Urban Complexity and the Dynamic Evolution of Urban Land Functions in Yiwu City: A Micro-Analysis with Multi-Source Big Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In their manuscript titled "Urban complexity and dynamic evolution of urban land functions in Yiwu city: A micro-analysis with multi-source big data," the authors seek to identify the urban functions and distribution characteristics of the main urban area based on POI (point of interest) data and urban form data. The quality of paper is good. However, I’ve included several comments below.

 

The authors have used some ambiguous concepts. For example, they said, “With the continuous deepening of urbanization in China”. This sentence is unclear. What is the meaning of deepening? Is it an increase in urbanization or rapid urbanization? However, the urbanization rate in China is currently 66.2, which is low at the global level. Please fill this gap, and don't assume the readers will do so for you.

 

There seems to be a gap in what the authors are trying to say about the traditional urban development model with "incremental expansion." There is no traditional urban development model because urban development is an evolutionary process. I understand that there were traditional models such as civic design, rational approach, participatory approach, UN programs, and initiatives, but currently the traditional models of urban development have mostly changed. In China, there is real progress towards smart cities and the use of technology in managing services, urban resilience, and sustainable development.

 

In most cases, the term "incremental expansion" is used with unplanned urban development (i.e., informal settlements and slums). It is incremental because people do not have sufficient funds to build during a single phase. Please reassess this issue. Otherwise, explain what you mean by" incremental expansion".

Do you have evidence that the complex types of land use in urban areas lead to undesirable outcomes? The complexity of mixed land use often leads to saving time and costs.

The authors did not provide sufficient literature. They should provide previous studies with their conclusions. However, the literature should present other international practices in this context.

The authors have selected Yiwu City as a case study. How do the authors describe this city globally? For instance, is this city mentioned in the international indices in terms of business, economy, trade, technology, industry, etc.?

 

Author Response

Reviewer 1#:

In their manuscript titled "Urban complexity and dynamic evolution of urban land functions in Yiwu city: A micro-analysis with multi-source big data," the authors seek to identify the urban functions and distribution characteristics of the main urban area based on POI (point of interest) data and urban form data. The quality of paper is good. However, I’ve included several comments below.

1. The authors have used some ambiguous concepts. For example, they said, “With the continuous deepening of urbanization in China”. This sentence is unclear. What is the meaning of deepening? Is it an increase in urbanization or rapid urbanization? However, the urbanization rate in China is currently 66.2, which is low at the global level. Please fill this gap, and don't assume the readers will do so for you.

Response: Okay. We have revised this sentence.

Over the past 10 years, although China's economic growth has slowed down, urbanization still shows a rapid growth trend. China's urbanization rate rose from 51.3% in 2011 to 65.2% in 2022, an increase of 13.9%. With China's rapid urbanization, ……

2. There seems to be a gap in what the authors are trying to say about the traditional urban development model with "incremental expansion." There is no traditional urban development model because urban development is an evolutionary process. I understand that there were traditional models such as civic design, rational approach, participatory approach, UN programs, and initiatives, but currently the traditional models of urban development have mostly changed. In China, there is real progress towards smart cities and the use of technology in managing services, urban resilience, and sustainable development.

In most cases, the term "incremental expansion" is used with unplanned urban development (i.e., informal settlements and slums). It is incremental because people do not have sufficient funds to build during a single phase. Please reassess this issue. Otherwise, explain what you mean by" incremental expansion".

Response: Okay. We have deleted the word “traditional”, and modified "incremental expansion" as “urban construction land expansion or urban spatial expansion”, “stock renewal” as “urban renewal or urban smart growth”.

3. Do you have evidence that the complex types of land use in urban areas lead to undesirable outcomes? The complexity of mixed land use often leads to saving time and costs.

Response: Okay. We have made the following additions in the text.

