Fifteen years of observations from a very tall tower on the responses of landscape-scale productivity to climate anomalies and extremes

Significant advances have been made over the past decades in capabilities to simulate diurnal and season variation of leaf-level and canopy-scale photosythnesis in temperate and boreal forests. However, long-term prediction of future landscape-scale productivity in a changing climate and especially to climate extremes may be more dependent on how climate and biological factors influence interannual and decadal variability in ecosystem carbon exchanges. Until recently, multiple long-term (10+ year) high temporal frequency (daily) observations of canopy carbon exchange were not available to reliably assess this claim. An analysis of one of the longest running North American eddy covariance flux towers (16+ years) reveals that single climate variables do not adequately explain carbon exchange anomalies beyond the seasonal timescale. Lagged time-frequency domain analysis through empirical mode decomposition and Granger causality statistics of anomalies and covariance of moisture, temperature and net ecosystem exchange at the WLEF Park Falls tall tower site reveals both lagged effects of moisture deficits and carbon storage on productivity, especially in periods of major climate anomalies, such as droughts. These results lead toward a new conceptual framework for improving earth system models with long-term flux tower observations.