High-Level Project Summary
The power of ten can be a concept extended both to the past and the future. With our approach, we are trying to predict heat levels of the planet for the next 10 years, based on the previous 10 years of historical data, provided by NASA satellites that track heat signatures in areas around the planet.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
Using sample data with temperatures for Athens city of Greece, and Python Prophet framework we can predict the future temperatures in a 10 year span.
All we need is a running python environment and the dataset itself.
Installing relevant libraries such as matplotlib to visualize the data trend is also a step required.
The power of ten can be a concept extended both to the past and the future.
With our approach, we are trying to predict heat levels
of the planet for the next 10 years, based on the previous 10 years of historical data, provided by NASA satellites that track heat signatures in areas around the planet.
As a reference area we will be using country of Greece, a Mediterranean and on average a hotter country.
The objective is to identify a trend (is it getting hotter or colder in Greece) and generate a prediction for the next 10 years.
For this we will be using Data Series tabular data and the Prophet Python framework to analyze and visualize them. With the visualization we aim to create possibly a stimulus for even better frameworks that can predict heat-waves via recurring patterns.
Also help for early warning and mitigation of forest fires, as they are directly linked to persisting heat-waves in an area.
Space Agency Data
https://nsidc.org/data/atl21/versions/1
Accessing data from ATLAS (Advanced Thermal and Land Applications Sensor) and exploring downloaded XML's may reveal other patterns.
and
https://nsidc.org/data/SNEX17_KT15_TB/versions/1
Hackathon Journey
Installing Python and all relevant libraries was a journey itself.
As always alot of help was acquired of of the Internet from sites such as stackoverflow, that is always a good reference for answers to problems any developer can come across.
Also Prophet framework and online youtube lessons and guides played a critical role as I am not an expert in Python.
Just a curious learner.
The most difficult to find are NASA resources. For example I wanted to use image from heat-capable satellites.
This was not readily available, and no guide to it was available also.
Eventually I found data from https://nsidc.org/data/SNEX17_KT15_TB/versions/1
References
https://facebook.github.io/prophet/
https://www.python.org/
https://earthdata.nasa.gov/eosdis/eosdis-data-news-archive/eosdis-data-news-august-2021
https://nsidc.org/data/SNEX17_KT15_TB/versions/1 - (SnowEx17 KT-15 Infrared Brightness Temperature)
Tags
#heat #temperatureimages #satellite #python #prophet