Skip to content
 
 

Repository files navigation

WetLSP v2

Fork: Updated algorithms (July 2025) to derived a land surface phenology product from PlanetScope imagery for FLUXNET-CH4 v2 sites. Data developed and initially shared for wetland methane research applications in the McNicol ecophilab, the ESIIL AI for Natural Methane Working Group, and the USGS Powell Synthesis for Wetlands.

Presently, the pipeline generate daily pixel-wise EVI after PS processing, filtering and gap-filling using smooth spline. These procedures could be revised and improved.

To run updated code:

  • Clone the repo into your workspace

  • Create directories listed at top of PLSP_Parameters_gm_v1.json and updated filepaths

  • Use 00_img_download.py to retrieve data via PlanetScope API into created rawImage dir

  • Run series of 5 processing scripts

  • Ignore pipeline/validation and pipeline/timeseries for downstream analysis/time series retention

  • Downstream workflows:

    • nc-check-populate.Rmd to output standardized .nc phenometric files per site per year (2021-2024)
    • core-site-selection.Rmd to identify suitable sites with good coverage for analysis
    • review-final-wetlsp.Rmd to inspect phenometric products by site-year and observeed and gap-filled values by site
    • wetlsp-timeseries.Rmd to create fetch perimeter, extract and retain pixel-wise EVI (obs & filled) in .RDS
    • review-fetch-rds.Rmd to inspect the fetch perimeter and visualize/explore .RDS object

Quirks:

  • Note that original geojson files must be stored within /AMFLX subdirectory
  • _gm_v1 workflow skips water mask steps in 03_LSP_script, should not break process, but use if available

Origin: Algorithms to derive a land surface phenology product from PlanetScope imagery for AmeriFlux and NEON sites

  • Publication: Moon, M., Richardson, A.D., Milliman, T. and Friedl, M.A., 2022. A high spatial resolution land surface phenology dataset for AmeriFlux and NEON sites. Scientific Data, 9(1), p.448. https://www.nature.com/articles/s41597-022-01570-5
  • Data: Moon, M., A.D. Richardson, T. Milliman, and M.A. Friedl. 2023. Land Surface Phenology, Eddy Covariance Tower Sites, North America, 2017-2021. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/2033

About

Algorithms to derive a land surface phenology product from PlanetScope imagery for AmeriFlux and NEON sites

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages