Carpe Charmin - Space Apps Challenge

Carpe Charmin| An Integrated Assessment

An Integrated Assessment

Your challenge is to integrate various Earth Observation-derived features with available socio-economic data in order to discover or enhance our understanding of COVID-19 impacts.

A multivariate approach to understand differences in national severity of Covid-19

Summary

The deadliness of the Covid virus has varied dramatically between countries. There is a desperate need to understand these variations so that we can reduce the fatality rate for this and any future pandemics. The global nature of this pandemic provides a unique opportunity to assess the key determining factors of the severity of the virus in a multivariate approach attempting to correlate to data from each country about their population demographics and density (urbanization), healthcare and longevity, pollution, and population mobility as measured by energy and fuel consumption, and their governmental policy approaches (speed and effectiveness of quarantining strategies).

How We Addressed This Challenge

We have collected available data and used a multivariate analysis to assess the extent to which each of these factors can contribute to explain the variability of the fatality rates observed.

How We Developed This Project

We felt that the large geophysical data sets available gave a unique opportunity for understanding. For example we used the Nasa Sedac data on pollution (EPI) as an attempt to assess the importance of preexisting lung issues. Many data sets were taken from the web, particularly government response to Covid that was compiled in Oxford, and country data on virus fatality at John Hopkins and wolrdmeters. Energy consumption, demographics and urbanization data were compiled from various sources. The principle tools for analysis were Tableu, R, and Excel.

Project Demo

A generalized linear statistical model including 4 primary variables was able to account for 39% of the variance of the data. These variables were: 1) the ratio of people over 64 years old to the working age population (15-64 yrs), 2) Percentage of population in an urban setting 3) number of hospital beds/1000 people 4) Covid-19 tests per million people. All of these variables were significant (P value less than 0.05).

When we added EPI NASA pollution data to this analysis the new 5 parameter model accounted for 44% of the variance.

However The EPI was the most significant variable. In a single factor model it accounted for 35% of the variance. (P value of 2.5 e -7)!!!

So the pollution was by far the strongest predictor of Covid-19 mortality. It makes some intuitive sense that people who have lungs damaged by poor air quality would be at higher risk to Covid-19, but the correlation is startling.