Search this site
Embedded Files
Space Apps 2020 Challenge
  • Home
  • TEAM
Space Apps 2020 Challenge

A One Health Approach


Relationship between the air pollution on the planet, and the incidence of Covid-19

The Space IPN team is part of the challenge "A One Health Approach", which the objective is to analyze the levels of air pollution in some of the most important cities in North America and the relationship that these data have with the incidence of Covid-19, all this through Machine Learning techniques.












Air pollution poses an important environmental risk for health and it is one of the biggest problems that humanity has to deal with. Nitrogen dioxide (NO2) is the compound to analyze, it belongs to the category of gaseous pollutants, which are produced due to combustion engines and other processes like burning fossil fuels in power stations (1). A research issued by the University of Colorado explains that there is a relationship between nitrogen dioxide and respiratory diseases (2). There are proofs that concentration of nitrogen dioxide exists near main traffic routes, therefore it is important considering its effects on people. These oxides react with other particles in the environment and merges to the material; in presence of volatile organic compounds and solar radiation, they react generating ozone (O3) which can also have adverse reactions on the respiratory system of people. According to the level of air pollution in different countries, ozone can produce a morbidity rate caused by infections in the respiratory tract.

In such circumstances, several researches can suppose that there is a correlation between atmospheric pollution and infections of this new illness.

To explain these empirical conclusions, epidemiologists point out atmospheric pollution can produce effects in the pandemic of COVID-19 in three ways: raising the propagation, raising the susceptibility and worsening the infection. (3)

According to this context and knowing the new coronavirus also affects the respiratory tract, there is the hypothesis that there exists a correlation between air pollution and the number of deaths produced by COVID-19. For this reason, we are going to analyse some of the most important cities in North America, such as CDMX and Jalisco in Mexico, New York and Washington in U.S.A. and, finally, Ontario and British Columbia in Canada. We are going to compare the levels of nitrogen dioxide in each one of these cities and the mortality rates for COVID-19, so that we will analyze if high levels of nitrogen dioxide in the air can be related to the high number of deaths for COVID-19.

To achieve this, we are going to use techniques of Machine Learning, due to its capacity to identify patterns among the data and therefore to make predictions. Using the outcomes obtained and interference of these new data clusters the team can conclude if the hypotheses posed were correct or not.


Observe | The Challenge

Air pollution is a major global environmental health risk, causing an estimated seven million deaths across the globe annually. Your challenge is to take an interdisciplinary approach, using both Earth science and health science, and integrate different types of datasets and applications to study the effects of air pollution.



HOW WE DID IT?

COVID-19 is an unseen agent but so important in our discussions, conclusions and decisions. It came to change plans, projects, lifes. It came to stay, and now the society must pose new methods to adapt to this new reality. This will only succeed if we analyze the situation that we are living in, search for possible causes of why certain countries are more prone to develop a catastrophe than others, so that society can propose solutions for every challenge we will have.

Thus our team, Space IPN, chose the topic “A One Health Approach” to analyze the context we are struggling in and to set the statistical basis for concluding if effects of air pollution influence the number of COVID-19 deaths.

Like engineers, it is important to use tools which are the most accurate and practical.Therefore, the approach that we decided was “Relationship between the air pollution on the planet, and the incidence of Covid-19”. The project was carried out using Machine Learning in Matlab. But, how did we achieve this?


We chose Six emblematic cities of North America (two of each country) which have similar populations. In the first step, we obtained information about mobility by motorized transport. Also, data about levels of nitrogen dioxide (NO2) in each city were collected. Thus, the consequences of motorized transport on the environment were posed.

In the second step, we collected data of COVID-19 tests carried out for every 100000 inhabitants and the COVID-19 mortality rate in each city. All this information helps us to realize the next: each country carries out different analyzes. There are countries that carry out more number of COVID-19 tests for every inhabitant, such as U.S.A. and Canada, and, in opposite, there are countries that

carry out fewer number of COVID-19 tests, such as Mexico.

If its COVID-19 death rate is high and there are few tests for every inhabitant, this city could return a wrong outcome. On the other hand, if there are many COVID-19 tests, the outcome could be more certain.


