{"title":"Integrating stepwise residual refinement and explainable AI for interpretable forest volume modeling in Hokkaido, Japan","authors":"Kotaro Iizuka , Nobuo Ishiyama , Yasutaka Nakata","doi":"10.1016/j.rsase.2025.101740","DOIUrl":null,"url":null,"abstract":"<div><div>Accurate estimation of forest stem volume is essential for effective forest resource management and carbon accounting. However, spatial heterogeneity in forest conditions often leads to systematic modeling errors, especially across ecological and operational gradients. This study proposes an integrated framework that combines XGBoost-based modeling with a novel Stepwise Residual Refinement (SRR) approach and explainable AI techniques utilizing SHapley Additive exPlanations (SHAP) to enhance both prediction accuracy and model interpretability. The framework was applied to forest inventory and remote sensing data across Hokkaido, Japan, incorporating topographic, climatic, structural, and socioeconomic variables. The initial XGBoost model achieved a root mean square error (RMSE) of 170.21 m<sup>3</sup>/ha and a percentage RMSE (%RMSE) of 36.90 %. Following the application of SRR corrections, the final model improved significantly, yielding an RMSE of 105.75 m<sup>3</sup>/ha and a %RMSE of 22.93 %. KernelSHAP analysis revealed region-specific patterns of variable influence, highlighting how environmental and human factors differentially shape forest volume across regions. SHAP-derived zoning delineated clusters of forest quality that aligned with workforce presence and ecological conditions, particularly in conifer-dominated areas. These results demonstrate the importance of integrating explainable AI and spatial refinement to uncover nuanced forest dynamics and support adaptive, data-driven forest management. This study highlights how interpretable machine learning can simultaneously improve predictive accuracy and reveal latent socio-ecological processes that drive forest conditions.</div></div>","PeriodicalId":53227,"journal":{"name":"Remote Sensing Applications-Society and Environment","volume":"40 ","pages":"Article 101740"},"PeriodicalIF":4.5000,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Remote Sensing Applications-Society and Environment","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352938525002939","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
Accurate estimation of forest stem volume is essential for effective forest resource management and carbon accounting. However, spatial heterogeneity in forest conditions often leads to systematic modeling errors, especially across ecological and operational gradients. This study proposes an integrated framework that combines XGBoost-based modeling with a novel Stepwise Residual Refinement (SRR) approach and explainable AI techniques utilizing SHapley Additive exPlanations (SHAP) to enhance both prediction accuracy and model interpretability. The framework was applied to forest inventory and remote sensing data across Hokkaido, Japan, incorporating topographic, climatic, structural, and socioeconomic variables. The initial XGBoost model achieved a root mean square error (RMSE) of 170.21 m3/ha and a percentage RMSE (%RMSE) of 36.90 %. Following the application of SRR corrections, the final model improved significantly, yielding an RMSE of 105.75 m3/ha and a %RMSE of 22.93 %. KernelSHAP analysis revealed region-specific patterns of variable influence, highlighting how environmental and human factors differentially shape forest volume across regions. SHAP-derived zoning delineated clusters of forest quality that aligned with workforce presence and ecological conditions, particularly in conifer-dominated areas. These results demonstrate the importance of integrating explainable AI and spatial refinement to uncover nuanced forest dynamics and support adaptive, data-driven forest management. This study highlights how interpretable machine learning can simultaneously improve predictive accuracy and reveal latent socio-ecological processes that drive forest conditions.
期刊介绍:
The journal ''Remote Sensing Applications: Society and Environment'' (RSASE) focuses on remote sensing studies that address specific topics with an emphasis on environmental and societal issues - regional / local studies with global significance. Subjects are encouraged to have an interdisciplinary approach and include, but are not limited by: " -Global and climate change studies addressing the impact of increasing concentrations of greenhouse gases, CO2 emission, carbon balance and carbon mitigation, energy system on social and environmental systems -Ecological and environmental issues including biodiversity, ecosystem dynamics, land degradation, atmospheric and water pollution, urban footprint, ecosystem management and natural hazards (e.g. earthquakes, typhoons, floods, landslides) -Natural resource studies including land-use in general, biomass estimation, forests, agricultural land, plantation, soils, coral reefs, wetland and water resources -Agriculture, food production systems and food security outcomes -Socio-economic issues including urban systems, urban growth, public health, epidemics, land-use transition and land use conflicts -Oceanography and coastal zone studies, including sea level rise projections, coastlines changes and the ocean-land interface -Regional challenges for remote sensing application techniques, monitoring and analysis, such as cloud screening and atmospheric correction for tropical regions -Interdisciplinary studies combining remote sensing, household survey data, field measurements and models to address environmental, societal and sustainability issues -Quantitative and qualitative analysis that documents the impact of using remote sensing studies in social, political, environmental or economic systems