NEA addresses questions over veracity of PSI, as alternative air quality readings emerge

SINGAPORE — The National Environment Agency (NEA) has moved to dispel questions over the veracity of its 24-hour Pollutant Standards Index (PSI) readings, as growing numbers of online users refer to alternative air quality readings online.

In some cases these websites report readings that may appear to indicate that the air quality is poorer than that suggested by NEA for the same period.

The agency cautioned against relying on these alternative measures and said, for example, that some may be based on air-quality sensors bought off the shelves.

Its remarks came as the haze situation here showed few signs of abating, with 24-hour PSI readings on Thursday (Sept 19) ranging between the "moderate" and "unhealthy" range, with readings of 97-to-110 as at 6pm. The 24-hour PSI dropped to 79-to-87 — the higher end of the moderate range — at 11pm. 

WHY READINGS ON AQICN.ORG ARE 'INACCURATE'

NEA referred to readings on a popular website measuring air quality —aqicn.org — and said that it was “incorrect” to compare those with its 24-hour PSI.

In a telephone interview with TODAY on Thursday, Dr Khairunnisa Yahya, a senior scientific officer who works at the Pollution Control Department at NEA, said that the website uses the incorrect Particulate Matter 2.5 (PM2.5) reading to derive its final air quality index.

Dr Khairunnisa, who has a doctorate degree in atmospheric science, said that instead of using the 24-hour average PM2.5 readings, the website is using the hourly PM2.5 readings. 

“The Air Quality Index (AQI) by the United States Environmental Protection Agency (EPA) advises that you’re supposed to use the average 24-hour PM2.5 readings (to measure air quality). But (aqicn.org) is using NEA’s 1-hour PM2.5 (to measure air quality), which is not accurate.”

“You’re not supposed to use the 1-hour PM2.5 reading. If you took the 1-hour PM2.5, compared to the 24-hour average PM2.5 readings… of course it will be higher,” Dr Khairunnisa said.

She explained that using the 1-hour PM2.5 readings may not be meaningful because “scientific findings (on the effects of particulate matter on a person’s health) are based on 24-hour exposure to PM2.5”.

“There are very limited findings on one-hour exposure to PM2.5,” she added.

The aqicn.org website does not claim that its reading is directly equivalent to the 24-hr PSI. For example, the site recorded a maximum AQI reading of 182 over 24 hours from 6pm on Wednesday.

Over the same period, NEA recorded a maximum 24-hour PSI reading of 110.

On its website, the aqicn.org stated that it uses data from four sources: The citizen weather observer programme, NEA Singapore, the World Meteorological Organisation and Malaysia’s Department of Environment.

MORE ALTERNATIVE READINGS 

Besides aqicn.org, there are various other readings available online, including those by web or mobile application developers.

For example, Originally US, a Singapore mobile app development agency founded in 2014, posted its version of “actual PSI readings” on its Facebook page last Saturday.

Users may access the PSI readings created by Originally US via a web link — a service which the developer has been providing for several years. 

On the webpage, the company said: “We have created what we think is a right implementation of NEA's official PSI formula on haze.gov.sg to deduce the actual, hourly PSI measurements from hourly PM2.5 readings.”

NEA did not respond directly to TODAY’s questions on the accuracy of Originally US’ readings. However, it said that, in general, NEA’s measurements are more accurate than other sites given that the authority relies on many sensors located around the island, and the other sites in turn rely on data from NEA.

The agency has a total of 22 air quality monitoring stations — each with six sensors — spread out across the island.

Dr Khairunnisa said that these sensors adhere to the “US EPA standards”.

“We would advise the public to look at our PSI readings, because those readings are based on measurements from our air monitoring stations.”

Dr Khairunnisa also said that alternative sites calculating air quality could be using “other devices”, such as sensors which can be used via mobile phones, including “low-cost air-quality sensors that can be bought off the shelf”.

“Those are not the same as (NEA’s) sensors, which… are calibrated on a regular basis. We cannot verify the accuracy of readings taken by those other PM2.5 sensors, which are likely cheaper than ours,” she added.

Other examples of alternative websites include IQAir AirVisual — a Switzerland-based air quality monitoring facility — which ranked Singapore as having the seventh-worst air quality in the world as at 10.45am on Thursday.

VISIBILITY AND AIR QUALITY 

On another matter, NEA cautioned that poor visibility might not always mean a high PSI reading. This is so because humidity is also a factor in visibility.

“Poor visibility is not always caused by air pollutants. A very significant contributor to visibility is water. Water is not captured in PSI or PM2.5 measurements as it is not considered a pollutant.”

NEA said that water can cause visibility to worsen by binding with certain air pollutant particles. These water-bound particles swell to a size that makes them scatter more light than when the air is dry.