Which Indian cities show the strongest train derailment reporting signal?
A comparison of 45 Indian cities covered by Prism Geo.
18 pass the evidence gates and are ranked by exact
train_derailment-tagged news reports per 10,000 civic reports.
This measures the prominence of a topic in coverage—not real-world incidence, prevalence, or safety.
Geographic view · 18 city-linked districts
Where the train derailment reporting signal is concentrated
Every highlighted polygon is the 2021 administrative district containing a ranked city's coordinate. Color shows the city's reporting-signal rank band—not a measured district-wide rate, prevalence, or severity assessment.
Boundary geometry: geoBoundaries, ADM2, 2021 representation, Open Data Commons Open Database License 1.0. Districts are visual locators for city-ranked data and may not reflect later administrative changes.
How to read this ranking
Kanpur leads the normalized comparison at 87.75
tagged reports per 10,000 civic reports. Kanpur leads raw volume
with 65 reports. The two leaders can differ because normalization
controls for each city's overall civic-news footprint.
Cities evaluated
45
existing India profiles
Cities ranked
18
passed both evidence gates
Normalized leader
Kanpur
87.75 per 10K
Raw-volume leader
Kanpur
65 tagged reports
Largest recent rise
Kanpur
+203.65 per-10K points
National comparison
+1.30
recent per-10K change
Computed findings
What stands out
Normalization changes the picture: Kanpur ranks first by coverage share,
while Kanpur ranks first by article volume.
Coverage is concentrated: Kanpur, Mumbai, Bhubaneswar account for 55.4% of all
qualifying tagged reports in this table.
The aggregate signal moved from 2.44 per 10,000 reports in 2021–2023
to 3.74 in 2024–2025 across the 45-city universe.
Kanpur has the largest recent rise among cities with at least ten
tagged reports: +203.65 per-10K points.
Complete ranking
Primary and companion measures
Train Derailment reporting signal by city
The primary rank uses tagged reports per 10,000 civic reports. Raw rank,
recent change, and active years show whether a result depends on city size, timing, or a narrow burst.
65 tagged reports appeared within
7,407 civic reports in 2021–2025, equal to
87.75 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 6.73 in 2021–2023 versus 210.38
in 2024–2025. The city's peak normalized year was 2024.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
15 tagged reports appeared within
3,489 civic reports in 2021–2025, equal to
42.99 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 58.16 in 2021–2023 versus 0.00
in 2024–2025. The city's peak normalized year was 2022.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
15 tagged reports appeared within
5,388 civic reports in 2021–2025, equal to
27.84 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 23.03 in 2021–2023 versus 36.55
in 2024–2025. The city's peak normalized year was 2025.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
29 tagged reports appeared within
17,398 civic reports in 2021–2025, equal to
16.67 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 23.76 in 2021–2023 versus 7.78
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
6 tagged reports appeared within
3,699 civic reports in 2021–2025, equal to
16.22 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 13.97 in 2021–2023 versus 19.33
in 2024–2025. The city's peak normalized year was 2024.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
4 tagged reports appeared within
3,256 civic reports in 2021–2025, equal to
12.29 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 19.35 in 2021–2023 versus 0.00
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
5 tagged reports appeared within
4,191 civic reports in 2021–2025, equal to
11.93 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 0.00 in 2021–2023 versus 36.23
in 2024–2025. The city's peak normalized year was 2024.
The classifier marked authority or institutional action in 3 reports (60.0%) and conclusion language in 3 (60.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
16 tagged reports appeared within
27,784 civic reports in 2021–2025, equal to
5.76 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 7.22 in 2021–2023 versus 3.99
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
60 tagged reports appeared within
119,517 civic reports in 2021–2025, equal to
5.02 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 4.39 in 2021–2023 versus 6.09
in 2024–2025. The city's peak normalized year was 2024.
The classifier marked authority or institutional action in 1 reports (1.7%) and conclusion language in 1 (1.7%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
8 tagged reports appeared within
16,169 civic reports in 2021–2025, equal to
4.95 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 8.46 in 2021–2023 versus 1.27
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
8 tagged reports appeared within
26,153 civic reports in 2021–2025, equal to
3.06 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 4.12 in 2021–2023 versus 1.73
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
4 tagged reports appeared within
17,010 civic reports in 2021–2025, equal to
2.35 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 1.01 in 2021–2023 versus 4.22
in 2024–2025. The city's peak normalized year was 2025.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
3 tagged reports appeared within
13,958 civic reports in 2021–2025, equal to
2.15 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 0.00 in 2021–2023 versus 5.59
in 2024–2025. The city's peak normalized year was 2024.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
3 tagged reports appeared within
16,200 civic reports in 2021–2025, equal to
1.85 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 1.94 in 2021–2023 versus 1.69
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
6 tagged reports appeared within
35,340 civic reports in 2021–2025, equal to
1.70 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 2.34 in 2021–2023 versus 0.71
in 2024–2025. The city's peak normalized year was 2021.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
5 tagged reports appeared within
31,282 civic reports in 2021–2025, equal to
1.60 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 3.17 in 2021–2023 versus 0.00
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 1 reports (20.0%) and conclusion language in 1 (20.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
6 tagged reports appeared within
58,199 civic reports in 2021–2025, equal to
1.03 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 2.02 in 2021–2023 versus 0.00
in 2024–2025. The city's peak normalized year was 2023.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
20 tagged reports appeared within
297,207 civic reports in 2021–2025, equal to
0.67 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 0.39 in 2021–2023 versus 1.12
in 2024–2025. The city's peak normalized year was 2025.
The classifier marked authority or institutional action in 0 reports (0.0%) and conclusion language in 0 (0.0%). These are article-level signals: repeated coverage can
refer to the same underlying event, and the markers do not prove official resolution.
Annual values use all covered cities, including cities that do not pass the page-level ranking gate.
Transparent limits
How Prism Geo built this ranking
Universe: 45 Indian cities with an existing Prism Geo city manifest; conservative legacy names such as Bangalore/Bengaluru are merged.
Window: complete calendar years 2021–2025. Partial 2026 data is excluded.
Signal: the exact source tag train_derailment; topic variants are not silently combined.
Score: tagged reports ÷ all civic reports × 10,000; raw volume is shown alongside it.
Coverage, action, and conclusion figures are classifier outputs. Multiple reports can concern one event, and absence of coverage is not absence of a problem.
Source: marketai.civic_geo_agg. This dataset supports comparison of reporting patterns only; it cannot establish actual case counts, risk, prevalence, or official performance.
Need to go deeper than a city?
Fliar Prism answers any market, company or research question with the
same depth — real-time, across every industry.