Carbon Fluxes in a Managed Landscape: Assessing the Drivers of Temporal and Spatial Variability in Flux Tower, MODIS and Forest Inventory Data of the Pacific Northwest

This research focuses on the Wind Late Successional Reserve of Southern Washington where clear-cut logging over the past 100 years has created a fragmented landscape of coniferous forests that range in age from 0 to 500 years. In this study, we integrate several datasets to examine the environmental drivers of carbon exchange in this region across time and space. These sources include: (1) network of flux towers across a disturbance choronosequence, (2) MODIS Enhanced Vegetation Index, (3) aboveground net primary production (ANPP) from forest inventories, (4) and regional precipitation and air temperature measurements from the NOAA network of weather stations and PRISM reanalysis data. Net ecosystem exchange of carbon (NEE) has been measured at the Wind River Canopy Crane AmeriFlux site since 1998. The canopy crane is located in an old-growth forest composed of late seral Douglas-fir (Pseudotsuga menziesii) and western hemlock (Tsuga heterophylla). Two flux towers were erected in early seral stands to study the effects of silviculture on net ecosystem exchange. CO2 uptake at the old-growth stand is highest in the spring before bud break when air and soil temperatures and vapor pressure deficit are relatively low, and soil moisture and light levels are favorable for photosynthesis, while maximum CO2 uptake is observed two to three months later at the early seral stands and coincide with peak leaf area index. This CO2 pattern is driven by different water conserving strategies. A reduction in carbon exchange is observed at the old-growth forest when moisture becomes limiting and canopy conductance rates drop sharply after mid-morning in the summer. In contrast, inhibition in canopy conductance rates and CO2 exchange is not observed at the early seral stands until soil moisture levels become critically low at the very end of the summer. The regional MODIS data (200 km X 200 km area) from 2000-2008 show that annual variability in the Enhanced Vegetation Index (EVI) also can be linked to precipitation and temperature anomalies at the stand level and across the region. Regional EVI anomalies are strongly negatively correlated with the annual precipitation and air temperature anomalies once the MODIS pixels are carefully examined with regards to forest age. EVI data from the tower-centered pixel also correlate well with annual NEE at the AmeriFlux site and show promise for scaling sparse flux tower observations, even over old-growth forests. Lastly, permanent plots have been continuously measured in the old-growth stand since 1947 and provide long-term data on tree demographics, recruitment, growth and mortality, and show evidence of decadal variability in response to precipitation and air temperature anomalies, as well as to disturbance (e.g., a Douglas-fir beetle kill in the 1950’s). We take advantage of the overlapping measurement period 1998-2004 and compare ANPP from the forest inventories to the flux tower estimates of NEE and MODIS EVI with focus on the regional environmental drivers.