The effect of climate variability factors on potential net primary productivity uncertainty: An analysis with a stochastic spatial 3-PG model

Yearly climate fluctuations introduce variability in forest productivity that impacts final yield. Over the decade-to-century timescale, climate change is likely to have similar effects. Therefore, yield forecasts based on average climatic values are expected to provide a biased snapshot of potential carbon fixation when using process-based models. To bridge this gap, we propose a spatially explicit framework to estimate the potential net primary productivity (NPPpot) as well as its expected uncertainty. We relied on the Physiological Principles in Predicting Growth (3-PG) model that was modified to include a stochastic climate generator based on observed climatic values. We called this framework the Stochastic Spatial 3-PG (3-PGS2) model. The 3-PGS2 model is a set of functions programmed in the R software, tailored to estimate the mean, standard deviation, and confidence intervals of NPPpot. This framework is able to identify and quantify areas more susceptible to climate uncertainty as well as allowing for sensitivity analysis. For illustration, we estimate the NPPpot for loblolly pine in the southeastern U.S. Our predictions for average NPPpot coincide with average values published in the literature. Most importantly, we provide a spatial assessment of the areas with the largest NPPpot uncertainty, showing that areas with lower productivity tend to have higher relative variation and those areas with higher productivity have lower relative variation.

Duke Scholars

Author Bin Mei Environmental Sciences and Policy