Enhancing Resilience to Poor Air Quality in India’s Economic and Political Hub
Region Asia
Summary
Industrialisation in the Delhi National Capital Region (Delhi-NCR) has led to a massive increase in air pollution and urban heat island effects. These two threats compound each year and are deeply detrimental to the economy, public health, and environment in this important region of India. Open wastelands–characterised as barren or exposed soil–could be exacerbating air pollution in the Delhi-NCR.
A group of researchers at the Indian Institute of Tropical Meteorology have developed the Air Quality Warning and Integrated Decision Support System for Emissions (AIRWISE), to predict city-scale air pollution events and provide warnings and mitigation scenarios to residents and planners. The tool is now being enhanced with public Earth Observation data, to better understand and model the widespread impacts of open wastelands. This project has two objectives to support regional urban planners and policy-makers:
- To identify and map open wastelands in the Delhi-NCR and estimate their contribution to particulate matter pollution using Earth Observation and machine learning;
- Use scenario-based simulations to assess the potential benefits of covering open wastelands in vegetation, via surface temperature reduction and improved air quality.
These efforts to improve the existing AIRWISE toolkit using Earth Observation are supported by the European Space Agency and Future Earth’s Joint Program.
Value Chain Diagram
Decide
- Supports operational use by urban planners, public health agencies, regulators, and policy makers in Delhi-NCR to issue warnings and protect the population ahead of poor air quality events
- To design targeted mitigation strategies for heat stress and air pollution, especially from wasteland-derived dust sources
- Enables evidence-based decisions on converting wastelands into green spaces
- Strengthens early warning and response actions for extreme pollution and heat events using AIRWISE outputs
- Provides a repeatable framework for adoption to other urban areas, as well as incorporating new policy decisions back into the boundary conditions
Apply
- AIRWISE Air Quality Early Warning System provides improved air quality predictions and advance warnings of extreme air pollution events (as maps, web dashboards, tailored indicators, and reports)
- The existing AIRWISE System provides information on the contribution of different emission source sectors and NCR districts to pollution in Delhi.
- This project will provide an enhanced Decision Support System (DSS) that builds on AIRWISE capabilities (sectoral and district-level pollution attribution) to support barren/wasteland management through targeted greening interventions for mitigating heat stress and improving air quality
- Simulates nature-based adaptation or stress-event scenarios for greening identified open wastelands, producing a variety of decision-relevant air quality indicators (as map-based outcomes and as guidelines for policymakers)
Data production
- AIRWISE provides air quality forecasts up to 72 hours in advance, and its DSS enables sector-wise and district-level pollution attribution for improved air quality management and policy decisions
- AIRWISE incorporates satellite-based aerosol optical depth, land surface temperature, and land cover, with in-situ air quality and meteorological data
- Processed into:
- EO-informed boundary conditions and inputs for air quality models
- High-resolution wasteland maps for all 19 NCR districts using Sentinel-2 and machine learning classifiers (Random Forest, k-means)
- Dust emission estimates from identified wastelands integrated into AIRWISE modeling framework
Sense
- Aerosol optical depth observations (NASA MODIS and VIIRS, ISRO Oceansat-3 EOS-06)
- Burnt area products (Copernicus Sentinel-2)
- Soil moisture (Copernicus Sentinel-1 Synthetic Aperture Radar)
- Land use/land cover change (Copernicus Sentinel-2)
- Land surface temperature (Copernicus Sentinel-3 Sea and Land Surface Temperature Radiometer)
- Atmospheric reanalysis data (ECMWF ERA-5)
Decision-Making Context and Needs
The Delhi-NCR is home to 46 million residents, and serves as the economic and political centre of India. The region has experienced rapid expansion and urbanisation, and subsequently, an increase in urban heatwaves and air pollution. While they are at their worst in winter, pollution levels in the Delhi-NCR cause excess deaths annually due to exposure to elevated harmful particulate matter, on average equivalent to smoking 9 cigarettes per day. Heatwaves further contribute to air pollution by increasing surface ozone. And these are not just medical emergencies; schools and businesses have to close, and productivity can drop by 19–40%, massively impeding economic progress every year.
The effects of both extreme heat and poor air quality are felt disproportionately by poorer and weaker residents as well, as they tend to spend more time outdoors with less access to air conditioning and air purifiers. The increased use of these appliances, while offering temporary relief, can further exacerbate the negative impacts by putting pressure on the electrical grid and further increasing air pollution from power plants. If left unaddressed, these cascading effects will only worsen with climate change.
