Published November 19, 2024 | Version 20241119
Dataset Open

Global Airborne Observatory: Example Rapid Water Quality Maps of Hawaiian Island Coastline

Authors/Creators

  • 1. ROR icon Arizona State University
  • 2. ROR icon Brockmann Consult (Germany)

Description

Example maps from the publication:

Nicholas R. Vaughn, Marcel König, Kelly L. Hondula, Dominica E. Harrison and Gregory P. Asner. 2024. Rapid Water Quality Mapping from Imaging Spectroscopy with a Superpixel Approach to Bio-Optical Inversion. Remote Sensing. (in press)

Data:

Maps of water quality and auxiliary information produced from Global Airborne Observatory (GAO) VSWIR Spectrometer data for five example sites:

Site Island Date Collected Time Collected (24hr) Latitude Longitude
1. Pelekane Bay Hawaiʻi 30 October 2023 10:02 20.0243 −155.8256
2. East Kaho’olawe Kaho’olawe 22 January 2024 09:15 20.6027 −156.5638
3. Hilo Bay Hawaiʻi 15 January 2023 11:30 19.7353 −155.0636
4. Mā’alaea Bay Maui 22 January 2024 10:19 20.7805 −156.4842
5. South Moloka’i Moloka’i 7 January 2024 11:31 21.0866 −157.2159

There are 41 maps for each example site:

  1. A single 3-band true-color map of the example site for reference (Site<X_Name>_3band.tif)
  2. A single full-resolution water quality map produced using a bio-optical inversion model on each pixel (Site<X_Name>_PixelbyPixel.tif)
  3. For each of 13 different superpixel size settings (50px, 100px, 250px, 500px, 600px, 750px, 1000px, 1250px, 1500px, 2000px, 3000px, 4000px, 5000px), a map of the superpixel ID of each pixel (Site<X_Name>_Size<####>_ClusterID.tif)
  4. For each of 13 different superpixel size settings, a superpixel-level map of water quality as estimated at the cluster level. (i.e., each pixel in a superpixel will have the same values, Site<X_Name>_Size<####>_Cluster_Preds.tif)
  5. For each of 13 different superpixel size settings, a full-resolution water quality map produced by interpolation of the superpixel level maps in 4. above (Site<X_Name>_Size<####>.tif)

All files are in GeoTiff format using 32-bit floating point values. Most are compressed with the LZW algorithm. Maps are not georeferenced. No data value for map types 2-5 is -9999.

Files

Zenodo.zip

Files (1.7 GB)

Name Size Download all
md5:259ce7f8a18b49e7d9b7913ad86ac894
1.7 GB Preview Download