The Capital has its first breakdown of air pollution by source in nearly three years. But the Indian Institute of Tropical Meteorology’s (IITM) new measurements cannot yet say where a large share of the pollution they measure comes from, and its air pollution prediction model still runs on a 2021 emissions estimate that its own website says needs updating, with a replacement expected only by June next year.
Particles labelled “secondary sources” --- these form in the air due to a phenomenon that involves an interaction between gases — made up 29.3% to 42.2% of the new lab’s daily estimate of PM2.5 from October 20 to 24. The five days IITM’s page showed on Saturday evening put these sources ahead of vehicles at 19.4% to 27.6%, HT’s analysis of the numbers showed. Officials said they have no system yet to break these secondary aerosols down by their origin.
A second problem was that the lab’s total PM2.5 reading was itself far below what Delhi’s monitors recorded. For instance, it was 55.3 micrograms per cubic metre (µg/m³) on October 22. That number was 47% of a middle reading of 117.1 µg/m³ among 37 stations that day, a comparison with Central Pollution Control Board (CPCB) air quality data from that day showed. When this was pointed out by HT, officials said the site was still new and facing teething issues. “The PM 2.5 data will reflect accurately, shortly,” an official said.
Sachin Ghude, the IITM scientist overseeing the lab, said a large chunk of particulate matter originated as secondary pollutants: “Unlike primary, which are emitted directly into the air, secondary aerosols form when gases undergo chemical reactions in the atmosphere and get converted into particles.”
Also Read | Delhi AQI 'moderate' at 191; environment minister says govt in 'war mode for clean air'
Secondary organic aerosols form from volatile organic compounds (VOCs), and secondary inorganic ones, such as sulphates and nitrates, from gases including sulphur dioxide, nitrogen oxides and ammonia.
“We are working on that now, to try and find how these are forming. While we know this is a complex reaction of gases and VOCs, the original source is not known,” Ghude told HT. The lab’s next phase, he said, will first split the vehicle share between diesel, compressed natural gas (CNG) and petrol vehicles, which he expects “in a fortnight or so”.
The big gap in understanding the source could well hamper the effort’s usefulness.
“This facility will help us understand the composition of pollution as well as the contribution of different sources,” said Union minister of state for science and technology Jitendra Singh, while inaugurating the lab, called the Advanced Atmospheric Chemistry and Aerosol Toxicity Analysis Laboratory (Aatal) at New Rajendra Nagar, on Saturday. He added that this was important for “ensuring that action is based on scientific evidence”.
The lab’s 25 instruments, for atmospheric chemistry, source apportionment (working out how much pollution comes from each source) and aerosol toxicity, cost about ₹80 crore.
The ministry of earth sciences said the lab will complement IITM’s Air Quality Early Warning System and its Decision Support System (DSS), the prediction model whose source split the new web page shows by default; the DSS forecast for Saturday put Delhi’s transport at 17.7%.
Also Read | New GRAP actions failed to curb Delhi-NCR pollution in 10 of 12 episodes, reveals study
The model’s 2021 emissions inventory, its estimate of what each source releases, was “compiled using surveys done in 2016”, according to a 2024 paper by IITM scientists, and a replacement with a 2026 baseline is due only by June 2027.
IITM’s DSS page acknowledges “the latest emission fields are required” for more accurate estimates, and a second 2024 paper, summarised on its methodology page, says the model does not simulate nitrate or secondary organic aerosols, which alone made up 15.5% to 21.2% of the lab’s daily estimate.
A senior government official said the lab’s data would add to the DSS, but the overall assessment would not become fully accurate until the inventory is updated, which officials expect to happen next year.
What the estimate shows
Vehicles and the two secondary categories were the lab’s three largest of 15 every day. Its seven burning categories together made up 20.5% to 35.6% in HT’s sum, crop-residue burning alone 2.5% to 4.7%, and road, soil and construction dust 3.8% to 20.6%. On October 22, after IITM’s revision, burning was the largest group in HT’s sums, at 35.0% against 33.8% for the two secondary categories, the only day it was.
IITM’s page names vehicles the “dominant contribution” each day by skipping both secondary categories. On October 23, secondary organic aerosols alone edged vehicles, 19.5% to 19.4%.
The new page calls the split “real-time” but gives one figure a day per category, with no hourly data or method. IITM’s older Analysis page shows the lab’s daily figures in a chart marked “Experimental Mode - Preliminary results” and labelled a source contribution of “NR-PM2.5”, a term the site does not explain, where the new page says “Total PM2.5 (sum of the listed sources)”.
At 6.18pm on Saturday the new page also showed Friday and a part-day Saturday reading, with vehicles at 25%; by 6.30pm both had gone. They were back by 9.28pm, revised along with Thursday, with Saturday’s total raised from 25.2 µg/m³ to 29.9 µg/m³ and its vehicle share from 25% to 27.6%; the older chart, updated at 8.49pm, now carries the revised figures, and the tab now shows October 20 to 24. The changes suggested tweaks were still being made.
Also Read | Delhi minister orders crackdown on pollution hotspots as air remains ‘poor’ for 3rd day
IITM’s note on the lab says the facility will quantify contributions from sources “including transport, industry, biomass burning, and regional pollution transport”. None of its 15 categories covers pollution from outside Delhi, and IITM has not said how it tells dung cake, wood, crop residue and other fuels apart.
Sunil Dahiya, founder and lead analyst at the think-tank Envirocatalysts, said the true value of real-time source apportionment, forecasting and meteorological tools “depends entirely on data accuracy, lower model uncertainties, and actual real time availability”.
Delhi’s previous source studies, by IIT Kanpur in 2016 and The Energy and Resources Institute in 2018, are dated. The Delhi government stopped a Washington University study in 2021 over “unsatisfactory” results and halted IIT Kanpur’s real-time study in November 2023, again dissatisfied with its methods and results.