Dear Transport Department,

I am conducting independent research on traffic accident blackspot junctions in Hong Kong. I have been analysing Transport Department quarterly blacksite data to understand junction-level road safety patterns. This enquiry is entirely standalone.
Enquiry: Site-Specific Records, Methodology Change, and Traffic Data

Q1 — Site-specific blacksite history for two named junctions
What I already know: Analysis of quarterly blacksite data for 2015-2024 identified two junctions with exceptionally high accident counts:
(a) Chatham Road South × Austin Road × Cheong Wan Road (Tsim Sha Tsui) — 222 total accidents 2015-2024, Hong Kong’s most severe accident hotspot. Peaked at 39 accidents in Q2 2024.
(b) Kweilin Street × Tai Nan Street (Sham Shui Po) — highest pedestrian-involved accident count in Kowloon. 12 pedestrian-involved accidents by Q3 2015 alone.
I have searched TD publications, LegCo replies, and Audit Commission reports and can find no public record of specific improvement works at either junction.
Request: For each of these two junctions, please provide: (a) The years each was listed on the Junction Blacksite list (covering 2004-2024) (b) The dates and nature of any improvement works completed at each location during this period
Why this matters: These are Hong Kong’s worst-performing junctions by accident counts. If improvement works have been completed, citizens should be able to see what was done and whether it was effective. If not, there are urgent public safety implications.

Q2 — Pre-2014 vs post-2014 data methodology
What I already know: Hong Kong Police Force traffic accident statistics and Census & Statistics Department releases divide road traffic accident data at the 2013/2014 boundary. A C&SD feature article (February 2025) covers data from 2014 onwards without explaining why earlier years are excluded. This break affects any attempt to conduct longitudinal analysis of road safety trends.
Request: Please confirm whether TD’s accident data recording methodology or Traffic Information System changed in or around 2013-14. If a change occurred, please provide the relevant methodology documentation or internal note explaining: (a) What changed (b) Why it changed (c) The implications for comparing pre-2014 data with post-2014 data
If no change occurred, please confirm this in writing so the HKPF/C&SD data break can be attributed to their systems rather than TD’s.
Why this matters: Any analysis spanning the pre/post-2014 boundary (including the 20-year blacksite trend analysis I am conducting, covering 2004-2024) requires understanding whether apparent trends reflect real changes in road safety or artefacts of data collection methodology. This is a basic requirement for evidence-based policy evaluation.

Q3 — Junction-approach traffic volume data (AADT)
What I already know: The Annual Traffic Census (https://atc.td.gov.hk) publishes road-section traffic counts at approximately 1,600 stations. This is valuable data but does not capture junction-approach volumes or turning-movement counts. Junction-level safety analysis requires approach-specific data to calculate accident rates per vehicle movement, not just raw accident counts.
Request: Please provide Annual Average Daily Traffic (AADT) data, by junction approach, for the junctions listed in the Annex below, for available years 2014-2024. If approach-level counts or turning-movement data are not routinely recorded, please confirm: (a) What traffic volume data is held at or near these junctions (b) At what spatial resolution (junction-level, road-section, district-level) (c) Whether the data can be made available under the Code on Access to Information
Why this matters: LCQ13 (30 July 2025, https://www.info.gov.hk/gia/general/2025...) attributes accidents mainly to driver behaviour. Without traffic volume data, it is impossible to test whether high-accident junctions simply have more traffic — a basic methodological requirement for any road safety analysis. The Code on Access to Information (https://www.access.gov.hk/) establishes the public’s right to government-held information that informs policy decisions.
Annex: Junctions for AADT Data Request
Chatham Road South × Austin Road × Cheong Wan Road
Tsuen Tsing Interchange × Tsuen Wan Road
Waterloo Road × Argyle Street × Princess Margaret Road
Nathan Road × Mong Kok Road
Kweilin Street × Tai Nan Street 6-21. [To be completed — the 21 junctions with the highest accident counts and longest blacksite history]

Yours faithfully,
Winnie FU