City function zoning is to arrange various material elements (such as factories, warehouses, houses, etc.) in the city according to the functional requirements, to form an organic whole with mutual connections and reasonable layout, and to create a good environment and conditions for various activities in the city. It is an important method to determine the form of land use and space layout according to the principle of functional zoning. In previous studies, the identification of urban functional areas was mostly based on horizontal scale and single land use [1-4]. However, in the process of rapid urbanization in developing countries, the trend of vertical, multi-functional and mixed urban development has become increasingly prominent. Mixed land use (MLU) is still one of the most recommended indicators for successful urban planning and urban regeneration. MLU may play a great role in solving urban problems such as pollution and over consumption of lands. It can usually save commuting cost and enhance urban vitality [5], provide more local jobs and business options to enhance the economy of the community and add vibrancy to environment [6]. However, on one hand, MLU may have a negative impact on land use compatibility (LUC) [5], urban planners are hard to find the balance between MLU and LUC. On the other hand, MLU will also increase the difficulty of urban functional area identification.

4. The authors did not provide sufficient literature. They should provide previous studies with their conclusions. However, the literature should present other international practices in this context.

Response: Okay. We have added some literature and accounts.

In OECD countries, population density is used to identify urban cores, and trav-el-to-work flows is used to identify the hinterlands whose labour market is highly integrated with the cores [7]. By introducing the concept of latent activity trajectory (LAT), which captures socioeconomic activities conducted by citizens at different locations in a chronological order, Yuan et al. [8] proposed a data-driven framework to discover urban functional zones in a city, developed a topic-modeling-based approach to cluster the segmented regions into functional zones, and identified the intensity of each functional zone using Kernel density estimation. Based on building-level social media data, Chen et al. [9] used a dynamic time warping distance based k-medoids method to group buildings with similar social media activities into urban functional areas. Zhang et al. [10] used hierarchical semantic cognition (HSC) to classify urban functional zones, which relies on geo-graphic cognition and considers four semantic layers, i.e., visual features, object categories, spatial object patterns, and zone functions, as well as their hierarchical relations. Their experimental results indicate that this method can produce more accurate results than Support Vector Machine (SVM) and Latent Dirichlet Allocation (LDA). In addition, some scholars evaluated the polycentric index of a city and its degree according to the urban functional area [11], or studied the social cohesion of different functional urban areas [12].

5. The authors have selected Yiwu City as a case study. How do the authors describe this city globally? For instance, is this city mentioned in the international indices in terms of business, economy, trade, technology, industry, etc.?

Response: Okay. We have made the following additions in the text.

Yiwu is a global commercial city. In 2022, the city has a licensed market construction area of 6.11 million square meters, market operating households of 84,800, market employees of 230,000, operating more than 1.8 million kinds of commodities, the total turnover of 232.283 billion CNY, an increase of 6.9%. Yiwu maintains stable trade relations with 219 countries and regions in the world, and 28,554 foreign investors entered the city in 2022. The total volume of imports and exports was 478.80 billion CNY in 2022, up 22.7% year on year. Among them, the export was 431.64 billion CNY, up 18.0% year on year; Imports reached 47.16 billion CNY, up 93.5% year on year. The turnover of cross-border e-commerce was 108.35 billion CNY. It can be seen that Yiwu's urbanization model is a typical global urban development model of industry and trade integration, which combines market clusters and industrial development.

Reviewer 2 Report

Comments and Suggestions for Authors

Article review: Urban complexity and dynamic evolution of urban land functions in Yiwu city: A micro-analysis with multi-source big data

The authors analyze urban complexity and determine the evolution of urban land functions in Yiwu city. "Presumably" - because it is not clear from the content - the goal was to use machine learning to identify urban functions. It is an interesting topic, but the article needs improvement.

Comments that followed in order:

Abstract - First thing one reviews - no purpose of the paper.

 Introduction contains a literature review, but unfortunately, very little regarding the research topic. There is no detailed explanation of issues such as POI, TF-IDF algorithm, and a machine learning classification algorithm. The purpose of the work needs to be specified.

