Bali's air is tracked today by a small network of independent sensors. This brief compiles what they show: PM2.5 readings across the island, how they compare to international guidelines, and where more monitoring would add value.
In April 2026, Bali's regulations on organic waste disposal were tightened. The Suwung landfill, long the island's primary waste endpoint, stopped accepting organic material. In the weeks that followed, residents across Ubud, Kerobokan, Canggu and Jimbaran reported a marked increase in the smell and visible presence of burning waste — particularly in the early morning and after dark.
This document sets out what the available air quality monitoring records can and cannot tell us about that change. It is written for journalists, researchers, and residents who want to verify the numbers themselves.
Public historical data exists for only a handful of Bali air quality monitors. The AQICN network preserves records for the Denpasar government KLHK station (548 days, offline since August 2025) and the Kerobokan community sensor (469 days, now reporting again through IQAir). OpenAQ preserves CSVs for three additional private AirGradient sensors in Ubud, Balangan and Kopernik — all now offline as well.
Two PurpleAir units in Klungkung and Jimbaran are still reporting. Between them, they cover roughly two years of continuous data.
Coverage across Bali is limited. There is no multi-year record for Sanur, North Bali, or East Bali. The current network is built from independently operated devices, voluntarily shared through platforms with open APIs.
This brief's purpose is to make that network legible — showing what it covers, what readings it has produced, and where additional sensors would close the largest gaps.
"The sensors that exist tell a consistent story. Whether that story extends to the whole island is the question that more monitoring would answer."— Section 04 · Historical Data
| Sensor | Network | Area | Days of record | Mean PM2.5 | % > WHO 24h | Status |
|---|---|---|---|---|---|---|
| Kerobokan / Umalas | AQICN · GAIA | Canggu corridor | 469 | 21.9 | 68% | Live · IQAir |
| Denpasar / KLHK | AQICN · Govt. | Central Denpasar | 548 | 15.1 | ~40% | Offline Aug 2025 |
| Kopernik / Ubud | OpenAQ · AirGradient | Ubud | 36 | 31.2 | 92% | Offline Mar 2026 |
| Ubud / Rozendal | OpenAQ · AirGradient | Ubud | 167 | 17.8 | 57% | Offline Dec 2025 |
| Balangan | OpenAQ · AirGradient | Bukit peninsula | 26 | 15.4 | 46% | Offline Aug 2025 |
| Klungkung | PurpleAir | East Bali | 687 | — | — | Active |
| Jimbaran | PurpleAir | South coast | 728 | — | — | Active |
Most of what we know about Bali's air comes from a small number of sensors, some of which are no longer active. This chapter summarises what they captured — and what is still reporting today.
The Lumintang station in Denpasar recorded continuously for 548 days from September 2023 to August 2025. Its final two days showed daily medians of 45.7 and 75.0 µg/m³ — firmly in the US EPA "Unhealthy" band — before the feed went quiet.
Its 548-day archive is preserved on AQICN and forms part of the historical record in this brief. The station is listed as under maintenance; its full daily dataset is downloadable from the archive link in Chapter IV.
The Kerobokan community sensor, part of the AQICN / GAIA network, recorded 469 days from November 2024 to March 2026. Its public AQICN feed went quiet on the 15th of March — two weeks before organic waste was banned from the Suwung landfill — though the same device has since resumed reporting through IQAir.
The sensor is privately owned. The record it left behind during that window — daily medians averaging 21.9 µg/m³, with 68% of days exceeding the WHO 24-hour guideline — remains one of the most detailed openly downloadable windows we have into the air above south Bali.
The regulation reflects a legitimate environmental goal. This chapter looks at how it intersected with the air quality monitoring record and what the available data shows in the months that followed.
On the first of April, Denpasar banned organic waste from the Suwung landfill — Bali's largest waste endpoint — as part of a plan to close the site by August 2026. The stated aim is a legitimate environmental objective: push households toward source-separation and reduce landfill dependency.
Implementation has outpaced infrastructure. Without adequate composting or collection alternatives in place, residents have turned in some cases to burning waste in backyards, roadsides, and ravines, and in others to dumping it in the rivers that feed the west-coast catchments.
