Response of Daytime Changes in Temperature and Humidity to Three-Dimensional Urban Morphology in Subtropical Residential Districts

Abstract

The combination of global climate change and the urban heat island effect has given rise to a deterioration in the livability of residential districts within cities, posing challenges to enhancing the health quality of urban environments. Meanwhile, the intensification of daytime changes in temperature and humidity in residential districts has rendered the sensory representation of the urban heat island effect more pronounced. This study selects the residential districts in Fuzhou City as the research case area, which have witnessed a discernible warming trend in recent years, and acquires temperature and humidity parameter data at three time periods (early morning, noon, and evening) to represent the daytime temperature and humidity change phase. Through aerial photography and field research, three types of spatial morphological indicators (buildings I, vegetation II, and the combination of buildings and vegetation II) of residential districts are quantified to represent the three-dimensional spatial form of the case study area. The analysis results show the following: ➀ Residential districts experience two phases of daytime changes in temperature and humidity: a warming and drying phase (WDP) in the morning and a cooling and humidifying phase (CHP) in the afternoon. The characteristics of changes in temperature and humidity show a spatial correlation with each other. ➁ The impact of urban three-dimensional morphology on changes in temperature and humidity in WDP is minor, whereas, in CHP, it is influenced by Class II and Class III indicators. The two types of urban morphology exert a synergistic regulatory effect on changes in temperature and humidity. ➂ Vegetation has a significant regulatory effect on temperature and humidity variations in residential areas through changes in its three-dimensional form. Enlarging the area of individual trees while reducing their canopy volume can restrain the warming and dehumidification of residential districts and promote cooling and humidification. In contrast to only planting trees, a vegetation configuration combining trees, shrubs, and grass can bring a more obvious cooling effect to residential districts. The research results can provide a reference for urban planners in the planning and design of residential areas as well as the optimization and improvement of urban living environments.

1. Introduction

The urban heat island effect causes a decline in the livability of urban residential areas, and the efforts to improve the quality of urban environmental health are faced with difficulties [1]. At the same time, the intensification of daytime changes in temperature and humidity in residential districts makes the somatosensory representation of the urban heat island effect more obvious. Studies have pointed out that changes in temperature may affect the spread of infectious diseases [2] and are also closely related to the incidence of cardiovascular and respiratory diseases [3]. Therefore, maintaining a relatively mild climate and reducing drastic temperature and humidity changes can help improve the livability of residential areas. Furthermore, exploring the mechanisms for urban changes in temperature and humidity has practical application value. Currently, scholars are studying how urban form regulates diurnal temperature and humidity variations [4,5], but the diurnal changes in temperature and humidity are greatly affected by solar radiation [6]. During the daytime, the temperature and humidity changes are relatively less affected by solar radiation. However, there is a lack of research focusing on the changes in temperature and humidity at different stages of the daytime. At the same time, urban residents’ outdoor activities are more frequent during the daytime. Therefore, it is of great significance to focus on daytime changes in temperature and humidity.

In terms of research scale, most studies on the changes in temperature and humidity focus on a global scale [7] or regional scale [8], and relatively few studies focus on the changes in temperature and humidity in different functional areas within cities. Studies have demonstrated that regional climate variations significantly influence thermal comfort. The impact of temperature and humidity differs not only across regions but also varies seasonally within the same region [9,10]. Some studies have shown that lowering temperatures and increasing humidity can promote thermal comfort in the climate during summer in subtropical regions. As an important place for the formation of the urban heat island effect, the street shape and layout of urban residential districts have a direct influence on the changes in temperature and humidity. Therefore, narrowing the research scale down to the street level of cities and paying particular attention to the changes in temperature and humidity in residential areas of subtropical regions will help us more accurately reveal the influence mechanism of urban morphology on the changes in temperature and humidity. This choice holds practical significance.

Aiming at the regulation of the changes in temperature and humidity in urban areas, some scholars have proposed such urban form optimization paths as rational planning of land development [11], a scientific increase in blue and green space [12,13], and scientific evaluation of energy-saving buildings [14,15]. Compared with the high cost of infrastructure improvement [16] and the long period of reducing anthropogenic heat sources, urban form adjustment is more universal and practical, and easier to implement [17]. Currently, research on the impact of urban form on the changes in temperature and humidity has largely focused on two-dimensional landscape patterns, such as land use, surface coverage, and green space ratios. Previous studies have confirmed that the three-dimensional urban form is closely related to the urban wind and thermal environment [18,19]. Therefore, fine adjustment of urban three-dimensional spatial forms is helpful to cope with the urban heat island effect and adjust changes in temperature and humidity.

