The Spatial Pattern and Influencing Factors of China’s Nighttime Economy Utilizing POI and Remote Sensing Data

Abstract

The nighttime economy (NTE) is one of the primary measures used by the Chinese government to promote urban consumption and capital flow. Especially after COVID-19, more regulations were introduced by both the central and local governments to accelerate this commercial activity. However, the relationship between the NTE and urban development is controversial. There has been controversy over the relationship between the nighttime economy and urban development. We believe that organizations/individuals embedded in different regional contexts have different behavioral patterns, which, in turn, can make cities develop nighttime commercial activities differently. We wonder whether the nighttime economy’s large-scale development fits the diverse regional development. There is a lack of discussions of the spatial distribution of nighttime commercial activities from an urban geographical perspective, especially the differences and mechanisms of urban systems based on the nighttime economy. Based on existing research arguments, this article collects points of interest (POI) and nighttime light (NTL) remote sensing data (RSD) to spatialize nighttime economic activities in order to provide a reference for reasonable regional and urban economic planning. The nighttime economy (NTE) is one of the primary channels used by the Chinese government to promote urban consumption and capital flow, and the relationship between the NTE and urban development is controversial. Based on existing research, we selected the Yangtze River Delta (YRD) region as an example. We found that there are core–peripheral spatial patterns in nighttime commercial urban systems. The core is Shanghai, and provincial-level core cities form the second category, largely overlapping with the administrative urban system. Although the NTE is primarily concentrated in economically developed coastal areas, it spreads in the northwest–southeast direction, indicating that opportunities will arise in the geo-periphery. Although regulations encourage the growth of the NTE, infrastructure cannot fully support large-scale centralized expansion. The interaction of critical factors, such as urban policies, residents’ consumption, industrial structure, and economic foundations, may affect nighttime activities.

1. Introduction

Methods and applications of geospatial big data (GBD) comprise a hot topic of common concern in geographic information science as well as information science; this matter has important value for the in-depth exploration of the human–land relationship system [1]. The application of GBD to urban planning, intelligent transportation, environmental protection, and public safety has received widespread attention [2]. Analysis of regional socio-economic activities and urban planning is one of the most widely applied fields of GBD [2]. Especially in the identification of urban functions and spatial structures, existing research has used human activity trajectory data, social media data, urban street view images, points of interest (POI), and nighttime light (NTL) remote sensing data (RSD) to identify urban functional zones and urban centers and to carry out land use planning.

The urban economy’s resurgence is increasingly driven by the nighttime economy, especially after COVID-19 [3]. The concept of the nighttime economy originated in Britain in the 1970s [4,5] and is widely used in research on regional economic development [6,7,8]. It is regarded as economic activities spanning from 6 p.m. to 6 a.m. [3]. These activities mainly comprise catering, culture, entertainment, arts, festivals, events, sports, nightlife, tourism, and transportation services [3]. Nighttime economic activity is also relevant to the regional context and the local tourism industry [9]. Thus, the nighttime economy (NTE) is an economic phenomenon accompanying the development of urban economic levels [10]; it provides more possibilities for urban or regional economic growth [11], such as reshaping urban spaces, increasing employment opportunities for residents, improving the usage rate of urban infrastructure, and easing employment pressure [12].

The existing research on the NTE has a certain value, but systematic research is lacking about urban spatial patterns and, especially, its mechanisms grounded in GBD. Existing research on the NTE focuses on analyzing the economic value of night activities or the main problems of the NTE. On the one hand, from the perspective of the functional value of the NTE, innovation in the NTE has promoted improvements in product quality, industrial structure innovation, and supply-side service innovation; moreover, it plays an important role in shaping urban social space and driving urban economic growth [10,13]. Moreover, the social and economic benefits of urban nightlife in employment, tourism, and residents’ daily life are highlighted [14]. However, from the perspective of the existing problems of the NTE, although it is highly correlated with the development of cities, it has also caused many disruptions in social, economic, cultural, and other aspects, such as alcohol abuse, disorder, security risks, and other problems caused by nighttime activities [15,16,17,18]. In addition, China’s urban NTE faces challenges such as uneven development, inadequate infrastructure, and homogenization of the NTE’s format system.

The existing research on the night economy mainly focuses on the organizational behavior patterns of its participants [19], issues arising from the nighttime economy [20], and its relationship with urban policies or planning [5]. Previous research provides a good foundation, and some research has begun to discuss the relationship between the vitality of the nighttime economy and urban land use and spatial structure [21]. However, there is still insufficient data from a geographical spatial perspective on the spatial patterns of the night economy, especially research on the influencing mechanisms embedded in different regional contexts. Controversies exist in the research on the relationship between the nighttime economy and the urban economy or space. We believe this is a complex socio-economic phenomenon related to actors participating in nighttime activities and the urban context in which economic activities are embedded. Therefore, discussing the spatial patterns of nighttime activities and their influencing factors contributes to understanding the characteristics and operational mechanisms of this economic phenomenon in different spaces and stages. It also helps provide references for the governance of urban economic development and the rational planning of urban space.

Based on this lack of research and the need for economic growth, we discuss the spatial pattern changes and mechanisms of China’s NTE using GBD [10]. Taking the Yangtze River Delta (YRD) urban agglomeration with the most developed economy in China as a case study, by examining its current situation, problems, and the mechanism of nighttime economic space, this study provides a reference for the development of the NTE and the planning of space for night economic activity in the region.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature and highlights the importance of nighttime economic development as well as the application of POI data and RSD. Section 3 covers the data and methods. Section 4 presents the regional context and discusses the spatial distribution of the NTE, the supply and demand status of nighttime service facilities (NSFs), and the influencing factors of the NTE’s spatial distribution. Section 5 provides concluding remarks.

