遇见数据集

CHLA-Z: Global chlorophyll-a vertical distribution (0–200 m) derived from PACE OCI and Bio-Argo (draft)

收藏
Zenodo2026-01-12 更新2026-05-26 收录
官方服务:

资源简介:

This is the initial public draft release of CHLA-Z, a global gridded dataset describing chlorophyll-a vertical structure (0–200 m) derived from PACE OCI remote-sensing reflectances and in-situ Bio-Argo observations using a boosted regression tree (BRT) model. WEBSITE This release is intended to establish provenance, citation, and a stable access point for the dataset and associated documentation. The large CHLA-Z data files are hosted separately in cloud storage and are not archived directly in this release. Dataset available via NOAA Open Data Dissemination (NODD) on Google Cloud Storage gs://nmfs_odp_nwfsc/CB/fish-pace-datasets/chla-z Public HTTPS access (no authentication required):https://storage.googleapis.com/nmfs_odp_nwfsc/CB/fish-pace-datasets/chla-z/ Included in this release: Dataset landing page and documentation (in docs folder) Link Notebooks with the exact pipelines used to create the dataset (in notebooks folder) Example notebooks (Python / Colab / R) Example notebook | Open in Google Colab Metadata describing variables, grid, and access patterns Model configuration and supporting utilities Training data subsets, model configuration, and supporting utilities Data access: Zarr v3 (cloud-optimized): gs://nmfs_odp_nwfsc/CB/fish-pace-datasets/chla-z/zarr Daily NetCDF files: gs://nmfs_odp_nwfsc/CB/fish-pace-datasets/chla-z/netcdf # Example access via xarray and python; zarr > 3 required # See example notebook or docs/index.html for more examples in Python and R import xarray as xr zarr_url = "gcs://nmfs_odp_nwfsc/CB/fish-pace-datasets/chla-z/zarr" ds = xr.open_zarr( zarr_url, consolidated=False, storage_options={"token": "anon"} ) Status: This is a research/draft product. Validation, uncertainty characterization, and finalized methods documentation are in progress. In particular, the training data are currently limited to Bio-Argo and OOI and we are actively working to increase the training data (in-situ CHLA at depth) from other platforms. Variable names, grid definition, and access paths are expected to remain stable; algorithm details and derived metrics may evolve in subsequent versions. This release is intended for exploratory analysis, method evaluation, and early scientific use. Citation: Please cite this specific release DOI for reproducibility. A concept DOI representing the CHLA-Z dataset as a whole will always resolve to the latest version. License: CC-BY 4.0 unless otherwise noted.

提供机构:
Zenodo
创建时间:
2026-01-10
二维码
社区交流群
二维码
科研交流群
商业服务