As important components of urban spatial form, urban buildings and vegetation, respectively, as heat sources and cold sources, have a significant impact on urban temperature. Many scholars have studied the relationship between physical building form characteristics, such as building height, surface area and volume, and surface temperature in urban areas. Among them, building height and floor area ratio and other indicators are increasingly applied in heat island effect research because they can be directly applied to planning and design [20,21]. Therefore, the three-dimensional form of the building plays a significant role in regulating the changes in temperature and humidity. Regulating vegetation forms is recognized as a nature-based solution for cooling cities. Compared with two-dimensional morphological indicators, three-dimensional morphological indicators of urban vegetation can more comprehensively reflect the volume and spatial structure of vegetation [22]. Some studies have pointed out that the amount of canopy green affects the ecological benefits of vegetation [23,24]. Tree height and canopy diameter have also been proven to have significant effects on temperature and humidity variations in urban areas [25]. However, there are relatively few quantitative studies on the synergistic effects of vegetation and building 3D forms on temperature and humidity change at small scales. Therefore, it is of practical significance to quantify the three-dimensional form of urban vegetation and the combination of buildings and vegetation and study the mechanism of regulating the changes in temperature and humidity.

In view of the above research gaps, this study uses aerial images and field research results to quantify three types of spatial form indicators, named architecture (Class I), vegetation (Class II), and the combination of architecture and vegetation (Class III), to represent the three-dimensional urban form of the residential districts. This study investigates the changes in temperature and humidity at different stages of the daytime in urban residential districts and examines their response to urban morphological characteristics. The study identifies the key categories of urban 3D forms that affect the changes in temperature and humidity during the daytime and compares the separate and synergistic effects of different categories of 3D forms on the changes in temperature and humidity during the daytime. In addition, this study identified key three-dimensional morphological indicators that regulate changes in temperature and humidity during the daytime and explored their underlying mechanisms. By employing multiple statistical methods, the study conducted an integrated analysis of the three-dimensional forms of buildings and vegetation to investigate how urban morphology influences changes in temperature and humidity at different stages of the daytime so as to promote the improvement of residential planning and realize the continuous improvement of the heat island effect.

2. Methodology

2.1. Study Area

This study selected Fuzhou City, Fujian Province, China, as the research area. This region has a subtropical monsoon climate, with a prominent heat island effect. The influence range of heat island exposure is gradually expanding [26]. The spring season is marked by a significant daily average temperature difference of 9 °C (Figure 1a). This study selects the residential districts of Fuzhou as the research area. In 2023, Fuzhou City was conferred the first Global Sustainable Development City Award, being one of only five cities worldwide to obtain this distinction [27]. This indicates that enhancing urban environmental health and quality is a significant pursuit for Fuzhou and that the heat environment issue requires effective addressing.

Numerous factors influence changes in temperature and humidity in urban areas, such as land cover types and human-made heat sources [28]. To minimize the interference from these factors, this study opts for residential districts with consistent land cover types as the study area. Some studies have shown that, compared with open low-rise areas, the morphological characteristics of intensive high-rise residential areas have a stronger impact on the heat island effect [19], so this study arranged the sample sites in dense high-rise housing areas. In the study area, the following steps were carried out: Select sidewalks of urban arterial roads with similar width and uniform distribution and arrange the measurement route. Set appropriate numbers of sampling points along each route and ensure that the sampling points are on the same side of the sidewalk and are spaced 200 m apart. Adjust the location of the sampling points according to the land use to reduce the influence of surface characteristics, traffic volume, and population density on the temperature and humidity of the sampling points, and finally, obtain 29 sampling points (Figure 1b). Combined with the refined requirements of current urban planning and existing research results [10,29], buffer zones with a radius of 50 m are established to calculate relevant statistical indicators (Figure 1c–e).

Figure 1. Study area and the distribution map of measurement points (modified from reference [29]): (a) location map of the study area; (b) distribution of sample points in the study area; (c) schematic of the 50 m buffer zones around sample points; (d) extraction of three-dimensional building morphological information at sample points; (e) extraction of three-dimensional vegetation morphological information at sample points.

Figure 1. Study area and the distribution map of measurement points (modified from reference [29]): (a) location map of the study area; (b) distribution of sample points in the study area; (c) schematic of the 50 m buffer zones around sample points; (d) extraction of three-dimensional building morphological information at sample points; (e) extraction of three-dimensional vegetation morphological information at sample points.

