Advanced remote sensing to quantify temperate peatland capacity for belowground carbon capture

Temperate peatlands typically dominated by grasses and sedges generate among the greatest annual rates of net primary productivity (NPP, up to 4 kg C m^-2) and soil carbon storage (up to 1 kg C m^-2) for natural ecosystems. Belowground tissues represent 20-80% of total NPP, thus understanding the controls on belowground NPP (BNPP) in these wetland ecosystems is particularly important to quantifying peatland carbon balances. In addition, there is a growing need to quantify large-scale belowground C sequestration rates in wetlands to better understand marsh resilience to sea level rise and to help define eligibility for carbon offset credits. Since plant productivity influences wetland C budgets, combining field and remote sensing techniques for estimating above and belowground productivity of wetland vegetation over a large spatial extent will help to address these needs. We are working in a USGS long-term experimental wetland restoration site on drained peatland in the Sacramento-San Joaquin River Delta. Using the spatial variability in water depth and residence time across the 7 ha wetland, our goal is to develop practical methods to quantify and map BNPP of emergent marsh vegetation from remotely sensed estimates of aboveground plant characteristics and aboveground NPP. Field data collected on wetland plants hardstem bulrush (Schoenoplectus acutus) and cattail (Typha spp.) were coupled with reflectance data from a field spectrometer (range 350-2500 nm) every two to three weeks during the summer of 2011. We are analyzing reflectance data to develop hyperspectral indices that predict the biophysical characteristics of wetland vegetation - biomass, leaf area index (LAI), and the fraction of absorbed photosynthetically active radiation (f_APAR) - which may be used to infer belowground biomass and productivity. Soil cores and root in-growth bags were used to calculate root biomass and productivity rates. Existing allometric relationships were used to calculate aboveground biomass and green LAI. Digital photos taken above and below canopy were also analyzed to calculate canopy gap fraction and LAI, and were validated with allometric LAI measurements. We calculated f_APAR with measurements from line and point quantum sensors. Remote sensing-based estimates of these parameters may be used to model NPP, which is determined at the site through CO₂ flux measurements at the leaf scale, eddy correlation flux at the field scale, and stem and leaf turnover measurements. The field spectrometer data were registered to new high resolution (1.8 meter), 8-band Digital Globe World View-2 satellite imagery with a unique red-edge band, which was used to create detailed vegetation maps. This frequent-repeat satellite imagery should provide the capability to extrapolate measurements of plant productivity to other wetland sites using indices that are field-tested and calibrated for wetland plant species that contribute significantly to belowground carbon storage.