Pine caterpillar (Dendrolimus) infestations threaten pine forests, causing severe ecological and economic impacts. Identifying the driving factors behind these infestations is essential for effective forest management. This study uses the APCIRD framework combined with an improved random forest model to analyze spatiotemporal changes in infestation risk and the driving effects of habitat factors in Northeast China. From 2019 to 2024, we applied SHapley Additive exPlanations (SHAP), frequency analysis, fitting functions, and GeoDetector to quantify the impact of key drivers, such as snow cover and soil, on infestation risk. The findings include (1) the APCIRD framework with the MLP-random forest model (MRF) accurately assesses infestation risks. MRF is composed of MLP and random forest. Between 2019 and 2024, areas with high infestation risk declined, shifting from higher to lower levels, with Eastern Heilongjiang and Southwest Liaoning remaining as key concern areas; (2) snow cover and soil factors are critical to infestation risk, with eight key habitat factors significantly affecting the risk. Their relationships with infestation risk follow complex, non-monotonic quartic and cubic patterns; (3) factors triggering high infestation risks are mostly at low to moderate levels. High-risk areas tend to have low to moderate elevation (<800 m), moderate to high solar radiation and temperature, gentle slopes (<30°), low to moderate evaporation, shallow snow depth (<0.02), moderate snow temperature (266.73–275), low to moderate soil moisture (0.2–0.3), moderate to high soil temperature (276.73–286.92), low to moderate rainfall, moderate wind speed, low leaf area index, high vegetation type, low vegetation cover, low population density, and low surface runoff. Interactions between factors provide a stronger explanation of infestation risk than individual factors. The APCIRD framework, combined with MRF, offers valuable insights for understanding the drivers of pine caterpillar infestations.

Forests, as one of the most vital ecosystems on Earth, play essential roles in climate regulation, water conservation, and resource provision. However, forest health is threatened by pests, among which the larch caterpillar (Dendrolimus superans) is one of the most destructive defoliators of coniferous forests in northern China. Previous studies have mostly relied on single data sources for pest detection, which are limited by insufficient spectral information or inappropriate selection of sensitive bands, making it difficult to achieve high detection accuracy. Therefore, this study integrates hyperspectral imagery from Zhuhai-1 and multispectral imagery from Sentinel-2, leveraging their high spectral resolution and broad spectral range, thus enhancing discrimination capability. Genetic algorithm (GA) was employed to select optimal features from spectral indices, texture features, and fractional-order derivatives (FOD). Random Forest (RF), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost) were compared, and model interpretability was further analyzed using Shapley additive explanations (SHAP). The results showed that XGBoost achieved the highest performance, with an overall accuracy and Kappa coefficient of 93.47% and 89.81%, demonstrating superior adaptability. Moreover, the integration of hyperspectral and multispectral data significantly improved detection accuracy compared to using either data source alone. Among the GA-selected features, Band 15 of Zhuhai-1 hyperspectral imagery exhibited strong sensitivity to pest infestation. This study provides a novel and practical approach for forest pest monitoring based on the synergistic use of hyperspectral and multispectral remote sensing data.