Team Gaia| Automated Detection of Hazards
Automated Detection of Hazards
Bloom Pointer
Summary
The Bloom Pointer is an interface in which, with the application of machine learning, is possible to obtain the analysis of information from the database provided by NASA and make a notification over the indication of a critical point observed. The main goal of it is the optimization of the data analysis over the determination of a pattern as far as possible to understand how a region reaches a critical point of cyanobacterial proliferation. So that, with this identified pattern, it can prioritize the data of a certain region that tends to reach a critical point providing useful data to the researchers.
How We Addressed This Challenge
The challenge requires us to develop a system that can help researches and decision-makers to understand the bloom phenomenon, which, once our project is completed, it should demonstrate in a simple and easy way, with graphics elements the areas affected and what could happen with that situation. Besides that, the program should acquire data direct from the satellites using images comparison and enhanced by the machine learning to optimize the global monitoring. Thus, our project fulfill all elements requested in the challenge.
How We Developed This Project
Recently occurred the case of cyanobacterial proliferation which had harmful effects in Africa that contaminated and killed many elephants, but that´s one of many cases, including those can affect not only animals, but also people. Our team chose this challenge because we understand how important is to take action and try to reduce these cases of cyanotoxins contamination, considering its frequency whom we are seeing this phenomenon, all over the world, including our country, Brazil. And also the fact it is something that unbalances the environment and can bring many ravages for animals, public health and economy.
We created the idea of Bloom Pointer by doing an interface with the application of machine learning that aims to detect and alert when there is some region in warning or critical state, using near real time satellite data.
We used python as the main code language and libraries such as SVM - Sklearn, opencv and cvtcolor for the machine learning.
As soon as the challenge started, we had our first problem, one of our team members had medical problems and could not help in the developing. Not being bad enough "loss" one member, it was our programmer. so the others members have to try to learn and understand what are the principles and how we should code to solve this challenge. Besides that, we could learn how to use new tools such as Invision and Powtoon to make the simulation of the final software (both websites focus on design).
How We Used Space Agency Data in This Project
The idea is to get data from NASA and JAXA satellites database by Chlorophyll a, Ahlorophill A (L2) e Fraction of Photosynthetically layers, creates a scraping image model, developed in Python Language, with the concepts of machine learning, being able to identify the critical points of cyanobacterial proliferation and make an alert notification over a worldwide map.
Project Demo
https://youtu.be/3vOCSHhv324
Data & Resources
- MS, Ministério da Saúde (Brasil). Cianobactérias/Cianotoxinas: procedimentos de coleta, preservação e análise. Brasília/DF, 2015. <https://portalarquivos2.saude.gov.br/images/pdf/2015/janeiro/19/cianobacterias-cianotoxinas-2...pdf>. – Acesso em: 2 de outubro de 2020.
- CDC, National Center for Enviromental Health. Cyanobacteria Blooms FAQs: When in doubt, it’s best to keep out!. - < https://www.cdc.gov/habs/pdf/cyanobacteria_faq.pdf>. Acesso em: 3 de outubro de 2020.
- MARSHALL, Michael. “The event that transformed the Earth”, 2 July 2015. <http://www.bbc.co.uk/earth/story/20150701-the-origin-of-the-air-we-breathe>. – Acesso em: 2 de outubro de 2020.
- CIANOBACTÉRIAS TÓXICAS E PROCESSOS DE REMOÇÃO, tese de Taíssa dos Santos Barbosa, 2009. <https://repositorio.ufmg.br/bitstream/1843/BUOS-99UHCT/1/monografia_ta_ssa.pdf>. – Acesso em: 2 de outubro de 2020.
- ESTUDO DAS FLORAÇÕES DE CIANOBACTÉRIAS PRODUTORAS DE TOXINAS PARALISANES E DESENVOLVIMENTO DE METODOLOGIA DE PURIFICAÇÃO DAS TOXINAS, tese de Wilson Alves Colvara. Rio Grande, RS, Brasil, 2012. <https://sistemas.furg.br/sistemas/sab/arquivos/bdtd/0000011382.pdf>. – Acesso em: 2 de outubro de 2020.
- NASA, Earth Science Division of the NASA Science Mission Directorate, Applied Sciences Program. Detecting Harmful Algal Blooms, August 13 2020. <https://appliedsciences.nasa.gov/our-impact/story/detecting-harmful-algal-blooms>. – Acesso em: 3 de outubro de 2020.
- EPA, United States Environmental Protection Agency. Cyanobacteria Assessment Network (CyAN). <https://www.epa.gov/water-research/cyanobacteria-assessment-network-cyan>. – Acesso em: 3 de outubro de 2020.
- EPA, Satellite monitoring of cyanobacterial harmful algal bloom frequency in recreational waters and drinking water sources. <https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NERL&dirEntryId=336500>. – Acesso em: 3 de outubro de 2020.
- INPE, Introdução ao Sensoriamento de Sistemas Aquáticos. Princípios e Aplicações, 2019.
<http://www.dpi.inpe.br/labisa/livro/res/conteudo.pdf>. – Acesso em: 3 de outubro de 2020.
- INPE, Nova Metodologia para o monitoramento sistemático de qualidade da água de reservatórios tropicais, 2019.
<http://www.obt.inpe.br/OBT/noticias-obt-inpe/nova-metodologia-para-o-monitoramento-sistematico-da-qualidade-da-agua-de-reservatorios-tropicais>. – Acesso em: 3 de outubro de 2020.
- NASA,Earth Data, Level-3 Browser, Chlorophyll Concentration. August 13 2020. <https://oceancolor.gsfc.nasa.gov/l3/>. – Acesso em: 3 de outubro de 2020.
- JAXA, Himawari Monitor. <https://www.eorc.jaxa.jp/ptree/index.html> – Acesso em: 3 de outubro de 2020.