High-Level Project Summary
We developed some code to read data from the satellites which use the InSAR technique to measure Earth’s deformations. This allowed us to visualise the data and look at deformation, coherence, connected components and amplitude. We came up with some theoretical methods to counter Earth’s Tropospheric signal interference, however, we struggled with applying this knowledge to our implementation.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
Our first step was to investigate and research the InSAR technologies used by the satellites whilst measuring Earth’s deformation.
We downloaded the .nc (NetCDF) Ridgecrest, Faultcreek and Kilauea data-sets using the python scripts provided. We then wrote code in Python to use the matplotlib, netCDF4, and the Python data analytics, numpys, and pandas libraries to read and plot and visualise these data-sets. We planned on using using a machine learning library in Python called sci-kit learn to utilize a clean set of data to train against test data. This would allow us to feed it raw InSAR data and clean it up from the noise and interference from Earth's water vapour. We came up with theoretical ideas that would help us clean the raw data-sets. We visually compared the plotted data. We then wanted to omit outliers, because of how the data was provided we were unable to amend the data.
We pushed all of our code and our notebook to the cloud. this is the link that will take you to the notebook. https://hub.gke2.mybinder.org/user/paxtonion-space-nasa-space-apps-1dzd6438/doc/tree/notebook
Space Agency Data
We used NASA's data-sets. We used it to plot the data into visual representations of the deformations measured and recorded by the InSAR technologies.
Hackathon Journey
This Hackathon journey really inspired us to work on this project! As university students this was a blast to work on. These last 36 hours consisted of a lot of collaboration, brain storming, and research. We were pushed within our limits to work under such tight time constraints on a project this big and complex. We chose this challenge as we thought we would have more oppurtunity to mathematically approach and solve then implement this solution but we struggled a lot during the interpretations of the data provided. We would like to thank everyone who was involved in making this possible for us to participate in and the company we're interning for hosting and supporting us throughout this event.
References
https://www.spaceappschallenge.org/space-apps-challenge-2022-example-resource-insar-change-detectives/
https://scikit-learn.org/stable/model_selection.html#model-selection
https://matplotlib.org/
https://unidata.github.io/netcdf4-python/
Tags
#software #math #InSAR #satellites