2. Research Background

2.1. Application of POI Data and Night Lights Remote Sensing Data

Nighttime lights detected by satellites are increasingly used by economists to reflect the absence or perceived inaccuracy of traditional economic statistics such as national or regional GDP [22]. Therefore, light remote sensing data can objectively reflect human socio-economic activities, production, and living conditions. The two main sources of night light data are the Defense Meteorological Satellite Program Operational Linescan System (DMSP for short) and the Day–Night Band (DNB) of the VIIRS [22]. Existing research has controversies regarding the application of data from different sources. For instance, DMSP data are unsuitable for studying small areas or examining boundary effects [22]. Consequently, more research is shifting towards the utilization of VIIRS data, which are considered effective in addressing the limitations associated with the DMSP data [22]. Therefore, when employing light remote sensing data to analyze social-economic activities, it is crucial to acknowledge potential issues with data from various sources. It helps prevent shortcomings and facilitates effective data application and analysis [23].

The night light remote sensing data objectively reflect human socio-economic activities, production, and living conditions. NTL RSD are free from the excessive interference of human factors, have high data accuracy, and cost less when it comes to long-term ongoing data collection [24], including data on the underground economy or the informal economy; NTL RSD also more comprehensively reflect economic activity [25]. Enabling researchers to observe the economic activities of different regions for a long time without being affected by the price level [26], NTL RSD are often used as a proxy indicator of gross domestic product (GDP) and its growth rate, or to adjust and evaluate real GDP through NTL RSD [27,28,29]. Second, NTL RSD can be used to measure the spatial distribution of social and economic activities. Currently, traditional social and economic data are based primarily on administrative divisions of statistics [26]. It is difficult to identify the hot spots of social activities and economic concentration points in geographic space with high quality; NTL RSD can overcome the limits of administrative units and reflect the spatial distribution characteristics of social and economic activities in a more detailed and comprehensive manner. Further, NTL RSD play an important role in the urban development process [30], nighttime economic hotspots [31], the distribution characteristics of large cities [32], and the determination of urban agglomeration [30]. Finally, NTL remote sensing images can be used to monitor disasters and environmental pollution by estimating regional resource consumption and carbon emissions [33,34], evaluating the impact of natural disasters on the environment and the economy [35], and analyzing the impact of war or regional conflict on national socio-economic dynamics [36,37].

The application of POI data in urban and regional economies helps us to understand the social and economic functions, vitality, and development of cities [38]. POI data can be used to identify urban land use types and functional areas [39], the spatial characteristics of human activity [40,41], industrial spatial development [42,43], the urbanization process [44], and population and economic distribution [39,45]. POI data are highly feasible for the above functions, which can enable the rapid identification of urban functional areas, population size, and business development. POI can also provide a reference for exploring the distribution law of different service facilities such as catering, accommodations, transport, sightseeing, entertainment, and shopping in urban spaces. POI data facilitate the exploration of urban spatial structures based on big data and support the planning of urban construction and development. From the perspective of POI data, with the continuous expansion of their application, their analysis methods have also been enriched [46], including the nearest neighbor index [43], kernel density analysis [43], spatial autocorrelation analysis [47], standard deviation ellipse analysis [48], spatial clustering analysis [49], accessibility analysis [50], and other techniques.

For this study, we referred to existing research and built a nighttime economic vitality index (NEVI), based on which—including NTL RSD and POI data—we analyzed the features of nighttime economic development and the spatial distribution of the YRD urban agglomeration from the standpoint of multi-source data. We also explored the supply and demand balance of NSFs.

2.2. Nighttime Economy and Regional Development

Nighttime economies and urban development are closely linked. Existing research on the NTE focuses on nighttime activities and regional economic growth, regional infrastructure, regional or urban service levels, and other aspects. Current research on the NTE has established a vitality index, including indicators such as catering, accommodations, tourism, shopping, entertainment, and fitness; the index has a highly significant correlation with the level of development of the regional economy and the tourism industry [3]. The NTE has brought social and economic benefits to the tourism industry, employment for residents, and citizen propaganda [14,51]. Nighttime economic development needs to promote transportation strategies corresponding to the dimensions of space and time to meet the night transportation needs of residents and tourists [52].

The NTE interacts with regional infrastructure; government supervision and management stimulate the effectiveness of private sector investment in infrastructure and increase the impact of building it on the economy [53]. The building of infrastructure is a decisive factor in regional livability, and urban transport infrastructure has a significant positive impact on nighttime economic growth [54,55,56]. The construction of commercial and entertainment infrastructure influences the time and spatial distribution of nighttime activities [57].

The development of the NTE is related to the capacity of the urban service level, and with the continuous upgrading of consumption demand, the NTE is playing an increasingly important role [58]. The ongoing innovation of the NTE format has promoted improvements in product quality, industrial structure innovation, and supply-side service innovation and guided the industry to continue to transfer to the high-end value chain. The NTE plays a vital role in meeting the social needs of residents, shaping urban social spaces, improving the charm of urban tourism, and driving urban economic growth [13]. First, tourism is the core sector in the expansion of the NTE. Based on the development and planning of the NTE, it is easier to build an industrial chain focusing on leisure and entertainment, tourism consumption, catering, and accommodations; this industrial chain plays a pivotal role in promoting the growth of the national economy. Second, the NTE embodies the integration of urban space and nighttime activities, which is of great significance to enhance the charm of the city and expand its influence. Finally, the NTE plays a vital role in meeting people’s needs for a better life.

2.3. Impact of Nighttime Economy in Different Context

The nighttime economy was initially seen as a strategic means for the UK to address deindustrialization and urban decline. Its development, on the one hand, led to a more dispersed and diverse urban spatial structure. On the other hand, it blurred the complexity of culture, replaced by an excessive focus on nighttime economic activities [59]. The global development of the nighttime economy has been proven to play a certain role in promoting urban economic development, but it has also given rise to some socio-economic issues. For example, McArthur argues that the expansion of nighttime activities in London can stimulate urban economic development, particularly in terms of consumption, extending the time and space for residents’ entertainment and spending [52]. The development of the nighttime economy also influences urban infrastructure and policy formulation. For instance, in the UK, public policy supports the growth of the nighttime economy, emphasizing the construction of restaurants, clubs, and bars.

