Hazy winter

AN unusual confluence of factors, the most important among them being meteorological, is what caused the unprecedented polluted atmosphere and smog in and around Delhi that hung low in the air for over a week following Deepavali (October 30).

Come winter, parts of north India, particularly the capital region of Delhi, are enveloped in smog during the night and early morning hours, which clears up in the morning as the sun moves higher up. This is owing to the formation of what is known as the “inversion layer”, a meteorological phenomenon that occurs during winter months, when the normal decrease in temperature with height is reversed across an atmospheric layer that forms close to the surface of the earth, at heights of 100-200 metres, and whose thickness can be a few hundred metres. “During November-December, this can be much lower, under 50 m at night and hovering around 200-500 m during the day,” said Sarath Guttikunda, founder of UrbanEmissions.Info.

Under normal conditions the air near the surface of the earth is warmer than the air above it because the atmosphere is heated from below as insolation warms the earth’s surface, which in turn warms the layer of the atmosphere directly above it. When the air is colder near the earth’s surface, like it is during winter, a warmer, less-dense air mass moves over the cooler, denser air mass forming an inversion layer. (Inversion layers form in summers too but they are formed at much greater heights of two to three kilometres.)

When temperature inversion happens, atmospheric mixing due to the convection that is normally present does not take place, resulting in the atmosphere becoming stable over the area experiencing inversion. The warm layer acts much like a lid trapping the cold air below it, which results in the collection of dust and pollutants at low heights as they can no longer be lifted from the surface and dispersed. Moreover, Delhi is a landlocked city where there is no natural horizontal sea breeze to sweep away pollutants. But in the mornings, local heating takes place because of sunshine, the inversion layer breaks, air rises because of convection and pollutants slowly get dispersed as the day progresses. If the wind circulation in the region is such that there are horizontal surface winds, they also will contribute to the clearing of the air.

This year, in the days following Deepavali, the dispersal of pollutants that were hanging below the inversion layer, and the decline in their atmospheric concentrations, did not happen as normally expected. According to M. Rajeevan, Secretary, Ministry of Earth Sciences, and an atmospheric scientist, coupled with this normal accumulation of pollutants close to the surface due to the thermal inversion layer, the air circulation also became unfavourable. “It was basically an anti-cyclonic circulation, which is a high-pressure system at a height of about only 1 km, which had a huge effect on the entire low level atmosphere,” he said (Fig. 1). An anti-cyclonic circulation has winds moving in the clockwise direction (in the northern hemisphere), much like the motion of a screwdriver, creating pressure on the atmospheric system below.

“Though such anti-cyclonic formations are part of winter wind circulations, and not even rare, this time it was rather unusual because normally we see them only at a height of 4 km and more. And it was also very deep. So, this air circulation coming down created a kind of subsidence [of the atmosphere below with accumulated pollutants]. Thus, the whole atmosphere became very stable and prevented any local convection. Otherwise, with local heating some convection takes place and the air goes up. But because of the anti-cyclone sitting so low, surface winds were very weak and could not advect any local pollutants outside,” Rajeevan said.

“So, there was Deepavali, when the amount of pollutants injected on a single day was very high (Table 1) and there was a spike due to pollutants arising from the large amounts of biomass burning in the neighbouring areas, which combined with our usual high-polluting local sources such as vehicular traffic, construction work and emissions from power plants. The two extreme polluting events met with very unfavourable weather conditions,” Rajeevan added.

There has been a lot of discussion on the contribution of pollutants (smoke, which is likely to be mixed with soil particles, dust and partially burnt plant material) from the burning of agricultural residue (stubble) in Punjab and Haryana, which are situated north-west of Delhi. It is true that pollutants from the stubble burning during September-November across most districts in Punjab and parts of Haryana have been a regular feature over the years, which does spread a stream of smoke over Delhi and beyond (Pic. 1). These fires are small, short-lived and burn at relatively low temperatures. So, this smoke tends to stay near the surface. According to Gufran Beig, Project Director of SAFAR (System of Air Quality and Weather Forecasting And Research) at the Indian Institute of Tropical Meteorology (IITM), Pune, the long-term average of the contribution of agricultural biomass burning to particulate matter (PM) pollution over the Delhi region is about 20 per cent and the rest is from local sources.