 Data - if I understand correctly, extracted here the functions that buildings perform. Unfortunately, it is so described that you have to guess it.

Methods - fortunately, there is a method scheme because the description leaves much to be desired.

Results - 4.1 Identification results of urban functional areas - question about the effect of scale - can the results be applied to individual properties?

In general, the developed method shows how to identify the so-called urban areas using the adopted data interestingly. The results are promising, and the process of automating the identification of functions seems original and interesting. Unfortunately, the description of the method is unreadable. And here, one should refer to the need for a clear purpose of the work. A little clarifies this issue in the discussion, but more time must be spent. In the conclusions, one might be tempted, for example, to make some prediction of changes.

Author Response

Reviewer 2#:

Article review: Urban complexity and dynamic evolution of urban land functions in Yiwu city: A micro-analysis with multi-source big data

The authors analyze urban complexity and determine the evolution of urban land functions in Yiwu city. "Presumably" - because it is not clear from the content - the goal was to use machine learning to identify urban functions. It is an interesting topic, but the article needs improvement.

Comments that followed in order:

1. Abstract - First thing one reviews - no purpose of the paper.

Response: We added the purpose of this paper in Abstract.

The purpose of this paper is to automatically identify the spatial heterogeneity and temporal variation of urban land use functions in the context of complex urban systems.

2. Introduction contains a literature review, but unfortunately, very little regarding the research topic. There is no detailed explanation of issues such as POI, TF-IDF algorithm, and a machine learning classification algorithm. The purpose of the work needs to be specified.

Response: Okay. We added some literature regarding the research topic. At the same time, we added detailed explanation of issues such as POI, TF-IDF algorithm, and a machine learning classification algorithm, and the work objectives of adopting these methods is clarified.

Delineating urban functional zones is critical to understanding urban dynamics, evaluating planning strategies, and formulating supportive policies. However, initial studies using point of interest (POI) data did not adequately consider the relationship be-tween POI categories in a spatial context, nor could they provide a direct way to classify urban functional areas. If POI, TF-IDF algorithm and machine learning classification algorithm are integrated, the following work objectives can be achieved: (1) The spatial relationship between POI data and urban form data can be established by using geographic information system (GIS) spatial connection method, and the POI data can be given spatial entity attributes. Therefore, in recent years, POI data with accurate geographical location and detailed attributes have become the main data source for exploring urban func-tional zoning from bottom to up perspective [13-16]. (2) TF-IDF (term frequency-inverse document frequency) algorithm can be used to identify and extract important words in text data and classify urban functions. Combining POI data with TF-IDF algorithm to analyze functional distribution and activity characteristics of different regions can improve the accuracy of urban function classification results [17-20]. (3) Machine learning classification algorithms can automatically learn rules based on the training data provided to identify and classify city functions [21,22].

3. Data - if I understand correctly, extracted here the functions that buildings perform. Unfortunately, it is so described that you have to guess it.

Response: Okay. We supplemented the description of building form data extraction.

Different types of land use usually have different building layout forms, which are reflected in various form indicators. In the research, it is necessary to collect all kinds of building form index data (such as the highest building height, average building height, average building base area, plot ratio, etc.) to establish the database. The architectural form index of each plot will change with the change of land use function. It is combined with the business weighted attribute to form a complete weighted attribute. Finally, the combination relationship of various morphological indicators in different land use types is found through machine learning to improve the accuracy of identification.

4. Methods - fortunately, there is a method scheme because the description leaves much to be desired.

Response: Okay. We added the following description in Methods section.

The identification of urban land functions only based on the business form POI data ignores the correlation between urban building form and urban land use. Coupled with the single data dimension, relying on the business form data classification will still lead to considerable errors in the final identification results, and cannot achieve the practical accuracy of identifying urban land use subcategories. Therefore, this study combines the business form POI data with the building form data, and realizes the transformation from the business form data classification to the fine identification of urban land use subcategories through artificial intelligence technology.

