AI-Event-Monitoring

The goal of this project is to create a deep learning model that can recognize and mask significant deformation events in InSAR interferograms. View the documentation here.

Table of Contents

• Installation
• Setup
• Commands
• Running Unit Tests
• Synthetic Interferograms
• Simulated Interferograms
• Preliminary Results
• References

Installation

pip

pip install insar-eventnet

From source

Clone the repository

git clone https://github.com/asfadmin/AI-Event-Monitoring.git

Activate the conda environment from the environment.yaml file

conda create env -f environment.yaml
conda activate insar-eventnet

then, install the insar-eventnet package and cli

pip install .

Setup

Run the setup command in your desired working directory

insar-eventnet setup

This will add the data directory which is structured like this

data/
      └──input/
        └──products/
        └──aoi/
      └──working/
        └──real/
        └──synthetic/
      └──output/
        └──models/
        └──mask/
        └──tensorboard/

You should now be ready to run everything.

Command line interface

There are two command line interfaces, inference.py which utilizes the cloud API for inference, and insar-eventnet which runs the models locally.

inference.py

Running inference

python inference predict-event [usgs-event-id] [product-name]

aievents.py

For a List of Commands

insar-eventnet --help

For Detailed Usage Information

insar-eventnet [command] --help

Viewing a Random Synthetic Masked/Wrapped Pair

insar-eventnet show-random

Viewing a Random Simulated Masked/Wrapped Pair

insar-eventnet simulate

Creating a Synthetic Masked/Wrapped Dataset

insar-eventnet make-synthetic-dataset [dataset-name] [dataset-size] --tile_size [nxn-size-of-images]

Creating a Simulated Masked/Wrapped Dataset

insar-eventnet make-simulated-dataset [dataset-name] [dataset-size] --tile_size [nxn-size-of-images]

Viewing a Pair from a Dataset

insar-eventnet show [path/to/dataset.npz]

Training a Model

insar-eventnet train-model [model-name] [path/to/training-set] [path/to/testing-set] --epochs [num-of-epochs]

Testing a Model

insar-eventnet test-model [path/to/model]

Mask a Real Interferogram

insar-eventnet mask [path/to/model] [path/to/product_folder] --tile_size [size-of-tiles-used-to-train]

Running Unit Tests

Currently, test coverage is limited. However, they can be run with pytest by simply typing:

pytest

in the root of the project directory.

Synthetic Interferograms

Synthetic Interferograms are generated using more simple math than the simulated ones. This means that the datasets can be created more quickly; although, the simulated interferogram generation is still fairly quick and recommended over this.

Synthetic Masked/Wrapped Pair Example

Simulated Interferograms

Simulated Interferograms are comprised of simulated deformation using Okada's model, simulated turbulent atmospheric error using a FFT method, simulated topographic atmospheric error, and simulated incoherence from turbulent atmospheric error. Most of the functions related to the simulation come from this project by Matthew Gaddes which was used for this 2019 JGR:SE paper.

Simulated Masked/Wrapped Pair Example

Simulated Masked/Wrapped Pair Example (Atmospheric Phase Error only)

Basic Model's Mask Examples

These results come from a basic model trained on a simulated dataset with 1000 samples, 900 for training and 100 for validation.

An Earthquake in Iran

A Negative from the Coast of Greenland

References

Gaddes, M. E., Hooper, A., & Bagnardi, M. (2019). Using machine learning to automatically detect volcanic unrest in a time series of interferograms. Journal of Geophysical Research: Solid Earth, 124, 12304– 12322. https://doi.org/10.1029/2019JB017519