A study published in Current Science last year by the Hyderabad-based National Remote Sensing Centre (NRSC) of the Indian Space Research Organisation, which analysed the trend of stubble burning between 2004 and 2014 based on satellite images at 1 km and 750 m resolution, found that over the period, while the number of fires had been decreasing in October, it had been increasing in November (Fig. 2). The study also said that fire occurrence was detected almost every year in about 60 per cent of agricultural land in Punjab. In fact, according to a report of the U.S. National Aeronautics and Space Administration (NASA), based on images from its satellite Aqua, the number of biomass fires from Punjab during the October 25-November 5 period this year was 75 per cent more than last year.

SAFAR monitors the air quality for the National Capital Territory region of Delhi from 10 monitoring stations spread in different micro-environments, and on the basis of the measured values of eight major pollutants, it gives the quality of air in terms of a quantity called air quality index (AQI) on a daily basis. AQI is a dimensionless number varying between 1 and 500. It is calculated on the basis of five main pollutants and when a value of AQI for the entire Delhi region is given, it is the 24-hour average of the most prominent, or lead, pollutant from all the 10 stations. Given that 2.5-micrometre-sized particulate matter (PM2.5) poses the greatest health hazard and also happens to be the lead pollutant in the Delhi region during the current winter period, the AQI is calculated on the basis of concentration levels of PM2.5.

The AQI scale is 0-100: Good; 101-200: Moderate; 201-300: Poor; 301-400: Very poor; and 401-500: Severe. Anything beyond 500 (usually denoted as 500+) is off the scale and is considered extremely harmful to human health. SAFAR also issues a three-day forecast in terms of the five AQI categories on the basis of a model in which the initial inventory of different pollution sources and measured values of meteorological parameters such as wind speed, wind direction, humidity and temperature are the inputs, which generates a value for AQI as the output.

According to Beig, on October 29, the day before Deepavali, particulate pollution (both PM2.5 and PM10) was typical of winter months, which was gradually increasing owing to the fall in temperature leading to the formation of low-altitude inversion layer. “[The concentration of the lead pollutant] PM2.5 was already more than 250 microgram/cubic metre. We had forecast a ‘Severe AQI’ for October 31 and that the situation would significantly improve on November 1. The PM2.5 level did indeed fall down to nearly half the value, but AQI was still ‘Severe’. Widespread biomass burning had already begun in several areas of Punjab and Haryana by October 15 or so, but it did not have much impact as winds were mainly north-easterly and winds continued to be mainly north-easterly till November 2. Its contribution to Delhi pollution then was only 1-2 per cent,” Beig said (Fig. 3).

The easterlies/north-easterlies would help in blowing away much of the pollution from stubble burning from Delhi. On October 30, the wind speeds of the north-easterlies (which blew for nearly one-fourth of the time) were also significant, being in the 1-4 m/s range. Although there were winds from the north and north-north-west, they were there for only fractions of the time.

According to a report on the Delhi pollution episode prepared by SAFAR/IITM, on Deepavali and the day after, winds were very calm over Delhi (less than <1 km/h), the inversion layer had come down to as low as 40-50 m, there was enough moisture in the air, and venting of pollutants from the surface was almost zero. These conditions resulted in the emissions from firecrackers getting stagnated at low heights, causing the PM2.5 levels to peak on October 31 (Fig. 4).