2.2. Acquisition of Temperature and Humidity

This study used the method of fixed-point mobile measurement to determine the temperature and humidity in residential districts by using a temperature and humidity recorder (model: TOP Cloud-agri TPJ-20-LG). The fixed-point mobile measurement method involves measuring the temperature and humidity at selected sample points and then moving to the next sample point in sequence to obtain temperature and humidity data. The specific operation is as follows: Volunteers used the temperature and humidity recorder at 29 measurement points in six groups to measure and record the temperature and humidity at a height of 1.5 m. Measurements were conducted at each point along the road. Upon completion of the initial measurements, a return trip was made to conduct subsequent measurements. The average value of the round-trip data was calculated. The observation time was from 8:00 to 9:00 (T1), 13:00 to 14:00 (T2), and 16:00 to 17:00 (T3) on May 18, 19, and 20, 2022. The weather for the three days had temperatures ranging from 18 to 28 °C, with light winds. The average changes in temperature and humidity across the three time phases reflect the daytime changes in temperature (DT) and humidity (DH). The warming and drying phase (WDP) from T1 to T2 was characterized by a maximum temperature increase of 4.53 °C and a maximum humidity decrease of 7.81%. The cooling and humidifying phase (CHP) from T2 to T3 exhibited a maximum temperature decrease of 2.77 °C alongside a maximum humidity increase of 5.72%. Additionally, the overall daytime phase (ODP) from T1 to T3 demonstrated a maximum temperature increase of 3.57 °C and a maximum humidity reduction of 7.64%. Based on the measured data (DT and DH) of changes in temperature and humidity at different phases during the daytime at sample points, the spatial distribution maps of changes in temperature and humidity at different phases during the daytime were drawn through the spatial interpolation tool in ArcGIS Pro. The observed characteristics in the changes in temperature and humidity indicated spatial correlations among them (Figure 2).

2.3. Translation: Urban 3D Morphology Measurement

This study acquires high-resolution digital orthophoto maps (DOMs), digital surface models (DSMs), and LiDAR point cloud data for the research area through aerial photography conducted by uncrewed aerial vehicles (UAVs). The ArcGIS Pro 3.0.1 software platform is utilized to establish a three-dimensional spatial model of the research area, enabling the quantification of various Class I indicators. For measuring vegetation indicators, this study references the findings of Jianhua Zhou [30] and Di Chen [23]. Using the remote sensing technology of UAV, the three-dimensional shape index data of trees around the measuring point were obtained by simulating the three-dimensional volume with the plane quantity. Firstly, a variety of segmentation schemes are used to obtain the best crown segmentation and improve the accuracy of crown width measurement. Following the methodologies outlined by Yao Qiu and Yi He [31], preliminary segmentation of tree crowns is performed using the ESP tool in eCognition Developer 9.0 software. Then, based on the criteria of a vegetation difference index (VDVI) > 0.04 and a normalized digital surface model (nDSM) > 2, visual interpretation and mask processing were used to correct and optimize the tree crown segmentation results, and to minimize the potential errors in the calculation of other tree form indicators. These results are then imported into ArcGIS Pro. In conjunction with the research of Biaojun Ji [32], correlation equations relating crown diameter to crown height along with equations for quantifying tree crown green quantity are employed to quantify Class II indicators. Research by Parmehr [33] shows that photogrammetry based on drone point cloud and liDAR measurements has a high correlation with canopy parameters such as crown width, with an R² value greater than 95%. This can provide a reference for the basic accuracy of canopy width data of low-altitude high-resolution aerial images in this study. The selection of Class III indicators is based on the three-dimensional spatial pattern. A panoramic camera is used to take fisheye images of the sample points at a height of 1.2 m, and the sky view factor (SVF) is calculated using the pixel points of the images [34]. In addition, the selection of indicators also relies on the basic attributes of the three-dimensional space (height and volume), which are derived from the urban three-dimensional model containing buildings and vegetation and quantified according to the established calculation formulas (

Table 1. Calculation of urban three-dimensional morphological indicators.

