make these comparisons based on an equilibrium future and do not consider transi- 

 tion pliases leading to that steady state. 



Study findings indicate that regression-based models such as the Osnabruck model 

 have serious drawbacks when used to predict changes in net primary productivity 

 associated with changes in carbon dioxide and climate. For example, important 

 feedbacks between temperature and ecosystem processes cannot be fully expressed. 

 The reliability of regression-based models decreases as any analysis goes beyond 

 the range of the data used in calibrating the regressions. Clearly, a potential of global 

 change is to produce new combinations of environmental variables for some ecosys- 

 tems. 



Study findings further indicate that process-based models need to consider the 

 linkages between climate and site fertility. While elevated carbon dioxide could, in 

 some cases, increase net primary productivity, the availability of nitrogen varies within 

 forest type and is currently limiting to increased net productivity in eastern forests. 

 Feedbacks between the carbon cycle and the nitrogen cycle, in response to changes 

 in temperature and moisture, will alter the productivity response to elevated carbon 

 dioxide. Because of the complexity of interactions among processes that are affected 

 by changes in carbon dioxide and climate, modeling the response of net primary 

 productivity requires the use of process-based models that integrate ecosystem 

 functions. 



Analyses of resource supply and demand have traditionally assumed an unchanging 

 climate. Ecological analyses examining a changed climate have not extended their 

 results to the social and economic systems. Growth and yield research has tradition- 

 ally had an economic basis of planning harvest and calculating the returns from 

 investments in timber management. A link needs to be made between ecological 

 models describing the underlying processes in forested systems and economic 

 models describing timber demand at the regional and national scales. We initiated a 

 study with the objective of linking the Terrestrial Ecosystem Model to the Aggregate 

 Assessment System and the Timber Assessment Market Model to examine the 

 sensitivity of the forest sector to changes in net productivity resulting from changes in 

 climate. We emphasize that transient climate change impacts remain unmodeled and 

 may override results of the current study. 



To investigate the potential implications of changes in carbon dioxide and climate for 

 U.S. temperate forests, we ran the Terrestrial Ecosystems Model with climate output 

 from four existing global change models. 1/ These climate models predict equilibrium 

 climate change that corresponds to an atmospheric carbon dioxide concentration of 

 approximately 625 parts per million, which is projected to occur by 2065. 



1/The simulations used were the Goddard Institute of Space Studies (GIBS) Global Change Model; the 

 Oregon State University (OSU) Global Change Model, and two from the Geophysical Fluid Dynamics 

 Laboratory (GFDL 1 and GFDL 2). 



21 



