MODIS-Aqua PFT composition: data, code and reproducibility package
Year: 2026
Type: dataset
Abstract: Research data and analysis code accompanying “Field-target support and phytoplankton composition estimation from MODIS-Aqua observations.” The package includes attributed field-source tables, 8,104 phytoplankton composition targets, 121,560 target-date candidates, decoded ten-band MODIS-Aqua pixel windows, partition and campaign assignments, saved predictions and model checkpoints, analysis and figure code, dependency versions, data dictionaries, and checksums. The reproduce.py command reconstructs reported statistics, checks saved-model predictions, and generates analytical figures. The refit.py command trains the current study configurations from the included inputs. Fresh CPU fits are documented separately from archived GPU results. Full global NASA rasters are accessed through NASA; exact product access records and extraction code are provided. Original analysis code is licensed under MIT, and original derived data and figures under CC BY 4.0. Third-party materials retain their respective licenses. Consult README.md, REPRODUCTION_CHECK.md, and license_index.csv for execution instructions, verification scope, and component-specific terms.
Author: Lee, Myungin
Institution: University of Maryland, College Park
Language: English
Subfield: Oceanography
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7219792124
