Search

Abstract: It is difficult to map forest growth potential across regions with different environmental conditions from limited field measurements of productivity. Climatically sensitive biophysical models and remotely sensed data on vegetation offer the potential to overcome this problem. We combine both approaches in this paper. We calibrated the midsummer value of NASA''s MODIS instrument''s enhanced vegetation index (EVI) against site indices (SI) mapped at 10 widely dispersed locations for Douglas-fir or ponderosa pine, ranging in height from 16 to 48m at 50 years (age at breast height). Median values of EVI derived from a 3km×3km grid centered on commercial forest lands of known productive capacities produced a linear regression with site indices (R 2 =0.83). We matched stand growth properties generated by a physiologically based stand growth model (3-PG) with site-specific yield tables and inferred from model predictions, as well as from the literature, that a close relation exists between maximum leaf area index (max L), maximum periodic annual increment (max PAI), and SI. We tested the ability of median EVI to predict SI values derived from height and tree age measurements made at 5263 federal inventory and analysis (FIA) survey plots in Oregon with comparable success (R 2 =0.53) to that derived from previous application of 3-PG using 1km resolution of climate and soil data. Based on the general agreement between the two approaches, we used mid-summer EVI valves to generate a 1km resolution map predicting spatial variation in SI of Douglas-fir over 630,000km2 in the Pacific and Inland Northwest, U.S.A. [Copyright &y& Elsevier]