Remote sensing of tropical forest recovery: A review and decision-support framework for multi-sensor integration

Tropical forest recovery plays a vital role in mitigating climate change and conserving biodiversity, yet accurately monitoring its ecological success remains a persistent challenge. Common proxies such as canopy cover or greenness often overestimate recovery by conflating rapid structural regrowth with the slower processes of compositional reassembly and functional reintegration. This review synthesizes recent advances in remote sensing that enable more comprehensive tracking of tropical forest recovery across structural, compositional, and functional dimensions and at multiple spatial and temporal scales. We evaluate the capabilities and limitations of key sensor types: LiDAR for mapping canopy structure and estimating biomass; optical sensors for assessing spectral diversity and phenological variation; synthetic aperture radar (SAR) for reliable structural monitoring under all weather conditions; passive microwave sensors capture plant water content, with Vegetation Optical Depth (VOD) serving as a valuable proxy for hydrological function; and thermal sensors for tracking evapotranspiration and plant stress. Crucially, we highlight how integrating these complementary data streams (e.g., fusing LiDAR with hyperspectral or SAR with VOD) overcomes single-sensor blind spots, revealing decoupled recovery trajectories and avoiding misleading "green deserts" assessments. We identify persistent challenges, including sensor saturation, cloud cover, calibration gaps, and computational barriers in under-resourced regions. To guide ecologically robust monitoring, we present a decision-support framework that aligns sensor selection with specific recovery dimensions, spatio-temporal scales, data availability, and operational capacity. This framework emphasizes the use of defensible sensor combinations and prioritizes open-access data and feasible validation strategies. Future progress hinges on equitable calibration networks, integrating functional physiological metrics such as solar-induced fluorescence and VOD, developing interpretable machine learning tools, and expanding scalable cloud-based processing platforms. Multi-sensor integration is essential for verifying restoration outcomes to foster resilient tropical forest landscapes.