| Summary: | This archive provides datasets related to the following publication: V. Tollenaar, H. Zekollari, F. Pattyn, M. Rußwurm, B. Kellenberger, S. Lhermitte, M. Izeboud, D. Tuia, Where the White Continent is blue: deep learning locates bare ice in Antarctica. Geophysical Research Letters, 50, e2023GL106285 (2024). DOI:10.1029/2023GL106285 Contact: Veronica Tollenaar, Veronica.Tollenaar@ulb.be Users should cite the original publication when using all or part of the data. About the datasets: it includes the seven-band MODIS composite ("merged_bands_composite3031080910.tif"), the outlines of our set of training ("train_squares.zip"), validation ("validation_squares.zip"), and testing regions ("test_squares.zip"), the predictions on the presence of blue ice ("BIA_map.nc"), the corresponding uncertainty estimates ("BIA_map.nc"), and the smoothed blue ice area outlines ("smoothed_BIAs.zip"). References to the literature are provided in the corresponding manuscript. - BIA_map.nc: netcdf in which the band "mean" represents the predictions on the presence of blue ice, and the band "std" the corresponding uncertainty estimates, in EPSG:3031 - handlabels_sqXXX.zip: archived .cpg, .bdf, .prj, .shp, and .shx files of the handlabeled blue ice areas (Figures S3 and S4) in the respective squares (indicated in the files "xxxx_squares.zip") - merged_bands_composite3031080910.tif: 7-band tif file of the 7-band MODIS composite (Figure S1, Text S1) - smoothed_BIAs.zip: archived .cpg, .bdf, .prj, .shp, and .shx files of the smoothed outlines of our BIA predictions (Text S5) - xxxx_squares.zip: archived .cpg, .bdf, .prj, .shp, and .shx files of the outlines of training, validation, and testing regions (Figure S2)
|
|---|