Denpasar cleanup crews have reportedly removed approximately seven tons of waste from rivers daily, according to local reporting citing the city's Public Works office (PUPR). What burns at roadside does not appear in any such tally. It appears only in the air.
Mixed-waste burning releases fine particulate matter, dioxins, furans, polycyclic aromatic hydrocarbons, and heavy metals. PM2.5 — particles smaller than 2.5 micrometers — penetrates deep into the lungs and crosses into the bloodstream. Long-term exposure is associated with respiratory and cardiovascular disease and with measurable reductions in life expectancy.
The point is straightforward: understanding the air-quality effect of a major waste-management transition — in real time, as it unfolds — requires sensors on the ground. Adding more monitors now would mean having a record to look back on.
Cited reporting
Hey Bali News · Bali's Organic Waste Ban Backfires · April 2026
Bali.live · No Sorting, No Collection · April 2026
What can be shown, with a downloadable receipt attached to every number.
548 days of daily medians, Sep 2023 — Aug 2025. Government-operated, central Denpasar.
Mean daily PM2.5 15.1 µg/m³ · 3.0× WHO annual guideline · roughly 40% of days exceeded the 24-hour guideline.
469 days of daily medians, Nov 2024 — Mar 2026. Community-operated, Canggu corridor.
Mean daily PM2.5 21.9 µg/m³ · 4.4× WHO annual · 68% of days exceeded the 24-hour guideline.
36 days merged from two co-located AirGradient units, Oct 2025 — Mar 2026. Surfaced via OpenAQ.
Mean daily PM2.5 31.2 µg/m³ · 6.2× WHO annual · 92% of days exceeded the 24-hour guideline.
167 days, Jun — Dec 2025. AirGradient unit near central Ubud; OpenAQ.
Mean daily PM2.5 17.8 µg/m³ · 3.6× WHO annual · 57% of days exceeded the 24-hour guideline.
26 days, Jul — Aug 2025. Bukit peninsula. Sensor went offline after three weeks.
Mean daily PM2.5 15.4 µg/m³ · 46% of days exceeded the 24-hour guideline.
687 / 728 days and counting. Two PurpleAir community units on the southern coast and in East Bali respectively.
Klungkung shows a pattern similar to Kerobokan — moderate baseline with daily peaks reaching "Unhealthy for Sensitive Groups." Jimbaran's coastal readings stay close to background.
Even taken together, these five historical records and live sensors cover a limited slice of Bali. There is no multi-year record for Sanur, Amed, North Bali, or the eastern regencies. The longest continuous dataset — Kerobokan, fifteen months — shows a consistent pattern above WHO guidelines. Whether that pattern holds across the whole island is the question that a broader network would answer.
Bali's public air-quality infrastructure, measured against the places it is most frequently compared to.
| Jurisdiction | Government AQ stations | Population | Stations per million |
|---|---|---|---|
| Bali (today) | 1 | 4.4 million | 0.2 |
| Hong Kong | 18 | 7.5 million | 2.4 |
| Bangkok, Thailand | ~70 | 10.7 million | 6.5 |
| London, UK | 100+ | 9.0 million | 11+ |
The gap in coverage is not marginal. To reach even Hong Kong's density, Bali would need approximately eleven additional public government-class stations. To reach Bangkok's, roughly twenty-eight. At London's ratio, fifty.
Four practical, low-cost steps that any actor — foundation, business, community group, or individual — could take or fund today.
The Lumintang station produced nearly two years of continuous data before going offline in August 2025. Restoring or replacing it would add a valuable anchor point to the network — central Denpasar is the island's most densely populated area.
Low-cost open-hardware monitors such as AirGradient (US$195) and PurpleAir (US$229) can fill coverage gaps immediately. Five to ten units across populated Bali would transform visibility at a total cost well under US$3,000.
Any operational sensor — hospitality, industrial, community, or official — that feeds an open platform such as OpenAQ, AQICN, or PurpleAir immediately adds to the public record. The cost is near-zero; the benefit is a shared dataset everyone can verify.
As Bali's waste infrastructure evolves, live air-quality data is the clearest feedback available. More sensors mean a faster signal — and a record that will be useful long after the transition is complete.