Abstract
Abstract. Grasslands provide many important ecosystem services globally, and projecting grassland productivity in the coming decades will provide valuable information to land managers. Productivity models can be well calibrated at local scales but generally have some maximum spatial scale in which they perform well. Here we evaluate a grassland productivity model to find the optimal spatial scale for parameterization and thus for subsequently applying it in future productivity projections for North America. We also evaluated the model on new vegetation types to ascertain its potential generality. We find the model most suitable when incorporating only grasslands, as opposed to also including agriculture and shrublands, and only in the Great Plains and eastern temperate forest ecoregions of North America. The model was not well suited to grasslands in North American deserts or northwest forest ecoregions. It also performed poorly in agriculture vegetation, likely due to management activities, and shrubland vegetation, likely because the model lacks representation of deep water pools. This work allows us to perform long-term projections in areas where model performance has been verified, with gaps filled in by future modeling efforts.
Highlights
Grassland systems span nearly 30 % of the global land surface (Adams et al, 1990) and play a prominent role in terrestrial carbon cycles (Parton et al, 2012)
We evaluate the model across different combinations of North American ecoregions and vegetation types to find an optimal spatial scale in which to parameterize and apply the model
We evaluated each of the 11 models using the Nash– Sutcliffe coefficient of efficiency (NSE; Eq 2), as well as the mean coefficient of variation of the mean absolute er
Summary
Grassland systems span nearly 30 % of the global land surface (Adams et al, 1990) and play a prominent role in terrestrial carbon cycles (Parton et al, 2012). Annual productivity of grasslands in central and western North America is driven in large part by precipitation (Sala et al, 2012). Even with consistent shifts in climate, different locations can experience different changes in productivity due to local-scale responses (Zhang et al, 2011; Sala et al, 2012; Knapp et al, 2017). This highlights the need for models which can be resolved at small spatial and temporal scales, making long-term grassland productivity projections as informative as possible
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