Open wasteland sites across the Delhi-NCR (defined as barren or exposed soil) act as a significant dust source in this region, contributing to particulate matter pollution and local heat stress. However, it is still unclear where these sites are, how they are changing over time, and how much their dust may be exacerbating air pollution and urban heat island effects.
Earth Observation Data Products and Information
The AIRWISE system (Figure 3) incorporates two types of workflows: the Air Quality Early Warning System and the Decision Support System.
- The Air Quality Early Warning System generates air quality forecast maps at different scales: 10 days in advance at 10 km x 10 km, 3 days in advance at 2 km x 2 km, and 3 days in advance at 400 m x 400 m.
- The Decision Support System is being further enhanced to develop an integrated decision support system at NCR district-scale, simulating future air quality to both perform chemical fingerprinting of different pollution sources, and test emission reduction scenarios (via greening of open wastelands).
The results of both workflows are disseminated on a web platform as geospatial data products and India Meteorological Department bulletins. To map NCR wastelands, machine learning algorithms like Random Forest and weka KMeans model classifiers have been applied, followed by an accuracy assessment. Google Earth Engine is being used to enable rapid and repeatable data acquisition and an operational early warning and decision support framework.
Underlying Observations and Data Sources
A number of inputs are fed into the toolkit:
- Aerosol optical depth observations from MODIS (NASA Terra and Aqua satellites), VIIRS (NASA Suomi National Polar-orbiting Partnership), and EOS-06 (ISRO Oceansat-3 satellite).
- Burnt area products and land use/land cover change (specifically open wasteland) from Copernicus Sentinel-2 multispectral imagery.
- Soil moisture using Sentinel-1 Synthetic Aperture Radar.
- Land surface temperature from the Copernicus Sentinel-3 Sea and Land Surface Temperature Radiometer.
- Meteorological forcings from the ECMWF ERA-5 atmospheric reanalysis data.
- Stationary geographical data and surface measurements from 320 air quality monitoring systems.
To improve the accuracy of the study, the team first updated the land use information for the Delhi-NCR region using recent Sentinel-1 and Sentinel-2 satellite data from 2021, instead of the older 1992 dataset. The finer detail of the newer satellite imagery gives a better picture of current land conditions. Open barren land, an important source of dust emissions, was then identified (Figure 4) using supervised machine learning algorithms in Delhi, and unsupervised in other districts, to understand the distribution and impact of these dust sources.
Figure 5 shows the level of dust concentration in Delhi-NCR from the open barren/wasteland areas identified. The pattern clearly highlights that dust emissions from open barren land regions can significantly increase dust concentrations in the atmosphere both locally and regionally through its regional transport.
Impacts and Benefits
Key stakeholders benefitting from the AIRWISE findings include the Commission for Air Quality Management (CAQM), the Central Pollution Control Board (CPCB), the Delhi Pollution Control Committee (DPCC), and the National Disaster Management Authority (NDMA). These bodies are able to use the forecasts and bulletins across the early warning system to respond and protect residents accordingly. The decision support system then allows local and regional authorities to test and implement adaptation solutions, such as increasing vegetation cover or installing natural or artificial air-filtering infrastructure. By enabling timely and targeted interventions—such as emission control measures and public-health restrictions—this system has demonstrated a clear operational benefit in reducing extreme pollution events. Notably, it helped cut peak pollution levels in Delhi by about 18–22% during a severe episode, highlighting its effectiveness in mitigating the intensity of high pollution periods. This ability to proactively “cut the peaks” significantly reduces public exposure during the most hazardous conditions and strengthens overall air quality management. The Commissions for Air Quality Management (CAQM) in the National Capital Region and the adjoining areas, which is responsible for managing air quality in the city of Delhi, has recognized the AIRWISE system as a key partner in a new investigative initiatives to reduce particulate matter pollution and improving air quality in extreme circumstances.
The ESA and Future Earth funded stage of the project aims to inform urban planners, policy-makers, and the general public, to what extent open wastelands are contributing to poor air quality, and provide a series of adaptation scenarios for greening these wastelands. Through the Earth Observation-driven analysis, the project will enable planners and decision makers to design more informed, targeted interventions to effectively reduce air pollution in the Delhi-NCR.
The satellite-driven system can simulate over a large enough region to accurately forecast air quality by accounting for regional weather-related forcings. But the coupling of Earth Observations with ground-based sampling stations and a variety of data resolutions means these forecasts can be delivered at street-scale, where it really matters.