Additionally, efforts are made to improve urban transportation conditions, enhancing the possibility of people engaging in nighttime activities and fostering the development of the urban nighttime economy [60]. While the nighttime economy promotes urban economic development, it has also triggered a series of socio-economic issues, leading to controversy. For example, studies using Montpellier and Bologna as examples have discussed the negative impacts of nighttime economic activities on the management, design, and planning of public spaces. These studies emphasize the need for effective urban policy planning measures to address the negative impacts of the nighttime economy [61]. The potential socio-economic issues associated with the development of Berlin’s nighttime economy include conflicts related to the excessive use of public spaces, nighttime noise, littering, and disorderly conduct [16]. In planning for the nighttime economy, various regions in Europe have significantly increased regulations to address potential issues arising from the development of the nighttime economy [15].

3. Methodology

3.1. Data Sources

To build an indicator system for the NEVI of the YRD urban agglomeration, we referred to the “2021 Zhicheng Nightlife Index”, “The Quality Index of Nighttime Economic Development in China 1.0 (2020)”, and “The Vitality Index of Nightlife in Shanghai” (

Table 1. Evaluation index system for the NEVI.

Data Description	First-Level Indicators	Second-Level Indicators
Nighttime economic vitality index	Consumer activeness	Mean value of the midnight movie search index
		Movie box office/expressed by the box office on October 1st
		Mean value of the bar search index
		Mean value of the KTV search index
		Total electricity consumption
	Traffic accessibility	Final subway time
		Number of buses and trams
		Total number of bus and tram passengers
		Number of taxis
	Service convenience	Percentage of nighttime merchants (in 2019)
		Mean value of the 24-h convenience store search index
		Number of people employed in tertiary industries
		Number of people employed in the cultural, sports, and entertainment sectors

). Among these, film box office figures are represented on an annual basis as of October 1st each year. The latest closing time of the subway is indicated by the ending time of the last train as of the year 2020. The proportion of nighttime operating businesses is expressed as a percentage of the 2019 nighttime operating business count. After dimensionless processing of the data indicators, we calculated the index weight based on the entropy method. Finally, we measured the night economic vitality index of the YRD urban agglomeration from 2011 to 2020 according to the weight value.

The calculation steps of index processing are as follows:

First, normalization:

$$Y_{ij}=\frac{X_{ij}-\min \left(X_i\right)}{\max \left(X_i\right)-\min \left(X_i\right)}\left(\mathrm{Positive}\mathrm{indicators}\right)$$

$$Y_{ij}=\frac{\max \left(X_i\right)-X_{ij}}{\max \left(X_i\right)-\min \left(X_i\right)}\left(\mathrm{Negative}\mathrm{indicator}\right)$$

proportion of the jth indicator in the ith plan:

$$p_{ij}=\frac{Y_{ij}}{{\displaystyle \sum _{i=1}^n{Y}_{ij}}},i=1,\dots ,n;j=1,\dots ,m$$

Then, calculate the entropy value, redundancy, and index weight of the j-th indicator, respectively:

$$E_j=-\ln {(n)}^{-1}{\displaystyle \sum _{i=1}^n{p}_{ij}}\ln p_{ij}$$

$$D_j=1-E_j w_j=\frac{D_j}{{\displaystyle \sum _{j=1}^m{D}_j}},j=1,2,\dots ,m$$

(in which $E_j$ ≥ 0; if $p_{ij}$ = 0, $E_j$ = 0)

According to the weight, the nighttime economic vitality index is determined:

$$S_i={\displaystyle \sum _{j=1}^m{w}_j}\cdot {\mathrm{Y}}_{ij}$$

① POI data: We obtained POI data from 26 cities in the YRD urban agglomeration. The data format consists of point-like vector data. We converted the coordinates into WGS-1984 geographic coordinates. The attributes included longitude, latitude, province, city, district, county, category, name, and detailed addresses. We classified the processed POI data of the service facilities according to business type (

Table 2. Classification system and the ratio of NSFs in the YRD urban agglomeration.

Tourism Features	Examples	Ratio
Catering	Chinese restaurant, fast food restaurant, foreign restaurant	25.87%
Accommodations	Hotel, inn	3.66%
Transport	Subway station, bus station, passenger port	11.28%
Sightseeing	Scenic area, science and technology museum, park square	1.61%
Entertainment	KTV, cinema, and chess room	3.74%
Shopping	Supermarket, commercial street, specialty store	53.84%

).

② Night lights remote sensing data: We derived the NTL images from the NPP-VIIRS NTL data of 2020 obtained by the NOAA National Centers for Environmental Information. The spatial resolution of the NTL image after resampling was 500 m. Considering the YRD urban agglomeration as the research area, we cut an NTL image map using the administrative division vector map of the YRD urban agglomeration.

③ Scio-economic statistic data: We obtained data on the cities and relevant attributes from the China Statistical Yearbook, the City Statistical Yearbook, the Statistical Bulletin, and the first and second batches of the National Night Culture and Tourism Consumption Cluster List published by the Ministry of Culture and Tourism.

The influencing factor analysis indicators were urban traffic construction, expressed by highway network density. ① Traffic construction expressed by highway network density, which has a direct impact on regional economic growth. The convenience of transportation and the level of urban transportation infrastructure development are intermediate inputs and endogenous explanatory variables for urban economic growth. ② Population agglomeration expressed by population density. The spatial distribution of population in the Yangtze River Delta is uneven. Population density is one of the key factors for regional sustainable development, which also promotes regional urbanization. ③ Industrial structure expressed by the proportion of tertiary industry. ④ Economic development level expressed in per capita GDP. The nighttime economy is a crucial component of the leisure economy, and per capita GDP, as an economic indicator, represents the regional economic development level of urban clusters. Regional economies provide the material foundation and resource support for the construction of nighttime economic service facilities and the implementation of nighttime economic policy planning in various cities. Simultaneously, nighttime economic consumption serves as a driving force for regional economic growth, creating a reciprocal influence between the two. ⑤ The implementation of government policies expressed by the number of national night culture and tourism consumption clusters. The consumption level of residents is expressed in per capita consumption expenditure. This article uses the number of national-level nighttime culture and tourism consumption aggregation areas in 2021 and 2022 as variables for government policy implementation. This is because it reflects, to some extent, the inclination of various cities in the Yangtze River Delta urban cluster toward the nighttime economy in government policies.

