Air pollution in megacities

The NextGEOSS Pilot “Air Pollution in Megacities” offers a service-on-demand for analysing air pollution and urban growth in cities and megacities worldwide by exploiting full times series of multi-sensor Earth Observation data:


It supports environmental agencies, governmental bodies and the health community in developing measures for mitigation, adaptation and prevention. By providing a quick integrated view on the air pollution levels and trends and urban growth, in any area of interest, users can benefit from EO data, the Copernicus programme and the NextGEOSS infrastructure.


The pilot addresses especially the Societal Benefit Areas “Sustainable Urban Development” and “Public Health Surveillance” and promotes “Open EO Data”.


The Research Pilot addresses the Societal Benefit Areas “Sustainable Urban Development” and “Public Health Surveillance” and promotes “Open EO Data”. The products will be derived following the objectives defined within the Societal Benefit Areas and in close interaction with users. (From ESA)


With reference to the UN Sustainable Development Goals (SDGs) for 2030, the megacity pilot directly addresses two of them:


(3) Establish Good Health and Well-Being

(11) Sustainable Cities and Communities


Urban settlements are hotspots for air pollution being a hazard for human health. Following the latest estimates of the European Environmental Agency, 400.000 thousand premature deaths can be attributed to poor air quality in Europe only. As the majority of emissions leading to air pollution are of anthropogenic origin, cutting emission sources is a main concern of authorities and decision makers. On the other hand, the general public should be aware of the local pollution levels and trends to assess political decisions that may impact their everyday life. From the global perspective, it can be helpful to compare urban pollution at different hotspots. For example to learn about the links to the local environment, settlement structures and economic development.



The NextGEOSS Megacity pilot combines efficiently the latest available information from European space sensors on air pollution and urban settlement to derive an objective estimate of NO2-trends in the vicinity of megacity and urban pollution hotspots. In this way, it helps to achieve the SDGs for health and sustainable urban development. It is also worth noting that both aspects have a close connection as poor air quality is directly linked to human activities and economic status. The way we organize our daily life and economic activities is crucial for sustainability and a healthy way of living. Combining information on settlement and pollution levels is thus a first step to better development indicators.


To achieve project goals, the megacity pilot had to tackle several challenges:

-       Remote sensing data is in general bulky and demanding to process

-       Urban settlement area is not well defined by geometry only

-       Trends are mostly calculated not related to the urban settlement 

-       To be useful on-demand processing must be highly efficient

Environment 1 Climate & Health | Global Pollution Hotspots | Pollution Trends in Megacities

Atmosphere & Climate change | Built environment/atmosphere | Urban Areas | Pollution & Health | Urban Expansion & Air Quality | NO2 Pollution Trends in Megacities


Data sets


Data collections


Community portal



The solution

The solution

The solution to the challenge has both scientific and technical aspects. The three main aspects are:


  1.     Combining remote sensing based information on NO2 and urban settlement (GUF) allowed deriving an objective indicator for human development. This can be achieved by calculating the pollution burden per settlement area, not by geographical area. 
  2.     Harvesting geoserver for NO2 and GUF data avoided the local handling of large volume data
  3.     Parallel processing of shorter time records allowed on-demand processing of long-term data records
  4.     Deploying the backend via docker allowed hardware independent development and flexibility

5.     Using a pre-configured interface from German Remote Sensing Data Center eased the development effort



Service Layer Description

From the user perspective, the application requires four steps to be conducted to trigger a time series analysis:


  1. Browsing of global map showing urban settlements and megacities via the Global Urban Footprint (GUF),
  2. Interactively define area of interest (AOI) / zoom region for on-demand extraction of full EO time series and model data,
  3. Chose the desired earth observation data source,
  4. Start calculation of long-term NO2 trends related to the built-up area within your AOI.

In detail, the following software components are being used:


1 Urban Area Selector (UAS)


The urban areas of interest can be defined by means of the Global Urban Footprint (GUF) which has been derived from EO data by worldwide mapping of settlements. A total of 180 000 TerraSAR-X and TanDEM-X scenes have been processed to create the GUF. The settlement pattern allows for the analysis of urban structures and the arrangement of rural and urban areas. The GUF is available as a DLR Geoservice for scientific use (


The urban area selector enables in an interactive way to define the reference area or agglomeration for the upcoming air quality analysis, like built-up areas or non-built-up areas. The implementation shall be generic so that any area can be selected by vector (administrative boundaries) or raster data (land use classes).