Category	Indicator	Calculation Formula	Meaning
Architecture 3D Morphology (Class I)	MBH [35] Mean Building Height	$\mathrm{M}\mathrm{B}\mathrm{H}={\displaystyle \sum _{i=1}^n}{\mathrm{H}}_i/n$	MBH represents the average height of buildings within the buffer zone. Hi is the height of the i-th building within the buffer zone, and n is the number of buildings within the buffer zone.
	BV [36] Building Volume	$\mathrm{B}\mathrm{V}={\displaystyle \sum _{i=1}^n}{\mathrm{V}}_i$	BV represents the total volume of buildings within the buffer zone, and Vi is the volume of the i-th building within the buffer zone.
	FAR [37] Floor Area Ratio	$\mathrm{F}\mathrm{A}\mathrm{R}={\displaystyle \raisebox{1ex}{${\sum }_{i=1}^n{\mathrm{S}}_i$}\!\left/ \!\raisebox{-1ex}{$\mathrm{A}$}\right.}$	FAR represents the floor area ratio within the buffer zone. Si is the building area of the i-th building within the buffer zone, and A is the land area of the buffer zone.
	SCD [38] Space Crowding Degree	$\mathrm{S}\mathrm{C}\mathrm{D}={\displaystyle \frac{{\sum }_{i=1}^n{\mathrm{V}}_i}{{\mathrm{H}}_{max}\times \mathrm{A}}}\times 100\%$	SCD represents the spatial congestion degree within the buffer zone. Vi is the volume of the i-th building within the buffer zone, Hmax is the maximum height of the buildings within the buffer zone, and A is the land area occupied by the buffer zone.
	BSI [39] Building Structure Index	$\mathrm{B}\mathrm{S}\mathrm{I}={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n}{\displaystyle \frac{{\mathrm{F}}_i}{{\mathrm{H}}_i}}$	BSI represents the Building Structure Index within the buffer zone. Fi is the ground area occupied by the i-th building within the buffer zone, and Hi is the height of the i-th building within the buffer zone.
Vegetation 3D Morphology (Class II)	MTCV [29] Mean Tree Crown Volume	$\mathrm{M}\mathrm{T}\mathrm{C}\mathrm{V}={\displaystyle \raisebox{1ex}{${\sum }_{i=1}^n{\mathrm{V}}_i$}\!\left/ \!\raisebox{-1ex}{$n$}\right.}$	MTCV represents the average green mass of individual tree crowns within the buffer zone. Vi is the volume of the i-th tree crown within the buffer zone, and n is the number of trees within the buffer zone.
	SGV [29] Shrub and Grass Volume	$\mathrm{S}\mathrm{G}\mathrm{V}={\displaystyle \sum _{i=1}^n}{\mathrm{V}}_i+{\displaystyle \sum _{j=1}^k}{v}_j$	SGV represents the total green mass of herbaceous and shrub plants within the buffer zone. Vi is the volume of herbaceous plants on the i-th green space within the buffer zone, and vj is the volume of the j-th shrub within the buffer zone. The n and k are the number of trees and shrubs within the buffer zone.
	GV [40] Green Volume	$\mathrm{G}\mathrm{V}={\displaystyle \sum _{i=1}^n}{\mathrm{T}}_i+{\displaystyle \sum _{j=1}^k}{\mathrm{S}}_j+{\displaystyle \sum _{r=1}^h}{\mathrm{G}}_r$	GV represents the total green mass of trees, shrubs, and herbaceous plants within the buffer zone. Ti is the crown volume of the i-th tree within the buffer zone, Sj is the volume of the j-th shrub within the buffer zone, Gr is the volume of the r-th green space within the buffer zone, and n, k, and h are the number of trees, shrubs, and green spaces within the buffer zone, respectively.
	TCD [22] Tree Crown density	$\mathrm{T}\mathrm{C}\mathrm{D}={\displaystyle \raisebox{1ex}{$\sum _{i=1}^n{\mathrm{S}}_i$}\!\left/ \!\raisebox{-1ex}{$\mathrm{A}$}\right.}$	TCD represents the tree crown density within the buffer zone. Si is the crown cover area of the i-th tree within the buffer zone, and A is the land area of the buffer zone.
	MTCA [41] Mean Tree Crown Area	$\mathrm{M}\mathrm{T}\mathrm{C}\mathrm{A}={\displaystyle \raisebox{1ex}{$\sum _{i=1}^n{\mathrm{S}}_i$}\!\left/ \!\raisebox{-1ex}{$\mathrm{N}$}\right.}$	In the formula, Si represents the projected area of the i-th tree; N represents the total number of trees in that area.
	MTCD [41] Mean Tree Crown Diameter	$\mathrm{M}\mathrm{T}\mathrm{C}\mathrm{D}={\displaystyle \raisebox{1ex}{$\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{D}}_{\mathrm{i}}$}\!\left/ \!\raisebox{-1ex}{$\mathrm{N}$}\right.}$	In the formula, Di represents the diameter of the crown of the i-th tree; N represents the total number of trees in that area.
	TCH [30] Tree Crown Height	$\mathrm{T}\mathrm{C}\mathrm{H}={\displaystyle \sum _i^n}{\mathrm{H}}_i$	In the formula, Hi represents the height of the crown of the i-th tree.
Combined Building and Vegetation 3D Morphology (Class II)	SVF [34] Sky View Factor	Captures photos with a fisheye lens and calculates the area of the sky in Photoshop as a percentage of the entire photo’s area.
	VR Volume Ratio of Vegetation to Building	$\mathrm{V}\mathrm{R}={\displaystyle \raisebox{1ex}{$\sum _{j=1}^k{v}_j$}\!\left/ \!\raisebox{-1ex}{${\sum }_{i=1}^n{\mathrm{V}}_i$}\right.}$	VR represents the volume ratio of vegetation to buildings within the buffer zone. Vi is the volume of the i-th vegetation within the buffer zone, and vj is the volume of the j-th building within the buffer zone.
	HR Height Ratio	$\mathrm{H}\mathrm{R}={\displaystyle \raisebox{1ex}{${\sum }_{i=1}^n{h}_i$}\!\left/ \!\raisebox{-1ex}{${\sum }_{j=1}^k{\mathrm{H}}_j$}\right.}$	HR represents the height ratio of vegetation to buildings within the buffer zone. hi is the height of the i-th tree within the buffer zone, and Hj is the height of the j-th building within the buffer zone.
	ER [42] Entity Ratio	$\mathrm{E}\mathrm{R}={\displaystyle \frac{\sum _{i=1}^n{\mathrm{V}}_i+\sum _{j=1}^k{v}_j}{\mathrm{A}}}$	ER represents the solid occupancy rate within the buffer zone. Vi is the volume of the i-th building within the buffer zone, vj is the volume of the j-th vegetation within the buffer zone, and A is the area occupied by the buffer zone.

