construction has gradually replaced the original surface types,
such as natural vegetation and wetland, thereby resulting
in the gradual increase of the regional mean land surface
temperature (
LST
). Furthermore, the spatial distribution of the
UHI
area is substantially concentrated. The results show that
surface temperature is positively correlated with the density
distribution of urban buildings. Moreover, large building den-
sity results in a substantial heat island effect. Therefore, the
rational control and improvement of an urban building layout
is conducive in reducing the heat island effect.
This research is based on 3S (Remote Sensing, Global
Position System, Geographic Information System) technology,
high precision map, and remote sensing data using ArcGIS
®
10.5 and ENVI 5.1 software and taking Guangzhou as the ex-
ample. First, the temperature and humidity of the heat island
effect is simulated and analyzed using the data on land use
from 2000 and 2009 and considering the high-resolution
WRF
model, which was developed by the National Atmospheric
Research Center (
NCAR
). Furthermore, the quantitative method
is used to detect the relationship between building density
and heat island effect and their time-varying characteristics
and mechanisms. The current study shows that the inten-
sity of the heat island increases with the development of the
urbanization process (Lang
et al
., 2016a; Sobstyl
et al
., 2018).
The heat island intensity during summer is substantially high-
er than that during winter. The relationship between building
density and heat island strength indicates that building densi-
ty has an important effect on heat island strength. This study
also emphasizes that the influence of urban building density
distribution on the
UHI
effect can provide a scientific basis for
urban planning and thermal environment management.
Research Area and Data
Research Area
As the research object in this study, Guangzhou is in southern
China, the northern edge of the Pearl River Delta, and brink of
China’s South Sea across Hong Kong and Macao. Meanwhile,
Guangzhou is also called China’s “South Gate” Metropoli-
tan area and is the starting point of Silk Road. In the past 20
years, the urban built-up area has expanded from 553.5 km
2
in 2002 to 1237.25 km
2
in 2015 because of the rapid progress
of urbanization and expansion. Furthermore, urban buildings
continue to upgrade and the heat island effect is also increas-
ing because of the rapid development of urban construction
(Lang
et al
., 2016b). Therefore, investigating the relationship
between building density and heat island intensity in the
main urban area of Guangzhou is necessary. The research area
(see Figure 1) includes the districts of Liwan, Yuexiu, Tianhe,
and Haizhu, which are surrounded by the Pearl River and
have a total land area of 376.37 km
2
.
Data
This research uses the spatial distribution data of the urban
building density and the remote sensing data of the land use
change model to simulate the heat island effect. Building
density is known as the building coefficient, which denotes
the ratio of the base area to the land area of all buildings in
a certain range of land use. Building density is an important
indicator of land intensity and can also reflect the vacancy
rate. The urban building footprint spatial distribution data of
this study is obtained by crawling from the Baidu map, which
contains the shape of each building within the city (see Figure
2a). The building density is shown in Figure 2b.
The global land cover data, namely, GLC2000 and
GLC2009, are obtained from the Systeme Probatoire
d’Observation de la Terre (SPOT) satellites and Envisat satel-
lites in 2000 and 2009, respectively. These data have a spatial
resolution of 1 km and 300 m, respectively, and the classifi-
cation criteria are from the land cover classification system
(LCCS) (see Figure 3).
Figure 1. Research area of Guangzhou
550
September 2018
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
