Working With Datasets

The SHIFT AVIRIS-NG data is located in several spots. The gridded Zarr data can be found in mounted S3 bucket dh-shift-curated, located in the ‘aviris’ folder. The L1 and L2a products can be found in efs/efs-data-curated. Both are organized by the date of the flight. Additional campaign raster pre-grid and vector data is available in the dh-shift-curated bucket in their respective folders.

This user guide has several example notebooks demonstrating how to work with AVIRIS data which cover: A quick start guide for SHIFT Data (reading and writing data, clipping with shapefiles), Xarray basics, aligning data with different resolutions, data visualization and clustering. Additionally, on the ‘dh-shift-shared’ mounted S3 bucket there is a notebook template for getting started.

• SHIFT Data Quickstart Guide
  • Working with SHIFT Gridded Data
  • Working with the Raw SHIFT Data
• Loading SHIFT Data with Intake
  • Reading in Data
  • Working with L1 and L2a Data
  • Orthorectifying a Dataset
• Xarray Basics
  • Streamlined Xarray Interface to AVIRIS Datasets
• Aligning Data with Different Resolutions
• Visualizing Data
  • Plotting an RGB image
  • Using Holoviews with a Pandas Dataframe
  • Using Holoviews with Xarray
• Clustering Example
  • Pre-processing
  • PCA
  • Kmeans
  • Gaussian Mixtures
• Accessing SHIFT Intake Catalog With R
• Aligning Field and SHIFT Data

Additional examples can be found here: https://github.com/marinadunn/SHIFT-STAC-demo; specifically, take a look at this notebook.