).

Given the potential strong correlations among various morphological indicators that may influence subsequent statistical analysis results, an autocorrelation test was performed on 16 morphological indicators (Figure 3) to eliminate significantly correlated variables. Simultaneously, considering the integrity of the three-dimensional morphological indicator system and the applicability of these indicators in planning, some variables exhibiting significant correlations were also included in the subsequent analysis. Ultimately, a total of 14 three-dimensional morphological indicators were retained for further analysis and were selected after screening, comprising 5 Class I indicators (MBH, BV, FAR, SCD, and BSI), 6 Class II indicators (MTCV, SGV, GV, TCD, MTCA, and MTCD), and 3 Class III indicators (SVF, HR, and VR).

2.4. Statistical Analysis

The research structure for the study is illustrated in Figure 4. This study utilized SPSS 26.0, ArcGIS Pro, and R 4.3.0 software. Initially, all urban three-dimensional morphological indicators were normalized to establish a database, followed by the generation of spatial distribution maps depicting the changes in temperature and humidity for WDP, CHP, and ODP. Subsequently, correlation analysis, variance decomposition analysis, and redundancy analysis were performed. The specific analytical process is outlined as follows:

(1)

Correlation Analysis: A correlation matrix was constructed to examine the pairwise relationships between various categories of urban three-dimensional morphological indicators and changes in temperature and humidity during WDP, CHP, and ODP. This study employed the Pearson correlation analysis method, establishing a 95% confidence interval to identify indicators that are significantly associated with changes in temperature and humidity for further research.

(2)

Variance Partitioning Analysis (VPA): VPA facilitates the understanding of how multiple indicators independently and collectively influence the variance of the dependent variable. Indicators identified as significantly related to changes in temperature and humidity through correlation analysis were selected, and the Vegan package in R was employed to quantify both the independent and joint contributions of different categories of spatial morphology and the changes in temperature and humidity.

(3)

Redundancy Analysis (RDA): RDA aims to identify one or a set of variables among numerous factors that can explain significant changes in the dependent variable. In this study, RDA is employed to determine which morphological indicators exert a significant influence on changes in temperature and humidity during different phases under the combined effects of multiple indicators, thereby exploring the mechanisms underlying daytime changes in temperature and humidity within residential districts [29].

Figure 4. The methodological flowchart used in this study.

Figure 4. The methodological flowchart used in this study.

3. Results and Analysis

3.1. Correlation Between Changes in Temperature and Humidity and Urban 3D Morphology

The correlation analysis between the three-dimensional morphological indicators of residential districts and changes in temperature and humidity is shown in Figure 5. The data indicate that the three-dimensional morphology of urban districts is highly correlated with daytime changes in temperature and humidity. The three-dimensional morphological indicators are significantly correlated with changes in temperature and humidity during the CHP and ODP, while those are not significantly correlated with changes in temperature and humidity during the WDP.

The three-dimensional morphological indicators that exhibit significant correlations with changes in temperature and humidity differ during various phases. Class I morphologies demonstrate the weakest correlation with changes in temperature and humidity. During the WDP, both BV and FAR display relatively high positive correlations with changes in temperature. In contrast, during the CHP, BSI shows a strong positive correlation with changes in temperature, while BV and FAR transition to negative correlations at this time. Regarding humidity, only BSI consistently exhibits a negative correlation during different phases, with its strongest correlation occurring during the CHP. For Class II morphologies, significant indicators related to changes in temperature and humidity are observed in both the CHP and ODP; these indicators show positive correlations with changes in temperature and negative correlations with changes in humidity. Notably, SGV does not present any significant correlation during these daytime phases. However, in the CHP, it reveals a negative correlation with changes in temperature alongside a positive correlation with changes in humidity. For Class III morphologies, both VR and SVF indicate significant correlations with changes in temperature and humidity during the CHP; VR demonstrates a significant positive correlation with changes in temperature while exhibiting a significant negative correlation with changes in humidity. Conversely, SVF displays the opposite trend compared to VR. Correlation analysis can evaluate the strength and direction of the relationship between variables, but it has certain limitations in considering the interaction between variables and the complexity of the data, thus requiring further analysis.

