Monitoring of phenological control on ecosystem fluxes using digital cameras and eddy covariance data

Digital repeat photography is an emerging platform for monitoring land surface phenology. Despite the great potential of digital repeat photography to yield insights into phenological cycles, relatively few studies have compared digital repeat photography to in situ measures of ecosystem fluxes. We used 60 site years of concurrent camera and eddy covariance data at 13 sites, representing five distinct ecosystem types - temperate deciduous forest, temperate coniferous forest, boreal forest, grasslands and crops - to measure and model phenological controls on carbon and water exchange with the atmosphere. Camera-derived relative greenness was strongly correlated with estimated gross primary productivity among the five ecosystem types and was moderately correlated with water fluxes. Camera-derived canopy development was also compared with phenological phase as predicted by a generalized, bioclimatic phenology model and Moderate Resolution Imaging Spectrometer (MODIS) imagery to assess the potential for cross-biome phenological monitoring. This study demonstrates the potential of webcam networks such as Phenocam (phenocam.unh.edu) to conduct long-term, continental monitoring and modeling of ecosystem response to climate change