We compared levels of these common pollutants after a lockdown was imposed to pre-lockdown levels. For most of these pollutants, there were some signs of a reduction of levels. However, in order to determine if these reductions were caused by the lockdown and not as a result of other factors like seasonal variation, we analyzed temporal trends across many years. In this way, we were able to analyse the impact of the lockdown measures imposed due to the COVID-19 pandemic on these pollutant levels. In essence, we used information about lockdown dates to analyse the reductions in economic activity in industries through remotely sensed pollutant data. This allowed us to gain insights about the reduced functioning of the economy during COVID-19 restrictions, as well as the impacts such a lockdown had on variations in levels of pollutants, which could provide interesting insights for reduction of contaminants in the atmosphere over longer durations of time.
We chose this challenge because Nepal, our home country, has been in lockdown for over two months now. We wanted to see if the effects of similar lockdowns could be seen in industrial areas through levels of the pollutants. We used available datasets from NASA's Aura satellite for Ozone and Nitrogen Dioxide levels, as well as Aqua satellite's MODIS, derived Aerosol data. We used Python to access data from the databases, used R for data processing and visualization and manipulated JSON files in order to create custom shapefiles for clipping. We faced problems in identifying "bad data" from different datasets; we realised that some places had the same readings consistently over long periods of time, which could only be attributed to missing or bad data. It took us some time to rectify this and amend our final products. We also faced compatibility issues on R, because the version of R we used didn't support gdal which we needed to read .hdf files. As a workaround, we used .tiff files available on an online repository instead. Based on our analysis, we were able to find significant reductions in NO2 levels in industrial areas that went into lockdown, such as in Hubei, as compared to temporal trends in past years. However, we weren't able to find such a trend for Aerosol and Ozone; for Ozone seasonal variations are similar and can be explained despite the lockdown, whereas for Aerosol, we simply didn't have enough good data available in Western Europe for January, February and March. We were able to analyse seasonal variations in these levels, however, and noticed that levels of pollutants peaked around May/June for Hubei and fell down during the autumn. We were also able to witness a declining trend in values for pollutants out of China after 2015, indicating some level of pollution control, although follow-up observations and more long-term data are needed to ground this claim.
https://docs.google.com/presentation/d/1aXRH-nFS_IfNvtUfLtJGONLdVwKdhNqYEI1ItvGjQjM/edit?usp=sharing
https://neo.sci.gsfc.nasa.gov/view.php?datasetId=AURA_NO2_M, https://neo.sci.gsfc.nasa.gov/archive/geotiff.float/AURA_OZONE_M/,https://neo.sci.gsfc.nasa.gov/archive/geotiff.float/AURA_NO2_M/,https://neo.sci.gsfc.nasa.gov/archive/geotiff.float/MOP_CO_M/,https://neo.sci.gsfc.nasa.gov/view.php?datasetId=MODAL2_M_AER_OD, https://neo.sci.gsfc.nasa.gov/archive/geotiff/MYDAL2_M_AER_OD/, https://earthdata.nasa.gov/, https://worldview.earthdata.nasa.gov/