Automated Detection of Hazards
SETH
Summary
The SETH project is based totally in Data Science and Machine Learning, we propose our idea as a form of prevention and greater understanding of climate change and how it evolves, leaving repercussions on our ecosystems, but equally In order to find out what human actions have considerable repercussions on it, thus achieving greater knowledge and therefore better control in the face of the generation of natural disasters that impact the socioeconomic sector and obviously the environment.
How We Addressed This Challenge
There are innumerable phenomena that recurrently affect ecosystems and, derived from this, human security. These phenomena can and are tracked by satellite data, so the challenge to be solved is to use the open source data provided by NASA in such a way that one or more predictive and analysis models can be built using Machine Learning. These models will help the analysis and prediction of a specific phenomenon: drought. Likewise, for the construction of the prediction model and the analysis it will be intended to use auxiliary data such as factors of climate change, change in soil structure, fires, etc., having as a final result a graphical interface accessible through an API.
How We Developed This Project
By consulting satellite data regarding climatic conditions, as well as factors related to the phenomenon of drought, it is proposed to carry out a mathematical model that manages to determine the correlation of events and variables in such a way that with said model a model of machine learning for predicting droughts anywhere in the world. In the same way, models can be created, based on the analysis of the correlation of variables, that allow the analysis of the factors involved during droughts as well as the socio-economic effects that they can cause in different regions. For this, the initial model will start from a delimited area, as well as from events in a given period of time. Said tool will be implemented through a RESTful API model that allows future data to be consulted in such a way that the people involved in the investigation of said phenomena can access in a more direct way and can receive clear information.
How We Used Space Agency Data in This Project
The proposal to solve the problems related to the impact caused by droughts is the construction of a web service that provides the necessary data and tools to the timely and accurate prediction of this phenomenon. For this, a system composed of three main phases is proposed:
- Open source data consumption: In this phase, the system would consume data (through web requests) to the service provided by the platform of the Japanese Aerospace Exploration Agency (JAXA) on the site JASMIN (described later). Given the characteristics of the platform, the requested data would be those corresponding only to the Asian area, and from these the information corresponding to variables such as precipitation, soil moisture, surface temperature, solar radiation and mainly the drought index would be consulted. throughout a given epoch. The initial proposal is for the period of a decade (2010-2020), and in the future it may be constantly consulted depending on the updates that are made in the data provided.
Project Demo
https://drive.google.com/file/d/1A3Z7gjQa-N9H6kqcSIWFZpOzaHYL9Ex5/view?usp=sharing
https://drive.google.com/file/d/1nqBdJ0tKMEW-WEn2AbPJ-UWSQNutya9m/view?usp=sharing
Project Code
Data & Resources
In this project we only use data of JAXA because of the complete datasets that we found in the 48hrs. of realizing our project.