3.2. Research Method

① Nearest neighbor index: The nearest neighbor index is expressed as the ratio of the observed mean distance to the expected mean distance. We used the POI data of NSFs in the YRD urban agglomeration as the research object, and we calculated the nearest neighbor index of different types of service facilities. Therefore, it can indicate the concentration or dispersion level of service facility POI data in geographical space. Using the nearest neighbor index to measure the degree of clustering of six types of nighttime economic service facilities related to travel in the Yangtze River Delta urban cluster (food, accommodation, transportation, entertainment, shopping, and sightseeing) provides the potential for optimizing the layout of nighttime economic infrastructure for the rational development of the region.

If the index is less than 1, the NSFs show a clustering pattern in geographic space; if the index is greater than 1, the NSFs exhibit a discrete pattern in geographic space. The formula is [62]:

$$R=\frac{{\overline{r}}_1}{{\overline{r}}_E}$$

(1)

in which:

$${\overline{r}}_E=\frac{1}{2\sqrt{n/A}} {\overline{r}}_1=\frac{{\displaystyle {\sum }_{i=1}^n{d}_i}}{n}$$

where $R$ is the nearest neighbor index, ${\overline{r}}_1$ is the mean value of the distance between the nearest points, ${\overline{r}}_E$ is the theoretical nearest neighbor distance, $A$ is the area of the region, $n$ is the number of research objects, and $d_i$ represents the distance between the centroid of each feature and the centroid position of its nearest neighbor feature.

② Kernel density estimation (KDE): Kernel density estimation (KDE) involves employing kernel smoothing for the estimation of probability density. In other words, it is a non-parametric approach to assessing the probability density function of a random variable, utilizing kernels as weights. In our research, KDE reflects the degree of agglomeration of POI data in a study area and is widely used to analyze the spatial agglomeration of POI data. This helps us effectively assess the spatial balance issues of nighttime economic service facilities in various formats within the Yangtze River Delta urban cluster, especially providing recommendations for the integrated and coordinated development of nighttime economic infrastructure in the core and peripheral areas of the Yangtze River Delta. The formula is as follows [63]:

$$\widehat{f_h}(x)=\frac{1}{nh}{\displaystyle \sum _{i=1}^{n}K}\left(\frac{x_i-x}{h}\right),x\in R$$

(2)

where $\widehat{f_h}(x)$ represents the kernel density estimation value, $K\left(\frac{x_i-x}{h}\right)$ is the kernel function, $h$ represents bandwidth, and $n$ represents the number of POI for nighttime service facilities.

③ Standard deviation ellipse analysis: This method is used to summarize the spatial characteristics of geographic features such as central tendency, dispersion, and directional trends and to compare and analyze the distribution direction and degree of concentration of different NSFs in the YRD urban agglomeration. It helps us understand the spatial scope and development trends of differentiated nighttime economies, providing valuable insights for future nighttime economic development. We conducted a standard deviation ellipse analysis on the POI data of NSFs. The size and shape of an ellipse can describe the dispersion and concentration of NSFs. The major and minor axes of the standard deviation ellipse represent the diffusion direction and degree of concentration of NSFs, respectively, whereas the azimuth of the standard deviation of the ellipse reflects the distribution direction of NSFs. The formula is as follows:

Weighted mean center of features [64].

$$x_w=\frac{{\displaystyle \sum _{i=1}^n{w}_i}{x}_i}{{\displaystyle \sum _{i=1}^n{w}_i}}, y_w=\frac{{\displaystyle \sum _{i=1}^n{w}_i}{y}_i}{{\displaystyle \sum _{i=1}^n{w}_i}}$$

(3)

Standard deviation along the X-axis and Y-axis:

$${\sigma }_x=\sqrt{\frac{{\displaystyle \sum _{i=1}^n{\left(w_{i}x{\prime }_i^{}\cos \theta -w_{i}y{\prime }_i^{}\sin \theta \right)}^2}}{{\displaystyle \sum _{i=1}^n{w}_i^2}}}, {\sigma }_y=\sqrt{\frac{{\displaystyle \sum _{i=1}^n{\left(w_{i}x{\prime }_i^{}\sin \theta -w_{i}y{\prime }_i^{}\cos \theta \right)}^2}}{{\displaystyle \sum _{i=1}^n{w}_i^2}}}$$

(4)

Azimuth:

$$\tan \theta =\frac{\left({\displaystyle \sum _{i=1}^n{w}_i^2}x{\prime }_i^2-{\displaystyle \sum _{i=1}^n{w}_i^2}y{\prime }_j^2\right)+\sqrt{{\left({\displaystyle \sum _{i=1}^n{w}_i^2}x{\prime }_i^2-{\displaystyle \sum _{i=1}^n{w}_i^2}y{\prime }_j^2\right)}^2+4\left({\displaystyle \sum _{i=1}^n{w}_i^2}x{\prime }_i^{2}y{\prime }_j^2\right)}}{2{\displaystyle \sum _{i=1}^n{w}_i^2}x{\prime }_{i}y{\prime }_j}$$

(5)

where ($x_i$,$y_i$) represents the coordinates of point features; n indicates the total number of features; $w_i$ refers to weight; ($x_w$,$y_w$) signals the weighted mean center of features; ($x_i^{\prime }$,$y_i^{\prime }$) represent the coordinate deviations of different POI point features from the mean center; ${\sigma }_x$ and ${\sigma }_y$ refer to the standard deviation along the x- and y-axes; and $\theta$ is the azimuth, an intersecting angle formed by rotating the ellipse clockwise from due north to the major axis.

We used a geographic detector to discern the spatial differentiation of the nighttime economic index and the extent to which factors such as urban traffic construction, urban population concentration, urban industrial structure, the level of economic growth, government policy implementation, and residents’ consumption level explain the spatial differentiation of the nighttime economic index.

