Using Remote Sensing and Process-based Growth Modeling to Predict Forest Productivity Across Western Oregon

Graduation date: 2016Accurate measurement of forest productivity is fundamental to understand the carbon balance of forested ecosystems. Recent changes in climate highlight the importance of developing methods to measure forest productivity so that sound economic and environmental decision can be made. Efforts to measure forest productivity across the landscape using remote sensing suffer from limitations inherent in individual sensors. We combine approaches to estimate forest productivity on Douglas-fir plantations using optical passive remote sensing to estimate stand ages, single-pass LiDAR sampling to estimate structural properties, and a process-based growth model (3PG) to assess variation in soil properties across western Oregon. Our results indicate that it is possible to estimate both site index and difficult-to-obtain soil properties across the landscape using two types of remote sensing instruments in combination with modeling. While we believe that the approach is sound, we found that the quality of extrapolated climatic data very much affects the derived values of soil properties assessed by modeling. In contrast, estimates of maximum leaf area index with LiDAR proved a reliable independent means to assess site productivity because plantations reach maximum values in less than two decades after establishment