Which Indian cities show the strongest air quality improvement reporting signal?
A comparison of 45 Indian cities covered by Prism Geo.
17 pass the evidence gates and are ranked by exact
air_quality_improvement-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 · 17 city-linked districts
Where the air quality improvement 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
New Delhi leads the normalized comparison at 12.08
tagged reports per 10,000 civic reports. New Delhi leads raw volume
with 359 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
17
passed both evidence gates
Normalized leader
New Delhi
12.08 per 10K
Raw-volume leader
New Delhi
359 tagged reports
Largest recent rise
Ghaziabad
+3.40 per-10K points
National comparison
-0.53
recent per-10K change
Computed findings
What stands out
Normalization changes the picture: New Delhi ranks first by coverage share,
while New Delhi ranks first by article volume.
Coverage is concentrated: New Delhi, Mumbai, Kolkata account for 81.5% of all
qualifying tagged reports in this table.
The aggregate signal moved from 5.35 per 10,000 reports in 2021–2023
to 4.82 in 2024–2025 across the 45-city universe.
Ghaziabad has the largest recent rise among cities with at least ten
tagged reports: +3.40 per-10K points.
Complete ranking
Primary and companion measures
Air Quality Improvement 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.
359 tagged reports appeared within
297,207 civic reports in 2021–2025, equal to
12.08 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 12.17 in 2021–2023 versus 11.93
in 2024–2025. The city's peak normalized year was 2021.
The classifier marked authority or institutional action in 33 reports (9.2%) and conclusion language in 33 (9.2%). 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
4,191 civic reports in 2021–2025, equal to
9.54 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 0.00 in 2021–2023 versus 28.99
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.
8 tagged reports appeared within
11,332 civic reports in 2021–2025, equal to
7.06 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 7.18 in 2021–2023 versus 6.87
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.
11 tagged reports appeared within
17,010 civic reports in 2021–2025, equal to
6.47 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 5.05 in 2021–2023 versus 8.45
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.
13 tagged reports appeared within
20,361 civic reports in 2021–2025, equal to
6.38 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 7.50 in 2021–2023 versus 4.78
in 2024–2025. The city's peak normalized year was 2021.
The classifier marked authority or institutional action in 1 reports (7.7%) and conclusion language in 1 (7.7%). 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
31,282 civic reports in 2021–2025, equal to
5.11 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 8.86 in 2021–2023 versus 1.29
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.
15 tagged reports appeared within
36,239 civic reports in 2021–2025, equal to
4.14 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 5.30 in 2021–2023 versus 2.58
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
12,851 civic reports in 2021–2025, equal to
3.89 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 4.06 in 2021–2023 versus 3.35
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.
3 tagged reports appeared within
8,318 civic reports in 2021–2025, equal to
3.61 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 4.63 in 2021–2023 versus 2.50
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.
8 tagged reports appeared within
23,841 civic reports in 2021–2025, equal to
3.36 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 5.01 in 2021–2023 versus 1.01
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.
4 tagged reports appeared within
16,200 civic reports in 2021–2025, equal to
2.47 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 1.94 in 2021–2023 versus 3.38
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.
29 tagged reports appeared within
119,517 civic reports in 2021–2025, equal to
2.43 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 1.73 in 2021–2023 versus 3.61
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.
5 tagged reports appeared within
27,784 civic reports in 2021–2025, equal to
1.80 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 2.62 in 2021–2023 versus 0.80
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.
3 tagged reports appeared within
22,311 civic reports in 2021–2025, equal to
1.34 per 10,000. The recent-period comparison shows
a broadly stable share of local civic coverage: 1.32 in 2021–2023 versus 1.39
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.
7 tagged reports appeared within
58,199 civic reports in 2021–2025, equal to
1.20 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 2.02 in 2021–2023 versus 0.35
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.
3 tagged reports appeared within
26,153 civic reports in 2021–2025, equal to
1.15 per 10,000. The recent-period comparison shows
a rising share of local civic coverage: 0.69 in 2021–2023 versus 1.73
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.
3 tagged reports appeared within
35,340 civic reports in 2021–2025, equal to
0.85 per 10,000. The recent-period comparison shows
a falling share of local civic coverage: 1.41 in 2021–2023 versus 0.00
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.
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 air_quality_improvement; 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.