Reproducibility Hub

7 published findings, reproduced as inspectable folia.yaml recipes and validated against independent references. Each card is the recipe, its paper, and its validation record.

Alpine Treeline Ecotone detection (Wei et al. 2020)
Wei et al. 2020 — Spatial detection of alpine treeline ecotones in the Western US
✓ validatedhuman
regionWestern US and Canadaperiod2013-2022dataLandsat 8 L2 + 3DEP DEMops7resultspatial correlation 98.1%forks0
Climate vulnerability model for Malawi (Malcomb et al. 2014)
Vulnerability modeling for sub-Saharan Africa using indicators of adaptive capacity, livelihood sensitivity, and physical exposure
○ unverifiedhuman
regionMalawiperiod2010-2010dataDHS + FEWS NET + drought/flood exposureops5forks0
MapBiomas LULC, fire, water — Rondonia (2023)
Reproduce MapBiomas products for a ~50km tile in Ji-Paraná, Rondônia, Brazil using Folia's browser-native WASM pipeline. Validated against MapBiomas Collection 10, JRC Global Surface Water, and FIRMS active fire data.
✓ validatedhuman
regionJi-Parana, Rondonia, Brazilperiod2023-2023dataLandsat C2 L2 + AlphaEarth embeddingsops4resultlulc overall accuracy 88.9%forks0
LandTrendr temporal segmentation — PNW (Kennedy et al. 2010)
Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr — Temporal Segmentation Algorithms
○ unverified
regionPacific Northwest, USAperiod2000-2024dataLandsat time seriesops2forks0
Solar-on-forest trajectory similarity — Massachusetts
Click a documented forest-to-solar reference site and rank every Massachusetts pixel by how closely its 2019-2024 AlphaEarth Foundations embedding trajectory resembles the reference. All compute runs in the browser via @folia/compute (WASM with JS fallback). No server.
○ unverifiedsynthetic inputs
regionMassachusetts, USAdataAlphaEarth Foundations v1 (2019-2024)ops2forks0
LandTrendr burn detection — 2021 Bootleg Fire (Kennedy et al. 2010)
Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr — Temporal Segmentation Algorithms
✓ validatedagent
regionBootleg Fire scar, Lake and Klamath counties, Oregon, USAperiod2013-2026dataLandsat C2 L2 annual NBR, 84 scenes via Microsoft Planetary Computerops1resultpeak_to_trough disturbed recall producers accuracy 77.7%forks0
Embedding-analog mineral prospectivity — Mberengwa belt, Zimbabwe
Does a place's 64-D AlphaEarth embedding resemble the embeddings at a belt's known-mineralized sites more than it resembles random country rock? What is graded is the held-out skill of that analog on this belt, measured leave-one-out in the same run that draws the map. That skill is a BAND, AUC 0.92 to 0.96 across background draw and sampling grid, not a single figure, and it is Mberengwa-specific: the reference script reads 0.9608 and the engine recipe a visitor can re-run today reads 0.9253. The site read itself is a resemblance diagnostic with no ground truth, never a detection.
✓ validatedagent
regionMberengwa-Belingwe greenstone belt, Zimbabweperiod2024dataAlphaEarth Embedding Fields v1 (2024) + 85 confirmed gold occurrences (of 140 mapped)ops5forks0