Skip to content

Fire VASE Developmental Morphology

Fire VASE is the population-scale fire-history workflow used for the current developmental morphology manuscript. It treats each FIRED event as a real observed life history, converts daily growth into a comparable VASE profile, and then attaches climate as an explanatory layer rather than as part of the shape definition.

The workflow has two linked products:

  • a durable analysis database under ignored local roots such as scratch/fire_vase_run_full/
  • rendered views of that database, including atlases, manuscript figures, PDF drafts, and Google Docs-ready DOCX files

Rendered figures are not the source of truth. They are reproducible views of the VASE tables, feature tables, climate-attribution tables, manifests, and figure-generation scripts.

Current Scope

The manuscript-scale run uses real FIRED-derived daily fire histories from 2000-2021. The current population contains 278,569 events and 626,102 daily VASE slices. Complete daily centroid gridMET climate is available for 237,235 fires.

The centroid climate table includes:

  • maximum and minimum temperature
  • vapor pressure deficit
  • wind speed and a simple wind-presence flag
  • precipitation
  • maximum and minimum relative humidity
  • specific humidity
  • 100-hour and 1000-hour dead fuel moisture
  • energy release component and burning index
  • reference and potential evapotranspiration
  • solar radiation

The perimeter exposure workflow is a companion product. It summarizes climate over active daily burned polygons, cumulative burned area, and configurable exterior perimeter-extension buffers. It is deliberately stored separately from the centroid baseline so figures can compare exposure definitions without silently changing earlier analyses.

Analysis Stages

Stage Main script Main products
Cache climate years scripts/cache_gridmet_years.py cached gridMET NetCDF files and gridmet_cache_manifest.json
Build centroid VASE tables scripts/fire_vase_build_climate_tables.py fire_catalog.parquet, fire_traits.parquet, vase_slices.parquet
Build perimeter exposure table scripts/fire_vase_build_perimeter_climate_tables.py vase_climate_exposures.parquet plus a build report
Build developmental morphospace scripts/fire_vase_developmental_morphology_analysis.py feature tables, medoids, shape categories, atlas PDF
Audit claims and null framing scripts/fire_vase_manuscript_claim_audit.py claim audit, reviewer notes, revised manuscript drafts
Generate climate-revision figures and manuscript scripts/fire_vase_climate_revision.py five main figures, supplementary figures, Science-style PDF, manuscript Markdown
Generate Google Docs-ready manuscript scripts/build_fire_vase_google_docs_docx.py output/docx/fire_vase_climate_revision_google_docs.docx

Rebuild Commands

Run the comprehensive gridMET cache first when the optional climate variables are needed:

python scripts/cache_gridmet_years.py --preset comprehensive --keep-going

Build or refresh the centroid VASE climate table:

python scripts/fire_vase_build_climate_tables.py \
  --config config/fire_vase_pipeline.yml \
  --table-root scratch/fire_vase_run_full/tables \
  --report scratch/fire_vase_run_full/climate_build_comprehensive_report.json

Build the perimeter and perimeter-extension exposure companion table:

python scripts/fire_vase_build_perimeter_climate_tables.py \
  --include-optional-variables \
  --table-root scratch/fire_vase_run_full/tables \
  --report scratch/fire_vase_run_full/perimeter_climate_build_comprehensive_report.json

Build the population morphology atlas:

python scripts/fire_vase_developmental_morphology_analysis.py \
  --table-root scratch/fire_vase_run_full/tables \
  --data-output-dir scratch/fire_vase_developmental_morphology \
  --output output/pdf/fire_vase_developmental_morphology_atlas.pdf

Build the climate-revision figures and Science-style manuscript PDF:

python scripts/fire_vase_climate_revision.py

Build the Google Docs-ready manuscript:

python scripts/build_fire_vase_google_docs_docx.py

What the Current Figures Show

The current manuscript figures are organized as a story:

  • Figure 1 shows that similar final burned areas can arise through different life histories.
  • Figure 2 places representative VASE shapes over the morphospace so readers can see how developmental forms change across VASE coordinates.
  • Figure 3 colors the morphospace by climate variables and compares climate groups, emphasizing that climate shifts the prevalence of developmental forms.
  • Figure 4 tests whether climate effects depend on current developmental state using leakage-safe next-day-growth baselines.
  • Figure 5 shows matched examples where similar climate can lead to different forms and similar forms can arise under different climate pathways.

The figure legends live in figures/climate_revision_main/figure_legends.md. They are written in Science style: the figure number and title are part of the legend, not the image artwork, and each legend explains how to read the figure and what scientific message it supports.

Validation and Audits

The manuscript workflow includes explicit checks that keep the story bounded:

  • reviewer-style critique rounds in docs/manuscripts/fire_vase_developmental_morphology/
  • a Science author-guideline compliance note
  • a citation audit and final citation check
  • render checks for PDF and DOCX outputs
  • repository-size checks to keep bulk products out of git

The current final citation check is docs/manuscripts/fire_vase_developmental_morphology/final_citation_check_2026-07-23.md.

Current Limits

The main manuscript analysis still treats centroid daily gridMET exposure as the complete-population baseline. The perimeter and perimeter-extension climate table is available as a richer companion product, but its coverage and role must be reported separately when used in figures or claims.

The workflow does not yet include topography, vegetation, suppression activity, ignition cause, wind direction, wind gusts, or local-normal climate anomalies as complete-population covariates. Those are the next data-integration steps.

The current prediction values are linear baseline results from blocked validation. They are useful for bounding deterministic climate explanation, but they are not optimized predictive models and should not be interpreted as causal estimates.