On November 2, however, the wind circulation significantly changed. As pollution from firecrackers was dispersing slowly, the wind direction became north-westerly and there was no easterly/north-easterly component at all. And as stubble burning was going on in the north-west and north-north-west directions, particulate matter from these fires started getting injected into the atmosphere over Delhi because of the altered wind direction. Also, the wind speeds over the region did not increase and remained as low as 0.3-0.4 km/h, a direct consequence of the anti-cyclone over the region. These conditions were ideal for pollution to remain trapped over Delhi, and PM2.5 levels once again increased. They dropped briefly on November 4 when wind speeds temporarily picked up and dispersed part of the pollution away from Delhi.

Air quality started to worsen in the following days. During November 5-6, all wind components became north-north-west and north-west. This was a directed channel for the smoke particles from stubble-burning areas. In fact, this resulted in the stubble smoke spreading all across the Indo-Gangetic plain (IGP). Also, thanks to the north-westerlies, there was an influx of stubble-burning pollutants from Pakistan, said Hiren Jethva, a research scientist at the NASA Goddard Space Flight Centre (Pic. 2).

“There was direct flux of particulate matter from stubble burning, and the area in which biomass was being burnt too had increased significantly during November 3-5. On November 6 the inversion layer too remained below 100 m throughout the day,” Beig said. “This is unprecedented and the highest since SAFAR started in 2010,” he added.

According to him, the share of biomass burning to PM pollution had shot up to 70 per cent during November 6-7, although Guttikunda feels that would be an extreme value.

“I feel that would be an extreme case, maybe for some hours. But saying that all of 22 million people, eight million vehicles, more than 10,000 industries, more than 1,000 brick kilns, and all the cooking and heating and waste burning contributed to only 30 per cent of the pollution in a day seems very unreasonable. There were times when the [modelled, not measured] contribution of fires was close to 50 per cent, but that was only for certain hours. The average during the peaks was around 20-30 per cent,” Guttikunda said.

On November 8, the winds became westerly and west-west northerly and picked up speed as well to 8-10 km/h and the share of stubble burning had come down to about 10-15 per cent. “Though at present the wind circulation pattern remains the same, the temperature has dropped by about 2 °Celsius and consequently the inversion layer has also come down to about 150 m,” Beig said. “Of course, pollution levels will continue to fall to around 200 microgram/cubic metre, but in terms of AQI still ‘Very Poor’,” he added.

Comparing the situations in 2015 and 2016 using satellite images, Jethva, who is analysing the pollution effects of stubble burning over the years, made an interesting observation. “In 2015, the wind circulation pattern was such that the pollutants from Punjab and Haryana were steered around towards Rajasthan [Pics 2 & 3]. Though I have not looked at what pollution effects it had over Rajasthan, this aspect is worth studying,” he said. One of the reasons for this happening could be the effect of the easterlies that would have set in under normal circumstances following the monsoon period.

One of the main reasons for the absence or very low speeds of surface wind over the Delhi region was, as pointed out by Rajeevan, the anti-cyclone hovering over the region. But what was the reason for the absence of, or at best weak, easterlies? Even the north-easterlies that were there during the pre-Deepavali period were not very strong. By now, following the monsoon, the easterlies should have been established over the IGP.

Monsoon withdrawal this year was quite late; it withdrew from the north by about October 20 and the total withdrawal occurred only on October 28. Was the late withdrawal from the north, a delay of nearly 15 days, the reason for the easterlies not getting formed? “It is very possible that the late withdrawal could have had an effect,” Rajeevan said. “Even now the easterlies are very weak. Normally, by November the easterlies set in. Gujarat and Maharashtra have received rains even in October. That also could have had a role in the circulation,” he added. But why was the withdrawal so late? Climate change or some other reason? That is a topic for a different article.

“The temperatures are higher than usual for a November. By now, the temperature would have been under 15 °C at night, kicking in the need for heating, which has not happened yet. While the open agri-fires will be gone now, as soon as the heating period starts we have another episode in the making,” said Guttikunda, warning of a possible intense pollution episode in the days to come.