HYPERLOCAL CO2 EMISSIONS IN LARGE CITIES @COP27

8 billion people live on planet earth and over 4 billion of us live in cities. It then makes sense that urban populations contribute disproportionately to carbon emissions owing to the burning of fossil fuels in industrial production, private vehicles, subpar waste management techniques, and resuspended dust. It follows suit that cities account for 70% of global emissions. Reducing emissions in cities is absolutely vital to get to net zero. This also makes cities the prime focus for carbon emission data collection which can then go on to inform climate policy that targets these emissions.

The top 50 cities in the world account for 25% of global emissions. So really, if we want to tackle global emissions, we have to tackle emissions in the largest cities in the world. Collectives such as the Carbon Neutral Cities Alliance are working towards reducing carbon emissions in cities by 80% to 100% by 2050 or sooner. Legislation like the Paris Agreement manifests as a call to action that urges member nation-states to cut emissions by mid-century.

Hyperlocal data, in contrast to local data, is much more locality-specific — and so, can be used toward identifying disproportionate carbon emissions down to particular street addresses. While hyperlocal data may be more difficult to collect, it is also more effective and efficient in the direct targeting of problematic locations, which results in a greater demand for this kind of data by businesses and climate change organizations.

Ambee is a climate and environmental intelligence company that seeks to democratize access to environmental intelligence for the whole world. This kind of data and information being open-access is a positive step toward more climate-conscious behavioural and policy changes. Employing technological advancements in domains such as AI and machine learning, Ambee collects data through sensors, PCB monitors, and satellite data, that will be actioned towards the generation of this hyperlocal data. Ambee’s Application Programming Interface or API also collects data on pollen levels, active forest fires, severe weather, soil, and on normalized-difference vegetation indices.

Mumbai, the economic capital of India with a population of 21 million is the first city in the world that now has hyperlocal CO2 and emissions data. The GHG API also collects data on water vapour, methane, and ozone levels in these hyperlocal geographical areas, in addition to data collected on CO2 levels. Ambee’s API, will allow businesses, government bodies, and administrators to access real-time data for GHG emissions in their immediate vicinity which can then be actioned towards reinforcing climate strategy, simplifying personalization, and enhancing productivity.

Ambee installed 65 sensors across the city - you can see the locations on the sensors on the map. We combined data from the sensors with satellite data and use AI and ML to determine real-time emissions in each hyperlocal area in the city. The results have been very interesting. The first big insight is that Colaba - on the southernmost tip of Mumbai - one of the wealthiest areas in Mumbai had the highest CO2 emissions levels in all of Mumbai during the month of September. In fact, the elevated levels of CO2 emissions are not only in Colaba. The average outdoor normal emission levels are 250-400 ppm; across all of Mumbai, CO2 varies from 410 - 1400. Clearly, CO2 levels are elevated in all parts of the city.

This data produced many interesting insights. One of which can be very powerful. We noticed HIGH CO2 concentrations in areas that DON’T have many cars. Mumbai is a low-car-density city compared to Delhi or Bangalore, with most people relying on electrified public transport. Despite this, we see early morning hotspots (7-8 am), which show the existence of very hyperlocalized industry processes (baking factories, etc) that emit CO2. This shows that vehicles affect a city’s CO2 emissions far less than other factors like industrial activity or construction. It is important to note that Mumbai is building a massive new trans-harbour bridge, and has a refinery. The collection of hyperlocal data allows for nuanced conclusions such as this to be drawn: the pinpointing of particular contributors rather than the drawing of conclusions solely based on macro-level data and existing research.

Now let me take you to Delhi the city I live in. Priya lives in Kalkaji. She loves to exercise outdoors. What is the best time for her to exercise when PM2.5 levels are the lowest? Most of you probably exercise outdoors in the morning or in the evenings. However, if you live in Kalkaji, as you can see the best time to exercise is between 11 am and 1 pm. This points to another strength of hyperlocal data collection: open access, temporally-specific data, can inform people’s behaviours in a way that mitigates their exposure to pollutants that can have severe consequences for public health. In countries like India, particulate matter is parallel to GHG emissions and has a brutal impact on our lives. It affects me personally because half of all children in my city grow up with underdeveloped lungs. This data can drive decisions against emissions and towards health. This is what motivated me to take on this project.