3.2. The Response of Different Categories of 3D Morphology to Changes in Temperature and Humidity

In this section, the study selected indicators that exhibit a significant correlation with changes in temperature and humidity for analysis. The results of the analysis indicate (Figure 6) that Class II and Class III morphologies exert varying degrees of influence on changes in temperature and humidity. Specifically, the explanatory power of Class II and Class III morphologies regarding changes in temperature is 44.87%, with Class II morphology alone accounting for 13.50% of changes in temperature, while Class III morphology contributes 17.24%. The combined explanatory power of both categories for changes in temperature is 14.13%, which surpasses the individual effect attributed to Class II morphology alone. Furthermore, it should be noted that 55.13% of changes in temperature remain unexplained by either Class II or Class III morphologies.

Regarding humidity, Class II and Class III morphologies collectively account for 55.78% of changes in humidity, with Class II morphology alone explaining 22.43% of changes in humidity and Class III morphology contributing a mere 0.62%. The combined explanatory power of both categories for changes in humidity reaches as high as 32.73%. In other words, although the individual explanatory contribution of Class III morphology to changes in humidity is minimal, it exhibits a significant synergistic effect with Class II morphology in explaining changes in humidity.

3.3. The Response of Specific 3D Morphological Indicators to Changes in Temperature and Humidity

A further redundancy analysis was performed on indicators that exhibit significant correlations with changes in temperature and humidity, as illustrated in the figure (Figure 7). Regarding changes in temperature, the total explanatory power of the seven indicators for daytime changes in temperature is 44.5%. The cumulative explanatory power of the indicators associated with Class II morphology is 16.1%, which is lower than that of the indicators related to Class III morphology, at 28.4%. This finding is consistent with the results of the variance partitioning analysis. This further suggests that the regulatory effect of vegetation on daytime changes in temperature in residential districts is not contingent upon the contribution of a single morphological indicator; rather, it is achieved through the nuanced regulation resulting from the collective influence of multiple three-dimensional morphological indicators of vegetation. Among these indicators, SVF and MTCV demonstrate the most significant contribution to changes in temperature, with contribution rates of 26.7% (p = 0.002) and 11.2% (p = 0.016), respectively. Conversely, MTCA exhibits the lowest contribution rate at only 0.1%. Therefore, SVF and MTCV can be considered critical indicators affecting changes in temperature.

In terms of changes in humidity, the total explanatory power of the selected seven indicators is 56.3%, with Class II morphology indicators accounting for 40.4% of daytime changes in humidity and Class III morphology indicators contributing 15.9%. This underscores the significant role of Class II morphology in regulating daytime changes in humidity in residential districts, which aligns closely with the results from variance partitioning analysis. Similar to changes in temperature, SVF and MTCV exhibit substantial contributions to changes in humidity. However, at this time, MTCV’s contribution surpasses that of SVF, measuring 28.8% (p = 0.002) compared to SVF’s contribution of 15.4% (p = 0.002). The indicator with the lowest contribution is VR, at only 0.5%. Therefore, both MTCV and SVF are critical indicators influencing daytime changes in humidity.

To further elucidate the specific mechanisms through which urban three-dimensional morphology influences changes in temperature and humidity at different daytime phases, we analyze the results of the redundancy analysis as depicted in Figure 8. In this diagram, solid arrows represent changes in temperature and humidity parameters during WDP and CHP, while hollow arrows denote key three-dimensional morphological indicators. The projection of the morphological indicator arrows onto changes in temperature and humidity arrows—and their extensions—reflects the strength of that indicator’s impact on changes in temperature or humidity during a given phase. The angle between the morphological indicator arrows and those representing changes in temperature and humidity indicates their correlation; a smaller angle signifies a stronger correlation. If the indicator arrow aligns in the same direction as the arrow of changes in temperature and humidity, it represents a positive correlation; conversely, if it points in the opposite direction, it represents a negative correlation. An angle approaching 90° suggests no significant correlation.

The regulatory effects of morphological indicators on changes in temperature and humidity during WDP and CHP are explored. Regarding changes in temperature, during the WDP, the indicator most strongly correlated with changes in temperature is SVF, followed by VR. SVF exerts a positive influence on temperature increase, whereas VR has a negative effect. This implies that during the WDP, an increase in SVF will lead to more rapid rises in temperatures over a specified duration. Concurrently, GV also negatively impacts temperature increase to some extent. In contrast, during the CHP, SVF continues to be the most significantly correlated indicator for changes in temperature, followed by VR, GV, and TCD. However, at this phase, only an increase in SVF positively influences cooling temperature. The indicator contributing most significantly to the cooling phenomenon during the CHP is SVF, followed by MTCV and MTCD, with VR showing minimal contribution.

Regarding changes in humidity, during the WDP, all indicators exhibit a negative correlation with changes in humidity, indicating that an increase in these indicators will decelerate the rate of humidity decrease over a specified duration. Among them, MTCA and MTCD demonstrate the strongest correlation, suggesting that increases in these indicators will mitigate the decrease in humidity during the WDP. Concurrently, MTCA is also identified as the indicator contributing most significantly to changes in humidity during the WDP. In contrast, during the CHP, only SVF shows a positive correlation with changes in humidity and exhibits the strongest correlation; thus, an increase in SVF will expedite humidity increases during the CHP. The remaining three-dimensional morphological indicators are negatively correlated with changes in humidity during the CHP. The indicator that contributes most significantly to humidity changes is SVF, followed by GV and MTCV.

