Space Apps Challenge 2018

Introduction

The influence of air quality on human and ecological health is hard to overstate, and as such we consider it paramount to complement current data with alternative methods. There is still a lot of room for improvement in terms of satellite precision and accessibility, while ground-level measurements require constant work and maintenance. So we decided to do our part looking for new solutions by making that job easier. We ended up finding such solution in a place we didn’t expect: everyday people and their phone’s cameras.

The key issue to solve was how anyone could generate that information, so that was when bioindicators came into play. By recognizing the predominant lichen species in our city’s trees and providing the user with pictures and a guide to identify them in situ, the information can then be uploaded to our collaborative map and leveraged by both scientists and citizens.

Background

When we started investigating about air quality in Córdoba we quickly found out information was sparse and unreliable. For example we found news sites reporting in April 2017 there hadn’t been any measurements of air quality since 2013[1], only to report by September that research had resumed and linking to a website that just a year later is already dead[2]. The last widespread article regarding the topic is over 8 years old, and paints a grim landscape regarding air quality[3].

We eventually realised we could use bioindicators such as lichens, which already had decades of scientific research backing our use case. We were eventually able to find out the lichen species present in Córdoba identifying them in pictures, how they reacted to different pollutants according to a study in our city, how to collect field information about these organisms using a grid method together with different specifications for the selection of the phorophyte, and most importantly what formulae we could use to make sense of the large volumes of data, like IAP (Index of Atmospheric Purity) and Shannon’s Diversity Index.

About the app

How it’s used:

The users are required to select the trees according to a a few conditions:

- Rough crust.

- Diameter of the trunk greater than 10 cm.

- Not injured or painted on.

- Take the picture from the southwestern side of the trunk, at a height of 1,5 to 2 m.

- Angle of inclination of the tree not greater than 10 degrees.

The app allows users to:

• Learn how to identify the different lichen species present in Córdoba. And how to take the measurements.
• Take pictures of a standardized area of a tree covered with lichens (The user will only need a 20 cm by 50 cm rectangle made of paper or other cheap material to use as reference)
• Identify quadrants in which specific species appear.
• Identify the surface of each species present. It is a manual process which will be replaced by advanced deep learning algorithms when the app has collected enough training data.
• Find out the IAP and Shannon’s Diversity Index on their current location based on the previous measurements.
• Compare, as more data is obtained, the IAP and DIversity Indexes of different locations.

Source code for the app is published on :
https://github.com/lauranvelazquez/LichenApp.git

Source code for the server is published on :

https://github.com/lorellana672/Open-Lichen/commit...

The server is published and running on :

http://openlichen.herokuapp.com/

Impact

• For the local and national government

It will provide valuable information for environmental administration and backup projects like the creation of more green spaces, and an affordable alternative to the usual air quality monitoring stations.

• For scientists

Scientist will be able to access all the information gathered in one platform and use it to replicate previous studies and produce new research.

• For the population of Córdoba

Once enough data is gathered we will be able to estimate air quality and concentration of different pollutants with health effects such as[4]:

- PM10y PM2,5: Increases risk of acute lower respiratory infections, cardiovascular disease, chronic obstructed pulmonary disease, and lung cancer.

- Tropospheric ozone: Causes breathing problems, triggers asthma, reduces lung function and causes lung diseases

- Nitrogen dioxide: Increases symptoms of bronchitis in asthmatic children in association with long term exposure and in some concentrations can reduce lung function

- Sulfur dioxide: Causes irritation of the eyes, coughing, mucus secretion, aggravation of asthma and chronic bronchitis, and makes people more prone to infections of the respiratory tract.

• For students

This project can be used as a tool to interest students of all ages in environmental responsibility and civic engagement by using the app and its findings at local schools. This may prove to be the catalyst for innovative projects and new scientific vocations.

• For NASA

Many research papers have demonstrated the correlation between lichen biodiversity and the concentration of pollutants, but this app could be used to further investig ate on the topic by studying specific species’ sensitivity and also provide ground level verifications for satellital findings.

Looking Ahead: Plans for Future Development

• We need a base measurement, that we may obtain either by transplanting known species to a lab and observing their biodiversity and behaviour with high air quality or by measuring them away from the city. This is necessary before we can estimate a scale of air quality that we can provide to the user.
• Generate forms to include the comparative data resulting from other studies with different methods like analysis of the lichen in labs or using gaseous pollutant measurements, and have the best estimations possible.
• Include ecological and climatic data such as wind direction, temperature and precipitations to define the limits of the studied area and to see if the model can be extrapolated to other places around the world.
• As we said before, we are working on improving the detection system, from manual to automatic using advanced deep learning algorithms.
• We want to add vectorial information about pollutant sources using official, user generated and company declared data. This analysis may include traditional factories, highways, farms among other sources.
• The ultimate goal is to be able to perform intertemporal analyses that help understand the evolution of pollution and predict future trends.

References

[1] http://www.lavoz.com.ar/ciudadanos/sin-monitoreo-del-aire-en-la-ciudad-y-en-el-pais

[2] https://www.cadena3.com/noticias/sociedad/volveran-medir-calidad-del-aire-cordoba_117695

[3] http://www.lavoz.com.ar/ciudadanos/ambiente/detectan-alta-concentracion-esmog-fino

[4] http://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health

Bibliography

• Estrabou et al. (2011). Air quality monitoring system using lichens as bioindicators in Central Argentina. Environ Monit Assess, 182,375–383.
• Filippini, E. (2017). Bases para un sistema de monitoreo de calidad de aire con bioindicadores en áreas bajo diferentes usos de suelo. Facultad de Ciencias Agropecuarias. Universidad Nacional de Córdoba.
• Gonzales Vargas et al. (2016). Aplicabilidad de líquenes bioindicadores como herramienta de monitoreo de la calidad del aire en la ciudad de Cochabamba. Acta Nova, 7(4), 455-482.
• Hawksworth et al. (2005). Líquenes como bioindicadores inmediatos de contaminación y cambios medio-ambientales en los trópicos. Rev Iberoam Micol 22(7) 71-82.
• Mares Rueda, I. (2017). Líquenes como bioindicadores de la calidad del aire. Facultad de Farmacia, Universidad Complutense.
• McCarthy, D. (2001). Digital analysis of lichen cover: a technique for use in lichenometry and lichenology. Artic, antartic and alpine research, 33(1), 107-113.
• Pardo González, A. (2017). Líquenes como bioindicadores de la calidad del aire. Facultad de Farmacia, Universidad Complutense.
• The Institute for Environmental Modeling. Diversity indices: Shannon's H and E. University of Tennessee.