Estimating forest productivity in the eastern Siskiyou Mountains of southwestern Oregon using a satellite driven process model, 3-PGS

Abstract

The 3-PGS (physiological principles for predicting growth using satellite data) model generates monthly estimates of transpiration, photosynthesis, and net primary production (NPP), the latter derived as a fixed proportion (0.47) of gross photosynthesis. To assess the reliability of a simplified process model (3-PGS) to predict the productive capacity of coniferous forest across diverse landscapes in southwestern Oregon, we first used a geographic information system to display and manipulate basic data. This involved the following steps: (i) extrapolate monthly mean weather data to reflect topographic variation; (ii) transform monthly temperature extremes to spatial resolution of 4 ha and estimate incoming solar radiation, subfreezing days per month, daytime vapor pressure deficits, and mean temperatures; (iii) convert statewide soil survey maps into topographically adjusted estimates of soil fertility and water storage capacity (θ); and (iv) acquire satellite-derived estimates of the faction of light intercepted by vegetation during midsummer. Model predictions of soil water availability during summer months compared well with those reported from published measurements of predawn water potentials at three contrasting sites and with measurements acquired at the end of seasonal drought at 18 sites (r2 = 0.78 with mean monthly modeled drought index; r2 = 0.57 with seasonal modeled drought index). Similarly, seasonal shifts in the relative importance of various climatic and edaphic variables closely matched those defined in previously published studies. Finally, model predictions of maximum annual aboveground growth were compared with those derived from forestry yield tables based on height-age relationships with a resulting r2 of 0.76, and a standard error of 1.2 m3·ha-1·year-1 (P < 0.01).