Qidu Dist. road accident hotspots and casualties

From 2024/01 to 2026/08, Qidu Dist. recorded 1,189 casualty road accidents, 13 deaths and 1,575 injured, which is 15.0% of Keelung City.

Accidents
1,189
15.0%
Deaths
13
Injured
1,575
Monthly average
37.2
32 months

Qidu Dist. averaged 37.2 accidents a month over this period. Within a day the peak hour is 08, at 11.4% of the total; the night hours from 22:00 to 05:00 account for 7.9%; weekends account for 21.7%. Excluding the month still being filled in, the last 12 complete months had 418 accidents against 496 in the 12 months before, down 15.7%.

Motorcycles account for 53.3% of parties involved. Deaths per thousand accidents is 10.93, above the national average of 4.24. That ratio swings sharply when deaths are few, which is especially true in districts that record few accidents to begin with, so do not draw conclusions from it alone.

Roads appearing most often in Qidu Dist.

Counted by road names appearing in the accident address. Where the address is written as a junction (“A Rd / B Rd”), both roads are counted once.

RoadAccidents
Mingde 1st Rd.178
Guangming Rd.139
Mingde 2nd Rd.98
Bade Rd.89
Shijian Rd.75
National Freeway No. 155
Chongzhi St.51
Mingde 3rd Rd.45
Zizhi St.45
Fuwu St.44

Long roads and busy roads naturally rank high. This is where accidents happened, not danger per kilometre. Road names are automatic romanisation, not official signage.

When accidents happen

HourAccidentsShare
00:00–01:00110.9%
01:00–02:0070.6%
02:00–03:0090.8%
03:00–04:0080.7%
04:00–05:0090.8%
05:00–06:00131.1%
06:00–07:00423.5%
07:00–08:0013010.9%
08:00–09:0013611.4%
09:00–10:00605.0%
10:00–11:00655.5%
11:00–12:00705.9%
12:00–13:00554.6%
13:00–14:00474.0%
14:00–15:00524.4%
15:00–16:00544.5%
16:00–17:00655.5%
17:00–18:001008.4%
18:00–19:00837.0%
19:00–20:00534.5%
20:00–21:00302.5%
21:00–22:00403.4%
22:00–23:00342.9%
23:00–00:00161.3%

By day of week

DayAccidentsShare
Mon18915.9%
Tue18715.7%
Wed18315.4%
Thu17815.0%
Fri19416.3%
Sat14011.8%
Sun1189.9%

Primary causes

Primary causeCountShare
Following too closely13411.3%
Distracted driving: inattention or anxiety12510.5%
Improper left turn897.5%
Signalised junction: turning vehicle failed to yield to through traffic716.0%
Other failure to yield504.2%
Pulling away without checking494.1%

This is the attending officer's assessment, not a court finding.

Collision type

TypeCountShare
Side impact30926.0%
Other29424.7%
Rear-end19516.4%
Same-direction sideswipe13211.1%
Overturn / fall on road534.5%
Opposite-direction sideswipe443.7%

Type is the collision pattern recorded by police; "Other" is how the source data classifies it and cannot be split further.

Lighting and weather

ConditionCountShare
Daytime / lights off54946.2%
Night, lit43536.6%
Night, unlit20517.2%
Clear76664.4%
Rain28624.1%
Cloudy13711.5%

"Daytime / lights off" is the source category "lighting present but off or faulty", which in practice is mostly daytime.

Party types

Party typePartiesShare
Motorcycle1,38753.3%
Car (incl. passenger/goods)81631.3%
Pedestrian27010.4%
Light truck702.7%
Bicycle / slow vehicle331.3%
Heavy truck281.1%

By month

MonthAccidents
2025/0640
2025/0734
2025/0830
2025/0942
2025/1043
2025/1143
2025/1242
2026/0135
2026/0230
2026/0331
2026/0434
2026/0525
2026/0629
2026/07 (partial)28

Months with no recorded accidents are not listed.

The National Police Agency refreshes its rolling data on the 1st and 15th of each month, so the last month is still being filled in and will always read low. Do not compare it with earlier months.

See Qidu Dist. on the map

The map lets you drag the time range, show fatal accidents only, and filter by vehicle type. Zoom in and you can see individual stretches of road.

Open the accident heatmap for Qidu Dist.

Flood potential in Qidu District — the same township in the flood potential tool: simulated flooded area under ten rainfall scenarios.

How to read these numbers

More accidents does not mean more dangerous. Places with more traffic, longer roads and denser population naturally record more. This data has no traffic volume to divide by, so a rate per ten thousand vehicle passes cannot be computed; it can only tell you where accidents actually happened.

The road ranking counts road names appearing in the address. In Taiwan 26% of accident addresses are written as a reference point along a road rather than an exact position, so this ranking is reliable at the level of a road but not at the level of a junction.

Only A1 and A2 accidents, which involve casualties, are included. Property-damage-only collisions are not, so the real number of collisions is higher.

Full limitations

Other districts in Keelung City

Ren’ai Dist. Anle Dist. Zhongzheng Dist. Xinyi Dist. Zhongshan Dist. Nuannuan Dist.  See all