 2 Time Series Processor (TSP)

To quantify the related air pollution levels for the selected urban area, satellite-based observations are examined using the NextGEOSS infrastructure. The discovery services are used to choose available EO data e.g. of GOME-2 (2007-present) and the upcoming Copernicus fleet (Sentinel-5P, Sentinel-3) but also available data from the Copernicus Atmosphere Monitoring Service (CAMS). Here the NextGEOSS DataHub plays the key role. The backend for the Time Series Processor (TSP) is deployed within the NextGEOSS Public Cloud.


As the major objective, the EO data and model data will be spatially integrated with respect to the reference area defined by the Urban Area Selector. Therefore, this will be an innovative approach to correlate air pollution levels to land use classes or to distinguish between built-up areas from non-built-up areas.


3 Time Series Viewer (TSV)

The Times Series Viewer is realized as a web client. To enhance the interpretation of the resulting time series, a time slider functionality is available.


4 Time Series Analyser (TSA)

A web client enables the user to interactively analyze the resulting time series. The main features comprise the computation of basic statistics and the detection of trends.


NextGEOSS Air Pollution Service


IT needs

The megacity pilot team holds great expertise in processing and analysing remote sensing and model data on air pollution and land coverage. 


NextGEOSS provided the megacity pilot with:


-       Cloud processing and storage capacity

-       IT support for system setup and performance improvements


NextGEOSS Air Pollution in Megacities Map


What are our users saying about the NextGEOSS user experience?

The NextGEOSS AirQuality in Megacity Pilot offers the potential to derive time series and trends of air quality-related information from a variety of sources: Tropospheric NO2 as derived from EO missions or from atmosphere models such as CAMS or POLYPHEMUS/DLR and moreover it offers the functionality to include other socio-economic data sources such as the COVID-19 cases for the year 2020. The NextGEOSS AirQuality in Megacity Pilot was developed and deployed recently and is about to be evaluated by the air quality modelling community.


The NextGEOSS platform was very easy to use

All required software components have been dockerized on DLR’s development system. The air quality pilot application consists of three docker containers: the Geoserver with remote WPS plugin, a container hosting “OpenFire” and one container hosting the scientific processor. All required data needed for time series processing are retrieved from those servers hosting the respective information via the standardized OGC protocols. Therefore, all data products which are accessible on OGC-protocol compliant servers may be used for this service.


DLR - German Remote Data Center

The pilot`s single contractor and coordinator German Remote Data Center (DLR) has a long-standing track record of successfully performing and coordinating applied science projects in the field of remote sensing as well as retrieval, modelling and data assimilation. It has well-proven experience with level 2, 3 data regarding trace gas and aerosol missions and sensors like GOME-1 and GOME-2, SCIAMACHY, MIPAS, AURA-MLS, AVHRR, MODIS and the Sentinels to name but a few.


This has allowed the contractor to develop level 2 and higher data and information products with high relevance for authorities and the general public. User-driven activities have been dedicated to developing and building up value processing chains for information services within scientific projects.


Benefits for the pilot

Economic: the NG approach allows linking air pollution to populated urban areas This is a prerequisite for assessing environmental stressors per capita and thus in calculating health care costs. Moreover, additional impacts on air quality such as the Corona pandemic (by taking into consideration the number of Covid-19 cases per country) is displayed as a timeseries for all data products covering the year 2020 (the year of the Corona-virus pandemic).


Environmental: through the pilot, air pollution burden can be assessed globally. Different hotspots can be compared by the user. Tropospheric NO2 is used as a well-established indicator for anthropogenic pressure on the environment.


Regulatory: as NO2 is mainly emitted by anthropogenic sources, measuring NO2 levels is a good method of assessing the effectiveness of political decisions e.g. on traffic reduction and control.


Scientific: the megacity pilot allows efficient access to and comparison of different satellite instruments, e.g. GOME2 from MetOP A, B and C. It is also possible to retrieve Copernicus model results from CAMS. These comparisons are much needed for assessing the capacities and limitations of individual measurements or model results.


Social: health is a decisive factor for human well-being, economic participation and social inclusion. However it is at risk by environmental stressors worsening personal distress and challenging society.



Read more about space & security