Back to news
3 May 2021 | Atmosphere

Investigating air quality in the Kathmandu Valley: The need for data

6 mins Read

70% Complete
What could be behind the deterioration in air quality in the Kathmandu Valley and elsewhere in the HKH?

The HKH is an atmospherically turbulent region with areas of high population density and which is experiencing rapid industrialization, further population growth, urbanization, and major shifts in land use and landscapes. Major regional sources of air pollutants include biomass burning, transport, power generation, and industry (e.g., brick kilns). Biomass burning includes cooking, heating, rubbish burning, agricultural practices, land clearance, and forest fires. In Nepal, the primary sources of emissions are the energy, residential, industry and transport sectors. Open burning, including forest fires, crop residue burning and refuse burning are also critical. Pollutant concentrations are affected by a range of factors including emissions, meteorology and transformation of emitted pollutants.

variable source of pollutants
Figure 1: Alongside the annual sectoral breakdown of emissions (left), and a similar plot for the March-April window (right), the figure suggests that forest fires are an important and variable source of pollutants, but contribute less than 10% of overall emissions.


Biomass burning is a primary driver in reduced air quality during the pre- and post-monsoon seasons, and is linked to farming practices across the HKH. Biomass burning and regional haze episodes can be extensive and are highly variable across locations and seasons. In the Nepali context, the main forest fire season is from March to April; and, open agricultural residue burning happens during October-November (post rice harvest), and March-April (post wheat harvest). Open waste burning, in contrast, is ubiquitous, happens at a range of scales, and occurs year-round (see Figure 1).


What can affect air pollution measurement and analysis?

Several factors can potentially affect air quality measurements and analysis. Our understanding of air pollution depends on data from atmosphere monitoring stations, satellite data on forest fires and levels of air pollution, data on air pollution transport, and meteorological factors. Our Air Quality Watch app provides a visualization for many of these data sources. It is clear that we need a good network of monitoring stations and other supporting data sets to monitor and analyse air quality and high pollution episodes.


What is the status of atmospheric monitoring networks in the region?

Existing atmospheric monitoring networks in the region are relatively sparse (where they exist at all), report hourly data with little or no personal exposure assessment information.

Atmospheric levels of PM2.5
Figure 2: Atmospheric levels of PM2.5 (ugm-3) between 17 March and 3 April in central Kathmandu. The grey points are hourly data, black points/lines show derived daily concentrations (24-hr mean), and the red line is the national ambient air quality standard (NAAQS) of Nepal for PM2.5


This affects our understanding since fine scale (in time and in space) distributions and levels of major pollutants are not well characterized. In the current context, data from a central Kathmandu air quality monitoring site (see Figure 2) indicates that the recent haze episode started on 27 March, yet daily averages for PM2.5 before this period already exceed the national statutory threshold limit of 40 µgm-3    outlined in the Nepal National air quality standards. PM2.5 was reduced during this period relative to the following episode but is still higher than the Nepal ambient air quality standard of 40 µgm-3 as was announced in a press release by the Department of Environment, Government of Nepal. This pattern was repeated at other available local monitoring sites (ICIMOD at Lalitpur) and for other pollutants such as O3.


The value of satellite data

Satellite data helps investigate columnar level of air pollution, but to extract maximum value this information should be assessed in conjunction with other data sets.

One of the products that the MODIS satellite platform provides is a measure of atmospheric particulate loading, known as aerosol optical depth (AOD), which indicates levels of particulate pollution. Figure 2 shows the MODIS AOD product for pre- and post- episode, with increased aerosol depth from 26 to 28 March indicating higher levels of pollution over the region.

Modis AOD
Figure 3: Modis AOD over Nepal and the wider region.The green dot (28oN, 85.5oE) highlights Kathmandu. Blank pixels represent data gaps.


When that AOD is overlaid on the satellite true colour image, a deep haze is visible over the Kathmandu Valley during that episode. Importantly, this is not reflected in the above pictured raw AOD product, because of constraints in retrieval methodology and potential mischaracterizing of “haze” as “cloud” during some of this period.


Modis AOD and true colour image for Nepal
Figure 4: Modis AOD and true colour image for Nepal.


