MétaCan
Menu
Back to cohort
Record W4415699064 · doi:10.1016/j.eiar.2025.108245

Integrating participatory GIS, remote sensing, and explainable machine learning to assess forest provisioning services

2025· article· en· W4415699064 on OpenAlex
Kamaldeen Mohammed, Daniel Kpienbaareh, Rachel Bezner Kerr, Jinfei Wang, Isaac Luginaah, Esther Lupafya, Laifolo Dakishoni, Mwapi Mkandawire

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.
fundA Canadian funder is recorded on the work.

Bibliographic record

VenueEnvironmental Impact Assessment Review · 2025
Typearticle
Languageen
FieldEnvironmental Science
TopicRemote Sensing in Agriculture
Canadian institutionsWestern University
FundersSocial Sciences and Humanities Research CouncilCornell Atkinson Center for Sustainability, Cornell UniversityDavid R. Atkinson Center for a Sustainable Future , Cornell UniversitySocial Sciences and Humanities Research Council of CanadaInternational Development Research CentreMcKnight Foundation
KeywordsProvisioningLivelihoodRandom forestGeospatial analysisVegetation (pathology)Participatory sensingForest inventoryCitizen journalismIndex (typography)

Abstract

fetched live from OpenAlex

Forests play a vital role in supporting rural livelihoods by providing essential resources such as food, fuelwood, and medicine. Ensuring the sustainable utilization of these resources while balancing environmental protection requires a data-driven approach that integrates advanced technologies and local knowledge to inform forest management. This study synthesizes data from Participatory Geographic Information System (PGIS) of 66 forest plots and 1864 trees, multisource remote sensing (i.e., radar and optical) and explainable machine learning to assess forest provisioning supply for community forests management. Key findings from the inventory include the multifunctional roles of trees for medicinal, food and culinary uses. Vegetation Indices such as Transformed Soil Adjusted Vegetation Index (TSAVI) and Normalized Difference Index 45 (NDI45) were identified as useful predictors of forest provisioning supply, capturing essential attributes of vegetation dynamics using random forest (R 2 = 0.76, RMSE = 4.51). Radar-derived texture metrics were equally relevant and can be especially useful under challenging climatic conditions, such as persistent cloud covers. The Shapley Additive Explanations (SHAP) and Partial Dependence Plots (PDP) revealed threshold relationships between Sentinel-2 indices and forest provisioning, with notable thresholds observed at NDI45 = 0.3 and TSAVI = 0.59. These thresholds signal possible ecological tipping points associated with forest health and productivity. Also, Independent Conditional Expectations (ICE) and Locally Interpretable Model-agnostic Explanations (LIME) provided site-specific explanations on the association between remote sensing indices and forests provisioning capacity, underscoring the spatial heterogeneity of forest ecosystems. The study fills an important research gap by providing a framework that integrates interpretable and explainable modelling with participatory geospatial methods, aiming to inform community-based forests management and support data-driven landscape and site-specific forest ecosystem management in the Miombo woodlands. • A participatory geospatial and explainable machine learning framework for assessing forest services. • Participatory GIS is used to identify the multifunctional role of forests/trees in agrarian communities. • Radar and optical remote sensing metrics play complementary roles in predicting forest provisioning services. • Global and local model interpretations can inform both landscape-scale and site-specific forest management.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.487
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0000.001
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.016
GPT teacher head0.322
Teacher spread0.306 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it