To answer the first question about our source of the data, the World Air Quality Index service only uses officially measured and published data. In other words, at the World Air Quality Index project we do not operate any Air Quality monitoring ourselves. Instead we rely on each country's EPA to operate and maintain their own monitoring network ^[1]

In the case of Warsaw, the data is measured by the Regional Inspectorate for Environmental Protection in Warsaw (Wojewódzki Inspektorat Ochrony Środowiska w Warszawie). They provide the data using both the Polish AQI (called Indeks Jakości Powietrza, aka IJP), as well as the raw concentrations for PM_2.5, PM₁₀, Ozone, etc. You can see the data from their map service.

The IJP (Indeks jakości powietrza) is defining its own color coding and scale, which is summurized in the table below. Note that the color standard used by the IJP is not completely equivalent to the US EPA AQI standard, but that's something we will look into details later.

The IJP is calculated using a formula which takes into account the concentration of several pollutants. Almost every single country is haivng into own formula, and in the case of Poland, the "max"-based formula is used:

That might look slightly barbarian, but actually, this formula is very similar to the one used by the US EPA (and therefore on the World Air Quality Index project), except for major 4 differences:

• The IJP scale goes from 0 to 10, while for the US EPA, it goes from 0 to 500. Note that other standard like the Hong Kong AQHI standard are also using scale from 0 to 10, this is completely fine.
  
  
• The breakpoints used in the IJP are linear for each pollutant, while for the US EPA they are not. The breakpoints used in the US EPA standard are explained in this article.
  
  
• The IJP is including Benzen in the list of pollutants, while it is not for the US EPA. This is a very positive point for the IJP since Benzen is known to be highly cancerous.
  
  
• The IJP is based on 1 hour readings, while the US EPA standard is theorically based on 24 hours averaging. Again, this is a very good point of the IJP, since we are promoting the universal use of 1 hour readings through our Instant Cast reporintg.

Because graphics are better than long sentences, the following interactive graphics is a compariosn of the AQI and IJP scales for PM_2.5 and PM₁₀. You can just move your cursor over the graph to see the compared AQI, IJP and raw concentration.

When comparing the IJP and AQI scales for PM_2.5, the confusion can come from both ends of the scale:

• When the Air Quality is Good, a Moderate reading using the US EPA standard is equivalent to a Good (Dobry) with the IJP. And a Moderate (Umiarkowany) using the IJP corresponds to a Unhealthy for Sensitive Groups using the US EPA standard.
  
  
• When the Air is polluted, the IJP standard is setting a very conservatively low threshold, at such level that a Unhealthy reading using the US EPA standard can be equivalent to a Harzardous (Bardzo zły) with the IJP.

The four graphs below summurize the differences between the two standards for the station Komunikacyjna in Warsaw for the last 5 days (PM_2.5 on the left, and PM₁₀ on the right).

One can clearly see the equivalence between the IJP green and AQI yellow for the PM_2.5 scales (on the left). If one abstract this color swapping difference, then it is obious that both scales report similar level of pollution.

As concerns PM₁₀, while for the US EPA AQI, the index is always lower than the PM_2.5 index, for the IJP, the PM₁₀ index is actually similar or higher then the PM_2.5 index.

Usually, one expects PM_2.5 to be more harmfull than PM₁₀ (as explained in this article), so we will have to come back on this difference in a later article (especially looking at the correlation factor between PM₁₀ and PM_2.5 based on empiral data for US, Europe and Poland).

Among all the other pollutants, Ozone (O₃) is the one most likely to be the second dominent pollutant. It is already known that the US EPA 1-hour Ozone standard is not so conservative compared to other existing standards (see this article), and the IJP confirms this rule. The two graphs below show the Ozone Index for both AQI and IJP.

The other pollutant worth having a deeper look at is the Benzen, since the IJP is one of the rare scale taking into account Benzen in the Index calculation. Fortunately, the corresponding index for the last 10 days shows that the benzen concentation is always within the [0-1] "very good" range, so not a "daily problem" like PM2.5 can sometime be.

On one side, there are some very good improvements in the Indeks Jakości Powietrza standard defined by the Polish EPA: Using 1-hour readings, including more pollutants like Benzen and lowering the Hazardous level threshold; All those should be promoted worldwide and we will be looking at extending the World Air Quality Index project AQI calculation to also take into account Benzen.

On the other side, trying to compare apple-to-apple the IJP and AQI is not straightforward. When plotting the composite AQI for both IJP and AQI (see graphs below), it looks more like comparing apple and bananas. This is, according to our experience, the most confusing factor when comparing figures provided by different websites.

Last, the key learning is to remember as that each scale has its own specificity, and the most important is to keep the diversity in the scales: We are starting to believe that having only one unique scale might not be the right solution, and we are now working on a global solution which will allow users the select the scale which best fits their needs.

US EPA Composite AQI (Air Quality Index)

Poland Composite IJP (Indeks jakości powietrza)

EU Composite CAQI (Common Air Quality Index)

Note: the composite calculation is based on PM_2.5, PM₁₀ and Ozone only.