 6/j.builden v.2023.1 10444.
32. Stewart ID, Oke TR. Local Climate Zones for Urban Temperature Studies. Bull Am Meteorol Soc. 2012;93(12):1879–900.
https://doi.org/10.1175/BAMS-D-11-00019.1.
33. DJI. Mavic 2 Enterprise: specs (M2ED thermal camera) [Internet]. DJI Support [cited 2026 Nov 12]. Available from: h t t p s : / / w
w w . d j i . c o m / s u p p o r t / p r o d u c t / m a v i c - 2 - e n t e r p r i s e ( a c c e s s e d 28 September 2026).
34. Messina G, Modica G. Applications of UAV Thermal Imagery in Precision Agriculture: State of the Art and Future Research
Outlook. Remote Sens. 2020;12:1491. https://doi.org/10.3390/rs12091491.
35. Sarricolea P . The urban surface heat island and its conditioning factors: the case of the metropolitan area of Santiago de
Chile [doctoral thesis]. Barcelona: University of Barcelona; 2012. http://hdl.handle.net/2445/41987.
36. Bonan GB. Ecological climatology: Concepts and applications. England: Cambridge University Press; 2002.
37. Aguila J. Evidence of the relationships between urban morphologies and meso- and microscale surface temperature.
Some contributions to territorial planning. The case of Curicó, Chile. University of Chile; 2020. h t t p s : / / r e p o s i t o r i o . u c h i l e . c l /
h a n d l e / 2 2 5 0 / 1 8 0 7 6 6 .
38. Smith P , Blanco E, Sarricolea P , Peralta O, Thomas F. Urban climate simulation model to support climate-sensitive planning
decision making at local scale. Journal of Urban Management; 2024.
39. Pappaccogli G, Esposito A, Buccolieri R. Summer diurnal LST variability across local climate zones using ECOSTRESS data in
Lecce and Milan. Atmosphere. 2025;16(4):377. https://doi.org/10.3390/atmos16040377.
40. Voogt JA, Grimmond CSB. Modeling surface sensible heat flux using surface radiative temperatures in a simple urban area.
J Appl Meteorol. 2000;39(10):1679–99. https://doi.org/10.1175/1520-0450-39.10.1679.
41. Rizwan A, Dennis L, Liu C. A review on the generation, determination and mitigation of Urban Heat Island. J Environ Sci.
2008;20(1):120–8.
42. de la Ochoa J. Ciudad vegetación e impacto climático. El confort en los espacios urbanos. Barcelona: Erasmus ediciones;
2009. p. 184.
43. Jundiani J, Fadli M, Qurbani ID, Kurniaty R, Hulwanullah H. Urban green space regulation: Challenges to water resources
conservation in Indonesia and Australia. J Hum Rights Cult Legal Syst. 2024;4(1):169–93.
44. Nouri H, Borujeni SC, Hoekstra AY. The blue water footprint of urban green spaces: An example for Adelaide, Australia.
Landsc urban Plann. 2019;190:103613.
45. Ramaiah M, Avtar R. Daily Water Requirements of Vegetation in the Urban Green Spaces in the City of Panaji, India. Water.
2025;17(10):1487. https://doi.org/10.3390/w17101487.

Page 23 of 23
Roco and Smith Discover Cities (2026) 3:234
46. Del Castillo M, Castillo C. Aproximación bioclimática para el diseño de espacios públicos, análisis inicial en distintas plazas
chilenas. Arquitectura y Urbanismo. 2014;XXXV(3):69–82.
47. Gaertner M, Wilson JRU, Cadotte MW, et al. Non-native species in urban environments: patterns, processes, impacts and
challenges. Biol Invasions. 2017;19:3461–9. https://doi.org/10.1007/s10530-017-1598-7.
48. Adiguzel F, et al. QSSI (QGIS summer simmer index) calculator plugin: an open-source tool for thermal comfort analysis in
GIS applications. Int J Biometeorol. 2026a;70(4):99.
49. Adiguzel AD, Zhang M, Kaya AY, Karadeniz ENES, Mantas S. Projecting tourism climate suitability in Türkiye under climate
change scenarios: insights from the summer simmer index. Advances in Space Research; 2026b.
50. Zhang M, Yiğit İ, Adigüzel F, Hu C, Chen E, Siyavuş AE, Kaya AY. Impact of urban surfaces on microclimatic conditions and
thermal comfort in Burdur. Türkiye Atmos. 2024;15(11):1375.
51. Kaya AY, Sajjad M, Dedekorkut-Howes A. Downscaling urban disaster resilience: toward building-scale geo-information
modeling and machine learning-assisted resilience knowledge systems. Trans GIS. 2025;29(8):e70153. h t t p s : / / d o i . o r g / 1 0 . 1
1 1 1 / t g i s . 7 0 1 5 3
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

论文原文与出版信息

Discover Cities
查看论文原文

继续探索