4. Discussion

Previous studies have mostly been conducted from a two-dimensional perspective to determine the mechanism by which urban form influences temperature and humidity changes during the day and night, and there is a lack of research on changes in temperature and humidity at different stages during daytime. Therefore, from the perspective of urban three-dimensional morphology, this study compared the influence mechanisms of different types of three-dimensional morphology on changes in temperature and humidity at different stages during the day. The results show that the three-dimensional form of buildings and vegetation has a synergistic effect on changes in temperature and humidity. Unlike in previous studies, it was found that an increase in three-dimensional vegetation morphology can inhibit cooling and humidification during the daytime. The cooling and humidification effect of vegetation can be changed by adjusting the configuration of the vegetation. Specific aspects of this are discussed below.

4.1. Urban 3D Forms Affecting Daytime Changes in Temperature and Humidity in Residential Areas

The daytime changes in temperature and humidity within residential districts exhibit significant spatiotemporal characteristics. Unlike previous studies [10], the overall influence of the three-dimensional morphology of buildings on changes in temperature and humidity is relatively modest, its regulatory effect on changes in temperature is more pronounced during the WDP compared to the CHP. This phenomenon may be associated with the energy balance factors that are prevalent in residential districts during the CHP. Firstly, buildings possess heat storage capabilities; after absorbing solar radiation, they also emit long-wave radiation as a heat source, resulting in a time lag in their regulation of changes in temperature [43]. During the CHP, as the solar elevation angle decreases, buildings enhance their interception of solar radiation, thereby facilitating a reduction in temperature. Furthermore, studies have indicated that anthropogenic heat emissions peak between 4 PM and 6 PM [44], which somewhat diminishes the regulatory effect of the three-dimensional morphology of buildings on changes in temperature. Given these multiple influencing factors, the impact of the three-dimensional form of residential buildings on changes in temperature during the CHP becomes less discernible.

Regarding the three-dimensional form of vegetation, the results indicate that there is no significant correlation between vegetation and changes in temperature and humidity during the WDP. However, an increase in morphological indicators of arboreal vegetation significantly mitigates both the decrease and increase in temperature and humidity during the CHP, thereby promoting daytime temperature rises while concurrently reducing humidity levels—this finding somewhat contrasts with the existing literature [3]. This discrepancy may be attributed to the fact that during the WDP, as the solar elevation angle increases, light penetrates through canopy gaps to create “sunflecks” on the ground [45], which affects solar radiation absorption in shaded areas of vegetation; consequently, this diminishes the regulatory effect of vegetation on temperature and humidity. During the CHP, increased morphological indicators of arboreal vegetation can impede the upward diffusion of heat, resulting in reduced changes in temperature. Simultaneously, the rise in temperature reduces the vapor pressure difference (VPD) between plant stomata and ambient air, thus suppressing the transpiration rate of plants [46]. This ultimately weakens the cooling and humidifying effects provided by vegetation.

In the integrated three-dimensional configuration of buildings and vegetation, reduction in the sky view factor (SVF) during the CHP was found to suppress changes in temperature and humidity within residential districts, aligning with existing research findings [47]. The measurement points of this study are mostly located along roads in the residential districts, where SVF reductions are frequently attributed to shading from street trees; this shading can impede air circulation and heat dissipation on roadways [48]. Furthermore, an increase in the volume ratio of vegetation to buildings also mitigated changes in temperature and humidity in residential districts during the CHP, which is consistent with prior studies.

4.2. Response Patterns of Urban Form to the Changes in Temperature and Humidity

The VPA results indicated that the integrated three-dimensional configuration of buildings and vegetation exerts the most significant influence on the spatial morphological types affecting changes in temperature in residential districts. However, it is essential to also consider the joint effect of this combined form and the three-dimensional morphology of vegetation on changes in temperature. This suggests that an increase in plant quantity does not necessarily yield better outcomes; rather, greater emphasis should be placed on the spatial relationship between vegetation and buildings, as this can impact wind speed—a critical factor for temperature regulation in subtropical regions [49]. The independent explanatory power of the three-dimensional morphology of vegetation regarding changes in humidity is considerably greater than that of the combined form of buildings and vegetation but remains lower than their collective explanatory capacity. Vegetation transpiration can enhance environmental humidity, while SVF and building density significantly influence vegetative transpiration [50].