Finally, in the third step, afterwards collecting necessary data, we processed the information of the first two steps in Matlab, where the tool KNN was used helping us to classify figures which were more similar. Thus, this is how we could obtain the next outcomes:

Data Analysis

The analyze take a part of the pandemic problem, and the mobility issue. We analyze the mobility for the case of the people who travel to their workplace during the pandemic because in the observed cities the most stayed in quarantine and the pollution made by cars still for those cases.

The Goddard Space Center was the tool to let us observe the concentration of Nitrogen dioxide (NO2) in different cities of the world. We started looking at the data and analyzing the possible ways to infer the problematic of pollution during this pandemic.

The baseline is the measurement of the predicted behavior that has already been made by machine learning and the median data that has been processed by NASA. We use the baseline median data and the median data to program and calculate the percent of change from baseline in MATLAB.


NO2 vs Movility percent change from baseline

Within these graphs, we can observe in most cases, a relationship between the decrease in motorized mobility and the decrease in NO2 levels in the air. Most of the NO2 has its origin in the oxidation of NO that occurs in the combustion of vehicle engines, mainly diesel ones. The NO emitted by the engines, once in the atmosphere, oxidizes and becomes NO2. Therefore, with no transport on the street due to confinement, nitrogen dioxide levels will also decrease.


TEST PER TOUSAND PEOPLE

These data released help us to analyze the certainty of the information released on the mortality rate, in each of the cities. If two populations have the same mortality rate, the city that has carried out more tests for COVID-19 per inhabitant will be more reliable than the city that has carried out the least tests. In these cases, the mortality rate shown will have a significant difference with the real one.


NUMBER OF CASES VS NUMBERS OF DEATHS FROM COVID-19

The relation between number of cases and number of deaths provides the mortality of the COVID-19 in the observed cities. Whether the increase of cases or the decrease of deaths such as the changes in the pollution concentrations can be related for the population in the observed cities.




With the outcomes obtained, we can conclude that there is a relationship between the level of NO2 and death rate of COVID-19. Which is the reason for this?

First, due to the lockdown produced for COVID-19, there was a decrease of motorized mobility and therefore, it led to a decrease in the levels of NO2 in most of the cases. This behavior can be seen in the first 6 charts shown “NO2 percent change from baseline vs workplace mobility percent change from baseline”



References

  1. (GCE), Grupo Consultivo de Expertos. United Nations Climate Change . https://unfccc.int/. [En línea] [Citado el: 03 de 10 de 2020.] https://unfccc.int/sites/default/files/7-bis-handbook-on-energy-sector-fuel-combustion.pdf.

  2. Lara, Abigail R. The Manual´s Editorial Staff. University of Colorado. [En línea] Marzo de 2018. [Citado el: 03 de octubre de 2020.] https://www.msdmanuals.com/.

  3. Narain, Urvashi. Banco Mundial. Contaminación atmosférica: confinada pero no detenida por la COVID-19. [En línea] 02 de julio de 2020. [Citado el: 03 de octubre de 2020.] https://www.bancomundial.org/es/news/immersive-story/2020/07/01/air-pollution-locked-down-by-covid-19-but-not-arrested.

  4. National Aeronautics and Space Administration. Atmospheric Chemistry and Dynamics Laboratory . Global Nitrogen Dioxide Monitoring Home Page. [En línea] 15 de septiembre de 2020. [Citado el: 03 de octubre de 2020.] https://so2.gsfc.nasa.gov/no2/no2_index.html.

  5. Cassidy, Emily. earthdata. [En línea] 26 de Marzo de 2020. [Citado el: 04 de Octubre de 2020.] https://earthdata.nasa.gov/learn/articles/feature-articles/health-and-air-quality-articles/find-no2-data.

  6. Lamsal, L. N., Krotkov, N. A. National Aeronautics and Space Administration. [En línea] 2020. [Citado el: 04 de Octubre de 2020.] https://so2.gsfc.nasa.gov/no2/FAQs.htm.

  7. G., Manuel Oyarzún. SciELO. [En línea] Marzo de 2010. [Citado el: 4 de Octubre de 2020.] https://scielo.conicyt.cl/scielo.php?pid=S0717-73482010000100004&script=sci_arttext&tlng=en.

  8. Google LLC "Google COVID-19 Community Mobility Reports". https://www.google.com/covid19/mobility/ Accessed: 04/10/2020

Connect with #SpaceApps
Google Sites
Report abuse
Page details
Page updated
Google Sites
Report abuse