4. Result

4.1. Spatializing the Nighttime Economy

4.1.1. Regional Context

According to the percentage of nighttime merchants in the YRD region in 2019 (Figure 1), Shanghai had the highest proportion of nighttime merchants (94.6%), followed by Zhejiang Province (94.3%). According to the compound growth rate of the nighttime transaction number of new business types in the YRD region (Figure 2), Zhejiang Province ranked first among the three provinces and one municipality (82.5%), followed by Anhui Province (79.8%). Jiangsu Province ranked third at 72.2%, whereas Shanghai ranked fourth at 56.2%.

In April 2021, we used the massive algorithm (an insight-based algorithm rooted in a content consumption trend under massive engine) to compare data on Tik-Tok activities from 20:00 in the evening to 2:00 the following morning. Urban agglomeration in the YRD had the highest nighttime vitality. According to the Alibaba Report on the Nighttime Economy 2020, from June 2020 onward, Shanghai ranked first among the top ten cities with the largest number of viewers on nighttime live-streaming platforms and the highest overall nighttime consumption. Jinhua and Hangzhou ranked first in terms of nighttime delivery amounts. According to the nightlife index released by the CBN, Shanghai ranked first among all Chinese cities in terms of NTL intensity and nighttime traveler activity. In 2021, Zhejiang Province was proposed to support the cultivation of an NTE, and the first list of 27 pilot cities was announced. Wuxi used an integrated development model centered on the concept of NTE+ to provide tourists with deeply immersive sightseeing experiences. The above policies and plans represent the great efforts made by the urban agglomeration of the YRD to develop its NTE. This is also reflective of citizens’ and visitors’ enthusiasm and support for developing an NTE and engaging in nighttime activities.

4.1.2. Spatializing the Nighttime Economy Vitality

We classified the NEVI of the YRD urban agglomeration for 2020 using the natural discontinuity method, as shown in Figure 3, which indicates that nighttime human activity in Shanghai is more active; the NEVI is the highest there. Hefei, Nanjing, Wuxi, Suzhou, Hangzhou, Ningbo, and Hangzhou fall under the second category; the NEVI is between 0.1141 and 0.5766. Yancheng, Yangzhou, Taizhou, Nantong, Zhenjiang, Changzhou, Jiaxing, Shaoxing, Taizhou, and Jinhua belong to the third category; the NEVI is between 0.0631 and 0.1140. Chuzhou, Wuhu, Huzhou, and Zhoushan comprise the fourth category; the NEVI is between 0.0297 and 0.0630. Anqing, Tongling, Ma’anshan, Xuancheng, and Chizhou fall under the fifth category; the NEVI is between 0.0027 and 0.0296. In general, the level of nighttime economic growth in municipalities directly under the central government, eastern coastal cities, and provincial capital cities is higher than that of other regions. Shanghai’s level of nighttime economic growth is higher than that of Zhejiang and Jiangsu provinces, and the levels of Zhejiang and Jiangsu Province are higher than that of Anhui Province.

We calculated the total NTL intensity and average NTL intensity and used the natural discontinuity method to classify the total NTL intensity and average NTL intensity of the YRD urban agglomeration in 2020, as shown in Figure 4. The coefficients in

Table 3. Fitting relationship between the NEVI and NTL RSD for 2011–2020.

Variables	Fitting Coefficient	Fitting Model	p Value	t Value	R2
Total light intensity	0.7954	Y = 0.7954X − 0.0051	0.000	31.86	0.7974
Average light intensity	0.7970	Y = 0.7970X + 0.0135	0.000	19.06	0.5847

and

Table 4. Linear fitting analysis of NTL RSD, POI data, and NEVI for 2020.

Variables	Fitting Coefficient	Fitting Model	p Value	t Value	R2
Total light intensity	0.8777	Y = 0.8777X − 0.0432	0.000	10.49	0.8210
Average light intensity	0.7165	Y = 0.7165X − 0.0021	0.000	5.78	0.5820
Catering	0.8965	Y = 0.8965X − 0.0703	0.000	13.45	0.8829
Accommodation	0.8347	Y = 0.8347X − 0.0570	0.000	11.75	0.8519
Transport	0.9732	Y = 0.9732X − 0.0236	0.000	17.83	0.9298
Sightseeing	0.7313	Y = 0.7313X − 0.0401	0.000	8.21	0.7373
Entertainment	0.9765	Y = 0.9765X − 0.0271	0.000	19.49	0.9406
Shopping	0.7170	Y = 0.7170X − 0.1018	0.000	6.96	0.6688

represent the fitted coefficients from the OLS linear regression. In Table 3, the independent variables include the total light intensity and average light intensity. The dependent variable is the economic vitality index at night. In Table 4, the independent variables are total light intensity, average light intensity, and the number of POI points of the six types of elements in 2020, and the dependent variable is the economic vitality index at night.

The spatial distribution trend of the total NTL intensity is similar to that of the NEVI. The total NTL intensity of Shanghai and Suzhou is between 219,806.7039 and 426,214.2813. The total NTL intensity in Nanjing, Wuxi, Hangzhou, and Ningbo is between 144,451.5565 and 219,806.7038, which falls under the second category. The total NTL intensity in the cities of Yancheng, Nantong, Hefei, Changzhou, Jiaxing, and Jinhua is between 88,754.27363 and 144,451.5564, which belongs to the third category. The rest of the cities fall under the fourth and fifth categories; the total NTL intensity is between 8484.660156 and 88,754.27362. The spatial distributions of the average NTL intensity, total NTL intensity, and the NEVI differ slightly. Shanghai and Suzhou belong to the first category; the average NTL intensity is between 8.164186577 and 15.18127441. The cities of Nanjing, Zhenjiang, Changzhou, Wuxi, Jiaxing, Ningbo, and Zhoushan belong to the second category; the average NTL intensity is between 4.71411839 and 8.164186576. The cities of Yangzhou, Taizhou, Nantong, Hefei, Huzhou, Hangzhou, and Shaoxing fall under the third category; the average NTL intensity is between 2.66746777 and 4.714118389. The rest of the cities fall under the fourth and fifth categories, respectively; the total intensity of NTL is between 0.269962758 and 2.667467769.