Focusing on specific indicators, changes in SVF and the mean tree canopy volume exert significant regulatory effects on daytime changes in temperature and humidity. Among these factors, SVF contributes most substantially to temperature fluctuations, while the mean tree canopy volume plays a more critical role in influencing humidity levels. A larger mean tree canopy volume indicates that trees can enhance transpiration and evaporation processes, which aligns with existing research findings [28]. However, daytime changes in temperature and humidity exhibit different trends during various phases. Some studies have demonstrated that summer thermal comfort is negatively correlated with temperature but positively correlated with humidity. Based on these insights, morphological optimization strategies for residential areas are proposed.

4.2.1. Residential Area Morphological Strategies Based on Suppressing Temperature Increase and Humidity Decrease

During the WDP, a reduction in SVF can mitigate temperature increases. This phenomenon may be attributed to sample points with high SVF receiving greater solar radiation, while residential areas characterized by extensive hardened surfaces have been shown to experience the most significant surface temperature rises, compared to grasslands and farmlands [51], resulting in elevated daytime maximum temperatures. Increasing the vegetation-to-building-volume ratio can further help suppress temperature increases, potentially as a higher ratio may insulate heat radiation from buildings at pedestrian height. Moreover, the weak correlation between the mean tree canopy volume and changes in temperature during this phase suggests that, in contrast, the green volume exerts a certain inhibitory effect on warming. This indicates that a vegetation configuration integrating trees with shrubs and grasslands can provide more effective cooling for residential areas than merely planting trees alone—a finding also supported by existing research results [52].

4.2.2. Residential Area Morphological Strategies Based on Promoting Cooling and Humidity Increase

During the CHP, an increase in SVF most effectively enhances cooling and humidification, which aligns with the correlation results. However, while increasing the mean tree canopy volume of individual tree crowns is most effective for the mitigation of cooling, augmenting the green volume alongside the mean tree canopy volume proves to be more effective in suppressing humidification. Research has indicated that higher environmental humidity leads to a noticeable warming phenomenon near the canopies of trees and shrubs [40]. The sample points were situated in Fuzhou, where the annual rainfall reached 1314.12 mm in 2022, with particularly high levels recorded in May; this contributes to the suppressive effect of mean tree canopy volume on cooling during this phase. Conversely, a larger average green mass signifies an increased capacity for heat storage; as solar radiation diminishes during this phase, vegetation—due to its high specific heat capacity—becomes a source of heat [53]. The measurement points were predominantly located near street trees within residential districts, resulting in enhanced warming effects. Additionally, elevated initial environmental temperatures during this phase can restrict vegetation transpiration by influencing the vapor pressure deficit [40].

5. Conclusions

This study employed UAV remote sensing technology and fixed-point mobile measurement methods to collect data. The findings indicated that changes in temperature and humidity in residential districts exhibit a pattern of warming and dehumidification from morning to afternoon, followed by cooling and humidification from afternoon to evening. Furthermore, the characteristics of these changes in temperature and humidity demonstrate spatial correlations. The main conclusions drawn from this study are as follows:

• Regarding the differences in impact among morphological categories, the three-dimensional form of buildings exerts a limited influence on changes in temperature and humidity. In contrast, the three-dimensional structure of vegetation and the combined three-dimensional forms of both buildings and vegetation demonstrated a more significant correlation with changes in temperature and humidity during the CHP as well as throughout the ODP. Notably, increases in morphological indicators of arboreal vegetation and reductions in SVF inhibited cooling and humidification effects within residential districts.
• Regarding the synergistic effects of spatial form, the combined three-dimensional structure of buildings and vegetation exhibited the greatest independent explanatory power for changes in temperature in residential districts. However, it is also essential to consider the interactive effects between this combined three-dimensional form and the three-dimensional structure of vegetation on changes in temperature. The independent explanatory power of the three-dimensional form of vegetation concerning changes in humidity was significantly greater than that of the combined three-dimensional forms of buildings and vegetation, yet it remained lower than their collective explanatory power.
• Vegetation plays a significant regulatory role in changes in the temperature and humidity of residential districts by altering its three-dimensional form. Reducing the SVF and the mean tree canopy area, as well as increasing the volume ratio of vegetation to buildings, helps suppress warming and dehumidification. Increasing the SVF, reducing the mean tree canopy volume, and balancing the total green mass of trees, shrubs, and grass help promote cooling and humidification.
• Increasing the mean tree canopy area while reducing the green mass of the canopy can suppress warming and dehumidification and promote cooling and humidification in residential districts. Moreover, a vegetation configuration that combines trees with shrubs and grasslands can provide more cooling effects for residential districts than simply planting trees.

The data used in this study were collected on site in real-world scenarios. However, due to the inability to completely eliminate the influence of human factors, such as human activities and artificial heat sources, there are certain limitations regarding the external validity of the data. In addition, although the study revealed the impact mechanism of urban three-dimensional form on changes in temperature and humidity in residential districts, land cover types—as another key factor affecting changes in temperature and humidity in urban areas—have not been fully considered. Therefore, future research should further investigate the comprehensive effects of urban three- and two-dimensional forms on changes in temperature and humidity in residential districts.