Clustering Analysis of the Spatio-Temporal On-Street Parking Occupancy Data: A Case Study in Hong Kong

Round 1

Reviewer 1 Report

The analysis presented in this paper is complete and convincing. What I suggest to improve the paper any further:

i)                    Analyze the road pricing more in depth within your analysis, above all highlighting the possible gains in terms of social costs, related to both pollution and carbon emissions. See the papers: Anas, A., Lindsey, R. (2011). Reducing urban road transportation externalities: Road pricing in theory and in practice. Review of Environmental Economics and Policy 5(1), pp. 66-88; Cavallaro F., Giaretta F., Nocera S. (2018): The potential of Road Pricing Schemes to Reduce Carbon Emissions. Transport Policy 67: 85-92;

ii)                   Add a paragraph somewhere about the cost estimations and implications of your method. Such theme is relatively complex, so I suggest to gain materials from here: Browne, D., O'Mahony, M., Caulfield, B. (2012). How should barriers to alternative fuels and vehicles be classified and potential policies to promote innovative technologies be evaluated? Journal of Cleaner Production 35, pp. 140-151; Cavallaro, F., Nocera, S. (2022). Are transport policies and economic appraisal tools aligned in evaluating road externalities? Transportation Research Part D: Transport and Environment 106, 103266;

iii)                 Discuss possible generalizations of the method outside of Hong Kong

Author Response

Please see the attached file.

Best,

Wei

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors analyze on-street parking occupancy in Hong Kong and highlight some specific patterns via direct observation. They also associate that to  land-use factors with the goal to contribute to the decision-making process and strategies development on this topic.

The manuscript is interesting, well written and few amendments are suggested to improve its readability, as follows:

1) in the manuscript different places/streets/areas are often mentioned, but for those not familiar with HK it is hard to understand, sometimes, the differences stressed by the authors. A very short description could help and possibly some pictures showing the differences in parkings and land use/streetscape

2) Figure 7, too many lines, too close one to another; make it bigger so to be more readable; ditto fig. 9 and 10

3) Figure 8, the color hues are too similar so it is difficult to understand the differences on some bars 

4) in the concusions it is reported: "and higher granular studies of 513 block or road level analysis can be more insightful". This is understandable, but since the goal of the manuscript is to help decision-makers, very high granularity might be not needed, being the decisions tale at city/district/region level. Or are there specific regulatory tools at street level in HK?

5) Eq 1, what is P? is it %?, please, clarify. Please note that P is also used in eq 4

6) the symbol R (446-448): use symbols/fonts in eq. editor (or similar) please

Author Response

Please see the attached file.

Best,

Wei

Author Response File: Author Response.pdf

Reviewer 3 Report

General comments

The paper examines the spatial and temporal pattern in Hong Kong, based on big data collected from new parking meters. Compared to other studies on on-street parking based on data from surveys or other sources (video recording, GPS traces, transactions data), the proposed methodology has the advantage that is based on real-time data collection provided by the new technology of the parking meters. The data mining includes t-SNE and K-means algorithms to visualise and cluster the parking data from over six thousand parking meters operating in Hong Kong. Considering the negative effects of cruising for on-street parking, the presented study can contribute to a better understanding of the parking pattern and support the development of parking management strategies.

The first part of the paper includes a synthetic review of on-street parking issues and a description of the on-street parking system in Hong Kong. The data processing and the steps of the developed methodology are clearly presented. The spatial and temporal distributions of on-street parking patterns are analysed based on occupancy rates computed based on data collected in July 2021, each day, every 5 minutes. The processed data are accurately discussed and interpreted for all the 3 regions and 18 districts in Hong Kong. The limitations of the study are well emphasised.

In conclusion, the paper is valuable through the methodology developed for processing and interpreting big data provided by new technologies included in the parking meters (even from a scientific perspective, the study brings no meaningful contributions - in my opinion). The presented methodology increases the accuracy of spatial and temporal parking pattern analysis and could provide valuable practical insight into the development of parking and traffic management systems.

Minor remarks:

The reference list needs revision. Some references are missing. E.g., Van der Maaten and Hinton [48] – line 268; McQeen [49] – line 293.

In section 3.1. & 3.2, for consistency and correctness of the equations, the range for used indices should be specified. Please revise the eq. (8).

Author Response

Please see the attached file.

Best,

Wei

Author Response File: Author Response.pdf

Reviewer 4 Report

In this research, authors performed cluster analysis on parking occupancy data. The paper is well written and organised. 

It would be interesting if authors elaborate the results more in comparison of results with previous studies. 

It is suggested to include policy implications section before conclusion and provide discussion on limitations of study at the end of conclusion. 

Author Response

Please see the attached file.

Best,

Wei

Author Response File: Author Response.pdf