The total and average NTL intensities from 2011 to 2020 are dimensionless and compared with the NEVI of each city during the same time span. The fitting results are listed in Table 3. All pass the significance test at the 1% level. The R² value of the total NTL intensity is 0.7974, and the R² value of the average NTL intensity is 0.5847. The fitting effect of the total NTL intensity is better than that of the average NTL intensity. Thus, in a follow-up study, we would view the area of the total NTL intensity as the hotspot area of the NTE, representing a region of strong demand. We calculated and verified the NEVI of 2020 by using NTL RSD, POI data, and linear fitting in 2020 (Table 4). The POI data for NSFs, the total NTL intensity, and the average NTL intensity of the different business types pass the significance test at the 1% level. It is, therefore, reasonable to use the POI data of NSFs to study the nighttime economic spatial distribution of the YRD urban agglomeration.

4.1.3. Spatializing the Nighttime Economy Facilities

The concentration area of total NTL intensity embodies the hotspot area of the NTE, denoting a region with strong demand. The POI data of NSFs indicate the spatial distribution of NSFs. By analyzing the data of the YRD urban agglomeration, we can see that the NTE in the region exhibits clear characteristics of spatial agglomeration, and there are obvious differences between different cities, with a directional discrepancy from northwest to southeast.

First, the spatial characteristics of the YRD urban agglomeration are significant. The six types of NSFs in the YRD urban agglomeration—catering, accommodations, transport, sightseeing, entertainment, and shopping—display significant agglomeration at the 99% level (

Table 5. Spatial concentration characteristics of nighttime service facilities.

Features	Observation Range/m	Expected Mean Range/m	R	z Value	p Value	Spatial Type
Catering	36.184	319.870	0.113	−1663.182	0.000	Significantly Concentrated
Accommodation	136.625	849.391	0.161	−591.770	0.000	Significantly Concentrated
Transport	133.106	483.455	0.275	−897.535	0.000	Significantly Concentrated
Sightseeing	460.736	1286.421	0.358	−300.147	0.000	Significantly Concentrated
Entertainment	218.184	836.517	0.261	−526.748	0.000	Significantly Concentrated
Shopping	28.941	222.107	0.130	−2352.853	0.000	Significantly Concentrated

). Catering facilities have the most prominent agglomeration features. Catering consumption has the universality of 24-h service and the rigidity of consumption demand, which has become the regular content of the NTE. Shopping facilities have been continuously developed and have flourished in the process of building an urban NTE; they play an important role in the nighttime economic activities of residents and tourists. Accommodations are necessary for tourists to continue their long-term travels and to participate in nighttime economic activities. The degree of development of the accommodations service industry affects the depth and breadth of nighttime economic activities. Leisure and entertainment facilities allow residents and tourists to relax and pursue pleasant experiences. Leisure and entertainment activities are intertwined and closely related to nighttime economic activities; they overlap in spatial distribution and facilities. Traffic diversification and networking, the duration of nighttime operations of subways, and other forms of public transit are closely tied to the growth of the NTE. Urban development must accelerate the construction of comfortable, convenient, and safe transportation facilities to meet the needs of residents and tourists at night. Among the six types of NSFs mentioned above, the agglomeration of sightseeing facilities is the smallest. Compared to other elements, the spatial distribution of natural and cultural landscapes (such as scenic spots and historical buildings) is more restricted and affected by geographic features and historical development.

Second, the spatial imbalance between the different cities is prominent. According to the nuclear density analysis (Figure 5), the spatial distribution of NSFs in different cities is considerably different and uneven, showing the spatial distribution features of many dense NSFs in the east, whereas there are far fewer and more dispersed NSFs in the west. In addition, we can see the distribution pattern of one core area and multiple sub-core areas, with the center spreading to the edges. Shanghai’s core area has a dense population and is experiencing rapid economic growth. When establishing an NTE, a city concentrates on building a nighttime economic system that integrates six types of NSFs: catering, accommodations, transportation, sightseeing, entertainment, and shopping. It has established a good business network and a solid nighttime consumer base. The sub-core areas consist of various provincial capitals, including Hefei, Nanjing, and Hangzhou. The levels of economic growth in these cities and in Shanghai differ substantially. Despite this, they are sub-core regions in the YRD urban agglomeration, second only to Shanghai in terms of economic potential, nighttime consumer resources, and adequate NSFs. The characteristics of the spatial distribution of small-core areas are closely linked to NSF type. In small core regions, catering and shopping services are more dependent on spatial distribution, whereas sightseeing and entertainment facilities are less distributed.

Third, the northwest–southeast diffusion is significant. Standard deviational ellipse analysis can be used to describe the spatial distribution of NSFs in the YRD urban agglomeration as well as their direction of development. The dispersion and concentration of NSFs can be described by the size and shape of the ellipse. The major and minor axes of the standard deviational ellipse represent the direction of diffusion and the degree of concentration of NSFs, respectively. The shorter the minor axis, the more obvious the centripetal force of the data. The more concentrated the future distribution of NSFs and the longer the long axis, the greater the diffusion of NSFs. The azimuth angle of the standard deviational ellipse denotes the direction of distribution of NSFs. From the perspective of the distribution range (Figure 6), the standard deviational ellipse, which comprises 68% of the data, covers the cities of Nanjing, Zhenjiang, Changzhou, Wuxi, Suzhou, Shanghai, Huzhou, Jiaxing, and Shaoxing. There is no major difference in the coverage of NSFs between the different types of businesses. From the viewpoint of directionality and centripetal force, the general development pattern of NSFs in the urban agglomeration of the YRD is such that NSFs are spread along the northwest–southeast direction. There is no significant difference in the directionality or centripetal force of NSFs between the different types of businesses.

