Team Updates

This video 

https://youtu.be/kvLZUbwTv0g

explains how to use NASA data for this modeling and prediction of COVID19 based on region, season, environment, weather, and human behavior.

m
meg noah

I created some matlab code to show how to acquire data

https://www.mathworks.com/matlabcentral/fileexchange/76451-nasa-earthshots-image-getter

Here are some videos that illustrate it

Short version:

https://youtu.be/EBvNaP-WRTI

Long version:

https://youtu.be/kvLZUbwTv0g

Next it is time to explain how the data are going to be used.

https://youtu.be/kvLZUbwTv0g


Modeling COVID19 Transmission

The transmission rate, R, of COVID19 can be found by best fit to an SIR model. Thus, it is an empirically determined value found from an assumed mathematical model to describe measured data. Once known, R can be used to predict outbreaks. But much of the dependence of R on environmental factors and weather conditions is not known. NASA data can help this process. Together with measured rates of infection from testing that confirms cases, local laws for social distancing and PPE protocols, knowledge of compliance with those laws, and finally, the NASA data, multivariate statistical analysis can be performed for regional, seasonal R values found from fitting measured data to an SIR Model.

Environmental Variables Believed to Influence SARS-CoV-2 Lifetime and COVID19 Transmission Rate

When temperature fluctuations are low and the relative humidity is 50% or less, the corona virus has a low survival rate on its own: less than 1% survive after 2 days. [1] When relative humidity is greater than 80% or lower than 20%, corona viruses survive on surfaces at constant temperature of 20 degrees Celsius.

Source: https://duux.com/en/the-effects-of-temperature-and-humidity-on-covid-19-corona-virus/

The wavelength of UV light (UV-C, 280–100 nm) that is used for disinfecting does not reach sea level as it is absorbed by the Earth’s Ozone layer. But it is not known how much damage various strains of coronavirus sustain when exposed to sunlight’s UV-B (315–280 nm) and UV-A (400–315 nm) bands.

Source: https://en.wikipedia.org/wiki/Ultraviolet

Humidity can play a role several ways. Aerosols with SARS-CoV-2 convectively move the virus allowing it to spread. If aerosols are quickly building mass, they will precipitate faster. Temperature factors are not well understood for SARS-CoV-2 lifetimes, but they do influence how rapidly they move.

Weather conditions also change human behavior which in turn changes their social distancing and frequency of contact. Air pollution reduces the bodies immune system and makes it in particular more susceptible to respiratory illness, leading to higher incidence of comorbidity conditions in the population.

[1] https://www.condairgroup.com/humidity-health-wellbeing/scientific-studies/effects-of-air-temperature-and-relative-humidity-on-coronavirus-survival-on-surfaces

NASA Data Likely to Correlate with COVID19 Transmission Rate

Data sets to begin to look for as correlated predictors:

· Ground Temperature

· Insolation

· UV Index

· Surface Relative Humidity

· Water column

· Cloud Cover (A-Train, GOES, Climate Data Maps)

· Precipitation Rate (A-Train, NOAA Rawinsonde)

· Terrain Type

· Population Density

· Air Pollution

· Surface Wind (A-Train, NOAA Rawinsonde)

Many sources of data were identified that likely correlate in some way with factors that contribute to the rate of COVID19 transmission.  Once modeled, the data can also be used to predict the probability of transmission for a person in a place and time.

My thoughts are that there needs to be 3 separate vectors/scores/metrics: 1) probability of currently being infected, 2) probability of becoming infected, and 3) probability that once infected it will lead to serious condition.  These are complicated to say, and we're only scratching the surface of what is available for data of environment, behavior, and health condition of a person.  I like the idea of bringing the concept of badged dosimetry reading into a person's life - out in public and on the worksite.  And then being able to make predictions that use estimates of viral load, exposure duration, and strains, to bring into the transmission probability.  Environment also plays a factor as temperature and humidity both impact the viability of the virus in the aerosols, air, and on surfaces; as well as change the precipitation out of aerosols because they can grow faster and gravity takes them out of the arena of likely contact.  It would be nice if people didn't just have to 'stay home' or wrap themselves in lots of PPE all the time by being able to take these into consideration and provide real-time risk assessments and make recommendations for future behavior much like a traffic rerouting App.


m
meg noah