Skip to content

Latest commit

 

History

History
76 lines (57 loc) · 2.88 KB

NEWS.md

File metadata and controls

76 lines (57 loc) · 2.88 KB
Raw

Known issues: https://github.com/PredictiveEcology/LandWeb/issues

LandWeb v2.0.1

App changes

  • fixed issue with downloading model outputs.
  • minor performance enhancements.

LandWeb v2.0.0

Input data

  • reworked data preparation steps and fixed many data errors
  • implemented and fit various statistical models to impute missing and inconsistent data

Model improvements

  • reparameterized the model to emulate a state-transition model to ensure burned pixels regenerate
  • several bug fixes and performance enhancements

App changes

  • reworked app to use pre-generated results only instead of computing on-the-fly; as a result, only selected results per FMA are available in-app.
  • several other performance enhancements.
  • removed shapefile uploads, as custom areas require rerunning various analyses, which were slowing down the app server and blocking other users from interacting with the app.

LandWeb v1.0.1

  • added axis labels and titles to several figures (#57)
  • added text to clarify meaning of “years” in simulations (#58)
  • added boxplots that summarize the vegetation cover histograms (#59)
  • added news feed and notifications (#60)
  • some performance improvements loading maps
  • fixed file-naming bug when downloadling model outputs (with #62)
  • improved caching
  • server upgrades have reduced several performance and disconnection issues

NOTE: There appears to be a bias towards deciduous species in some of the results. This may be due to the interaction between the LandMine fire model and the LANDIS Biomass-Succession models: deciduous species burn less than conifers but regenerate more quickly. I am currently taking a closer look at the simulations and code to verify these results.

LandWeb v1.0.0

  • time since fire maps, and histograms for all combinations of stand age class, leading vegetation type, and spatial area at a particular patch size;
  • vegetation cover histograms for all combinations of stand age class, leading vegetation type, and spatial area;
  • user-selectable spatial areas and patch sizes;
  • log in using your Google account to view current conditions, upload custom polygons, and download model results.

NOTE: There appears to be a bias towards deciduous species in some of the results. This may be due to the interaction between the LandMine fire model and the LANDIS Biomass-Succession models: deciduous species burn less than conifers but regenerate more quickly. We are currently taking a closer look at the simulations and code to verify these results.