4.2. Analyzing the Influencing Mechanism

4.2.1. Supply and Demand Analysis of Nighttime Economic Facilities

We measured the demand for nighttime activity in the urban agglomeration of the YRD using the NTL data. We measured the supply status of NSFs using tourist POI data. Based on fishing nets with 50 rows and 50 columns, we calculated the NTL data and the supply quantity of NSFs provided in the YRD urban agglomeration for each fishing net. Using natural breaks, we reclassified the data and divided them into five groups ranging from small to large. We assigned values of 1 to 5 to the data. After reclassification, we compared the NTL data score to the supply quantity of the NSF score (the reclassified score values of the NTL data minus the reclassified score values of the NSF supply quantity). (In Figure 7, because the range of values obtained by subtracting the reclassified score values of the NTL data from the reclassified score values of NSF supply quantity is between −4 and 3, the figure only depicts values within this range.) When the reclassified score values of the NTL data are greater than the reclassified score values of NSF supply quantity, the demand for NSFs exceeds the supply, referred to as the relative deficit area of NSFs. Conversely, when the reclassified score values of the NTL data are less than the reclassified score values of NSF supply quantity, the demand for NSFs is less than the supply; this is referred to as the relative surplus area of NSFs.

In general, the relative surplus area that corresponds to each type of business in the YRD urban agglomeration accounts for the smallest proportion. A relative surplus occurs in cities with a low level of economic growth, slow nighttime economic development, low population activity, and limited tourist flow. These cities include Yancheng, Chuzhou, Xuancheng, Huzhou, and Taizhou. A significant deficit exists in Shanghai, Hefei, Nanjing, Wuxi, Suzhou, Jiaxing, Ningbo, and Zhoushan, which have high levels of economic growth, high nighttime economic vitality, high population activity, and huge tourist flows. Current NSFs cannot meet the needs of residents and tourists. In addition, there are fairly well-balanced catering and shopping facilities for the NTE. However, the construction of transportation facilities must be reinforced.

4.2.2. Influential Factor of the Nighttime Economy

From the perspective of the influencing factors of the NEVI of the YRD urban agglomeration, government policy (0.9054) plays a leading role in regional differences in the NEVI. According to the outcomes of interaction detection (

Table 6. Results of differentiation and factor detection.

Independent Variable	Urban Transport Construction	Urban Population Agglomeration	Urban Industrial Structure	Economic Development Level	Governmental Policy Implementation	Household Consumption Level
q value	0.0722	0.4468	0.6069	0.4881	0.9054	0.6811
p value	0.6971	0.1277	0.0090	0.0089	0.000	0.0068

), when the implementation of government policy interacts with any of the other five influencing factors, the dependent variable is more likely to explain the regional NEVI. Hence, the interaction manifests as a two-factor enhancement. China’s nighttime economic policy was first issued by the Qingdao Municipal Finance Office on 12 May 2004. With nighttime tourism at the core of development, the NTE of Qingdao has been actively investigated. With directive documents issued by the General Office of the State Council in August 2019, such as the Opinions on Further Stimulating the Consumption Potential of Culture and Tourism and the Opinions on Accelerating the Development of Circulation and Promoting Commercial Consumption, which established the importance and necessity of “vigorously developing the nightly culture and tourism economy”, various cities have witnessed an upsurge of development and promotion of the nightly economy. Further, the implementation of government policies has stimulated the market vitality of cities’ NTEs. An NTE can rise, develop, and stand out without substantial support of government policies.

Household consumption level is the second most influential factor in explaining the regional NEVI (0.6811) after the execution of government policy. The results of interaction detection indicate that the interaction between household consumption level and the remaining influencing factors manifests as a two-factor enhancement (

Table 7. Types of interactions between the independent and dependent variables.

	Type of Interaction	Resulting Value	Type of Interaction
Urban Transport Construction ∩ Urban Population Agglomeration		0.8334	Non-linear enhancement
Urban Transport Construction ∩ Urban Industrial Structure		0.8056	Non-linear enhancement
Urban Transport Construction ∩ Economic Development Level		0.6233	Non-linear enhancement
Urban Transport Construction ∩ Governmental Policy Implementation		0.9495	Two-factor enhancement
Urban Transport Construction ∩ Household Consumption Level		0.7782	Two-factor enhancement
Urban Population Agglomeration ∩ Urban Industrial Structure		0.9338	Two-factor enhancement
Urban Population Agglomeration ∩ Economic Development Level		0.6140	Two-factor enhancement
Urban Population Agglomeration ∩ Governmental Policy Implementation		0.9333	Two-factor enhancement
Urban Population Agglomeration ∩ Household Consumption Level		0.8436	Two-factor enhancement
Urban Industrial Structure ∩ Economic Development Level		0.8492	Two-factor enhancement
Urban Industrial Structure ∩ Governmental Policy Implementation		0.9288	Two-factor enhancement
Urban Industrial Structure ∩ Household Consumption Level		0.7827	Two-factor enhancement
Economic Development Level ∩ Governmental Policy Implementation		0.9672	Two-factor enhancement
Economic Development Level ∩ Household Consumption Level		0.8652	Two-factor enhancement
Governmental Policy Implementation ∩ Household Consumption Level		0.9452	Two-factor enhancement

). With the improvement in residents’ consumption levels and rising demand for consumer services, the NTE has become a critical supply for meeting people’s diversified consumption needs. At the same time, the NTE can enhance the charm of as city, release the potential of the consumption market, promote the recovery of the service sector, extend tourists’ retention time, drive the growth of the local economy, and boost residents’ consumption levels.

The urban industrial structure ranks third in explaining the regional NEVI (0.6069). The results of interaction detection indicate that the interaction between the urban industrial structure and urban transport construction manifests as a non-linear enhancement. However, its interaction with the remaining four influencing factors—urban population agglomeration, the level of economic development, the implementation of government policy, and household consumption level—manifests as a two-factor enhancement. This interaction increases the interpretative ability of the regional NEVI. The NTE covers almost all tertiary sectors, such as catering, accommodations, tourism, and entertainment. An industrial pattern dominated by the service industry forms the basis for the expansion of the NTE. Hence, the urban industrial structure is a vital factor in the growth of the city’s NTE as well as a crucial reference for evaluating the city’s nighttime economic benefits and level of high-quality development.

The level of economic growth (0.4881)—based on the outcomes of interaction detection, in addition to the interaction between the level of economic growth and urban transport construction—manifests as a non-linear enhancement. Its interaction with urban population agglomeration, the urban industrial structure, the implementation of government policy, and household consumption levels manifests as a two-factor enhancement. This interaction increases the explanatory power of the regional NEVI. With improvements in economic growth and per capita income, the demand for leisure and entertainment among residents and tourists is rising annually. In areas with high levels of economic development, infrastructure construction is good and the consumption level of residents is high, which can meet the basic needs of residents at night. In areas with low levels of economic expansion, due to the small consumption demand of residents and tourists as well as the weak momentum underlying economic growth, the development of the NTE lacks the necessary economic foundation; thus, it starts late and develops at a low level.

The explanatory power of the single-factor urban transport construction (0.0722) and urban population agglomeration (0.4468) for the regional NEVI fails to pass the significance test. However, the interaction between urban transport construction and three of the remaining influencing factors (i.e., urban population agglomeration, urban industrial structure, and the level of economic development) manifests as a non-linear enhancement. The interaction between urban transport construction, urban population agglomeration, and other influencing factors is improved by two factors, which increases the interpretative ability of the regional NEVI. The convenience of travel and population size also affect the development of the NTE. However, without the support of government policies, industrial structures, and economic growth, even if the city has dense traffic and many traffic lines, it cannot play a role in promoting the NTE.

5. Conclusions and Discussion

Effectively applying GBD to the analysis of new socio-economic phenomena can provide a more comprehensive, in-depth understanding of their essence and mechanisms. The NTE is an important form of expansion in urban commercial activities. It is also one of the primary phenomena facilitated by the Chinese government to promote urban consumption and capital flow. We applied POI data and NTL RSD to analyze the spatial features and mechanisms of the NTE in the YRD from the perspective of urban geography, expecting to provide a valuable reference for urban commercial planning.

The NTEs of cities in the YRD have core–peripheral features. Shanghai occupies the core position. The sub-core cities are provincial-level cities such as Hefei, Nanjing, Wuxi, Suzhou, Hangzhou, Ningbo, and Hangzhou. Yancheng, Yangzhou, Taizhou, Nantong, Zhenjiang, Changzhou, Jiaxing, Shaoxing, Taizhou, and Jinhua are in the sub-peripheral position. The most peripheral regions are Anqing, Tongling, Ma’anshan, Xuancheng, and Chizhou. The spatial structure spreads in the northwest–southeast direction, possibly due to the imbalance of nighttime infrastructure services, especially transportation infrastructure. Government policies, resident consumption levels, urban industrial structures, and economic growth levels are the main factors affecting the regional NTE. Further, the interaction of these factors is more effective for the expansion of nighttime economic activities than a single condition.

Given the differences in urban spaces, building nighttime urban economic clusters and activity hotspots by classification is of great value. Examples include building a first-level nighttime economic agglomeration space in an urban comprehensive center with higher nuclear density, building a second-level nighttime economic agglomeration space grounded in the urban district center in an area with lower nuclear density, and finally building a third-level nighttime economic agglomeration space in township, community, and business district centers. Nighttime economic activities should be created that can highlight the characteristics of culture and art in tourist attractions as well as cultural and art centers. In addition, commercial catering activities can be implemented that can hold business banquets in office areas. Moreover, nighttime economic activities can be arranged that focus on community public services in public spaces where crowds gather.

In accordance with the spatial pattern of the nighttime economy in the Yangtze River Delta, it is essential to strategically coordinate the supply and demand relationship of corresponding service facilities. The proportion of relative deficit areas for service facilities related to the nighttime economy, including dining, accommodation, transportation, tourism, entertainment, and shopping, is higher than that of relative surplus areas. Specifically, there is a relative shortage in the supply of nighttime economic service facilities, with dining facilities having a relatively balanced construction. Shopping service facilities have relatively sufficient construction, while there is still a need for improvement in the construction of transportation infrastructure. Therefore, it is necessary to continue leveraging the radiating and driving role of nighttime economic activity hotspots in surrounding areas. This involves strengthening the infrastructure development in these hotspots and addressing the relative deficit phenomenon in service facilities. Creating distinctive neighborhoods and nighttime cultural tourism projects in the peripheral areas of the Yangtze River Delta can attract tourists to participate in nighttime economic activities, achieving the guidance and dispersion of nighttime pedestrian flow. Urban public transportation services should adjust their operating hours according to the anticipated scale of nighttime economic activities, considering appropriately extending operation times. Additionally, specific areas such as city pedestrian streets, shared streets, and slow commercial spaces could be opened during designated time periods.

This research provides a reference for economic growth and regional spatial planning and points out that the NTE is not applicable to all regions. Moreover, the planning and development of urban nighttime activities should not be done blindly or in a one-size-fits-all manner. It is necessary to consider basic regional economic differences and to develop diversified urban economic formats. From the practical aspect of regional planning, the research on the spatial characteristics and impacts of the nighttime economy in the Yangtze River Delta contributes to the coordinated development of the region, offering valuable insights for the harmonious development of diverse areas. The spatial analysis and service facility supply–demand analysis can accurately pinpoint the advantageous areas for the nighttime economy development in the Yangtze River Delta. On the one hand, it leverages the driving role of the core nighttime economy areas in stimulating the development of peripheral regions. Moreover, it identifies the inherent advantages of traditionally economically periphery regions, thereby providing a suitable path for rapid economic and social development.

Future research, however, should consider some limitations. This article, referring to the indicator system established based on the development reports of the nighttime economy industry, acknowledges that while existing research confirms that point of interest (POI) data, nighttime light data, and other factors, to some extent, reflect the current status of a city’s nighttime economic development, there are still shortcomings in comprehensively and completely depicting the spatial characteristics of urban nighttime activities. This represents a direction for future research. Our subsequent research will be based on more accurate, authentic, and open data, aiming to establish a more comprehensive, scientific, and rational research model and relevant theoretical framework for the urban nighttime economy.