5. Results - 4.1 Identification results of urban functional areas - question about the effect of scale - can the results be applied to individual properties?

Response:

The identification results of urban functional areas can be applied to individual properties. As long as there is enough sample data for the machine to learn, the data set can be classified into functional areas through algorithm selection and model parameter adjustment.

6. In general, the developed method shows how to identify the so-called urban areas using the adopted data interestingly. The results are promising, and the process of automating the identification of functions seems original and interesting. Unfortunately, the description of the method is unreadable. And here, one should refer to the need for a clear purpose of the work. A little clarifies this issue in the discussion, but more time must be spent. In the conclusions, one might be tempted, for example, to make some prediction of changes.

Response: Okay.

(1) In Introduction section, we have added some specific work objectives.

(2) In Methods section, we have made some supplementary explanations.

(3) In Discussion section, we further clarified 3 key issues that the study seeks to address, and pointed out the application value of this paper.

(4) In Conclusions section, we have added some prediction of changes.

Since the late 1970s, Yiwu has always adhered to the development strategy of "thriving commerce and establishing a city", successfully completed the leap from a regional commercial city to a national commercial city, and then to a global commercial city, be-coming a unique sample of global cities. In addition to a group of leading entrepreneurs and advancing institutional innovations, the main reasons for Yiwu's success lie in the diversification of industries and products, the complexity of the economy, and the wide connectivity of capacity networks. Yiwu is integrated into a global urban network that links its local industries to global economic activities and the flow of resources, capital, knowledge, talent and information. However, Yiwu also faces some practical challenges: (1) The land use layout is highly mixed, and the urban space quality is not high. The "four-and-a-half story" houses (i.e. living upstairs, renting downstairs, used as factories or warehouses) in the downtown area of Yiwu cover an area of about 10 square kilometers, and the functions of commerce, office, residence, logistics, warehousing, and production are highly mixed in the internal spaces. Excessive functional mixing has led to a series of problems such as freight transportation through the city, logistics throughout the city, traffic interference, lack of public space and service supporting facilities, etc., affecting the overall image and environmental quality of Yiwu. (2) The commercial space is saturated and the centrality is insufficient. The total commercial construction area of Yiwu city is 912,000 square meters, 0.48 square meters per capita, and the total volume tends to be saturated. From the perspective of spatial distribution, each business district basically has only one large commercial complex, which is difficult to form the agglomeration effect of commerce. (3) Scattered and inefficient distribution of industrial land.

According to the theory of complexity science, cities are complex systems formed by the interaction of agents such as people, institutions, markets and networks [48-50]. Moreover, cities interact with other cities at the regional, national and global levels, and may always be accompanied by various emergent outcomes. The change of trade pattern and economic conditions will affect Yiwu's market hub status. In the face of challenges such as the impact of global climate change, security threats, epidemics and emergencies, the competition between residential land, commercial and retail space, industrial land, transportation land, green land, recreational land and other types of land in Yiwu will be further intensified. The competition between residential land, commercial and retail space, industrial land, transportation land, green land and leisure and entertainment land in Yiwu City will be further intensified. The development trend will be moderately compact residential space, high quality and efficient commercial space, reduction and intensive use of industrial space, and expansion and quality improvement of public space. In order for Yiwu City to continue to occupy the market hub position in this emergent network, it must quickly and flexibly seize the first-mover advantage, actively connect the global networks, properly manage the complexity of the economy, ensure the balance between various types of land, and make the market gathering place become the gathering place of highly skilled talents and innovative capabilities. In addition, with the rapid development of digital technology, big data continues to emerge and present diversified sources, artificial intelligence technology will also be more widely used. Yiwu can make full use of big data and artificial intelligence technology to deeply study complex urban systems, simulate and plan future scenarios of the city, improve the efficiency of urban development, service efficiency and spatial quality, and make Yiwu more vibrant, more harmonious, more resilient and more livable.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors took into account all comments. The article has benefited from its current form. I recommend it for printing.