Fire counts are critical, but are only part of the story

Fire counts and extents are critical dimensions in the analysis of the impact of fires on air pollution

Fire radiative power indicates the intensity of fires, and localized land use cover indicates what has been burnt. Actual pollutant emissions can be calculated based on experimentally generated emission factors. Since satellite data is limited to overpass intervals, instrument capability, and data retrieval algorithms, it is therefore possible to miss smaller and or more transient fires.


How is the movement of pollution a factor in air quality?

The movement of pollution from source to receptor site is a key factor in understanding sources and levels of pollutants

A preliminary review of modeled air mass trajectories (see Figure 3) suggests that a stagnant air mass could have been a potential contributory factor to the particularly high PM2.5 levels experienced from 19 March across the Kathmandu Valley. From this it can be seen that the air mass over Kathmandu was relatively still at the start of the episode, which may have led to a continuous buildup of pollutants in the Valley.

Calculated trajectories for Kathmandu
Figure 5: Calculated trajectories for Kathmandu (black dot) showing modeled air mass histories by frequency. The red dots show fire counts.


Data on the Planetary Boundary Layer (PBL) –  the bottom part or lowest part of the atmosphere which comes into contact with the surface of the planet and is directly influenced by this contact – provides important insights into mixing as well as indicating potential stagnant atmospheric conditions.

Mixing can be either vertical mixing – which is air mixing upwards or downwards, or horizontal mixing – which is air mixing sideways. Although winds have a vertical and horizontal component, we humans experience them mostly as horizontal components. A gliding eagle, for instance, may experience more equal horizontal and vertical components.

In general the lower the PBL, the less vertical mixing is possible (so pollutants can build up and aren’t diluted by the wider atmosphere).


Could meteorological conditions be a factor?

Data suggests that meteorological factors led to a very stable air mass with low winds and  a consequent buildup of air pollution during the studied period.

Meteorological analyses of winds at high altitude, and comparison with those nearer the surface from images based on suggest low pressure systems may have contributed to atmospheric stagnation and buildup of air pollution during the studied timeline. For atmospheric scientists, this scenario is very interesting and if confirmed (this initial analysis is based on preliminary analysis of images from a single model product) may have been reported elsewhere, with a potentially similar event described in China in 2017.

Stay current

Stay up to date on what’s happening around the HKH with our most recent publications and find out how you can help by subscribing to our mailing list.

Sign Up
A more efficient brick sector through collaboration and mitigation actions

At the second Policy and Advocacy Network Asia (PAN) meeting on 20 and 21 June 2019, ICIMOD provided government representatives ...

25 Sep 2017 Atmosphere Initiative
Signing of an agreement between ICIMOD and SIDA

As part of its mission to reduce poverty in the world through development cooperation, the Swedish International Development Cooperation Agency ...

Air pollution research into use

ICIMOD responds to key research questions at the local to regional level, generating highly technical air pollution-related data in a ...

8 Jun 2022 Air and air quality
Air quality monitoring station and dashboard established

Air quality monitoring data is crucial in identifying problem areas to target programmes geared towards mitigation. On 6 May 2022, ...

14 Dec 2015 Atmosphere Initiative
Motivating Journalists on Atmospheric Issues

  Twenty eight highly motivated journalists from eight countries of the Hindu Kush Himalayan (HKH) region —Afghanistan, Bangladesh, Bhutan, China, India, ...

Draft FABKA roadmap highlights future as an independent legal entity

Also in October 2019, members conducted the 4th FABKA meeting in Pokhara where, following rigorous discussions, a roadmap was developed ...

29 Feb 2016 Atmosphere Initiative
Winter Study on Outdoor/Indoor Air Quality Measurements in Chitwan

A study of particulate matter (PM10) in ambient conditions was initiated in January 2016 by the Atmosphere Initiative of the ...

15 Feb 2017 Atmosphere Initiative
Emissions Study in Lumbini Analyzes the Effects the Agricultural Crop Residue Burning

While general awareness of worsening air quality in the Hindu Kush Himalaya (HKH) has risen in recent years, this attention ...