<p dir="ltr">Forest ecosystems play a pivotal role in climate change mitigation. Sustainable forest management practices necessitate accurate quantification of forest aboveground carbon stocks (FACS). In the first part of this study, I compared the 13-year changes in FACS across three silvicultural systems, including even-aged management (EA), uneven-aged management (UEA), and non-harvested controls (NH), in Indiana's hardwood forests. Forest stands within each silvicultural system were assigned with one of the six treatment types, including clearcutting, shelterwood, or prescribed burning for EA, single-tree selection or patch cutting for UEA, or untreated controls. From 2008 to 2021, the FACS of the study area exhibited an increase from 91.5 ± 9.0 Mg/ha to 115.3 ± 2.1 Mg/ha. Single-tree selection, shelterwood, and prescribed burning were found to have minimal impacts on FACS. However, clearcutting and patch cutting resulted in a significant reduction in FACS, with subsequent recovery reaching only 30-37% of their pre-treatment levels after 13 years. Further investigations may use long-term inventory data to analyze the chronic recovery patterns on these sites.</p><p dir="ltr">In the second part of this study, I evaluated the feasibility of using 3DEP LiDAR in conjunction with the random forest algorithm for multiscale FACS prediction. It was found that the stand-scale model outperformed the plot-scale model, primarily due to a stand’s higher positioning accuracy and reduced boundary effects than the plot-scale model. This led to a reduction in RMSE from 25.43 Mg/ha (26%) to 16.74 Mg/ha (20%). Moreover, the stand-scale model exhibited robust landscape-level prediction performance even in scenarios where point density decreased from 7.7 points/m² to 2.0 points/m². However, the partitioned model including solely clearcut and patch sites produced a higher RMSE of 59% (17.82 Mg/ha) due to inaccurate LiDAR return classification and biased canopy height metrics extraction. Future research should delve into the mechanisms of point cloud classification to improve the FACS prediction accuracy for clearcut forest monitoring.</p><p dir="ltr">Overall, this thesis contributed to a deeper understanding of carbon dynamics in managed hardwood forests, highlighted the potential of using LiDAR technology for improved landscape-level carbon monitoring, and informed the decision-making processes in the context of climate change mitigation.</p><p><br></p>