R.P. Neilson, S. Running, D. Schimel, D. Ojima, W. Parton, J. Melillo, W.M. Post, A.W. King, T. Kittel, I.C. Prentice, M. Sykes, D. Bachelet, P. Thornton, VEMAP participants. 2001. IHDP/WCRP/IGBP Global Change Open Science Conference, Amsterdam, the Netherlands 10-13 July 2001.

Abstract

 A recent hypothesis suggests there may be non-linear carbon feedbacks between the terrestrial biosphere and the atmosphere.  A relatively small amount of warming might enhance carbon sequestration due to increased productivity under slightly higher temperatures, increased precipitation and a physiological benefit from elevated CO2.  However, with additional warming, the terrestrial biosphere could become a source of carbon to the atmosphere, possibly due to saturation of the CO2 effect, an exponential increase in evaporative demand with increased temperatures, increases in plant respiration, shifting vegetation distribution and increased vegetation disturbance from drought and fire.
 This non-linear hypothesis was examined through a sensitivity analysis using six terrestrial biosphere models from the Vegetation/Ecosystem Modeling and Analysis Project (VEMAP) under two global warming scenarios over the conterminous U.S.  A Canadian scenario was near the high end of potential temperature increases (ca. 5.5o C, CGCM1), while a Hadley scenario was near the lower end of possible increases (ca. 2.8o C, HADCM2SUL).  Both scenarios simulated increases in precipitation of about 22% over the U.S.  Each scenario begins with a timeseries of observed weather from 1895 to the present using observed increases in CO2, then continues to 2100 using a 1% per year increase in CO2, as in the IPCC IS92a emissions scenario.
 The VEMAP models consist of four BioGeochemical Cycling models (BGC), which calculate productivity and nutrient cycles, but rely on a static vegetation map and either lack disturbance or have prescribed fire intervals.  Two additional models are Dynamic General Vegetation Models (DGVM), which combine the BGC processes with both dynamic vegetation distribution and fire simulation.  Under the above hypothesis, the six models should simulate carbon sequestration throughout the entire modestly-warming Hadley scenario.  However, under the warmer Canadian scenario, the simulated vegetation might initially show carbon sequestration, but with increasing temperatures could become a carbon source.  Diagnostics were used to examine model differences in terms of four separate processes; 1) elevated CO2 and carbon fixation, 2) water limitation, 3) nitrogen limitation, and 4) disturbance processes, specifically drought and fire.
 All six models showed increased carbon sequestration throughout the 21st century under the Hadley scenario.  However, under the hotter Canadian scenario, the DGVMs appeared to support the non-linear hypothesis, showing initial carbon gains followed by later carbon losses.  By contrast, the four BGC models showed carbon gains throughout the Canadian scenario.  Different sensitivities among the models appeared to be related to diagnostic differences and will be discussed.  Thus, the DGVM results suggest there could be a threshold temperature below which terrestrial biota are a carbon sink, slowing global warming, but above which the biota could become a source, accelerating global warming.

HOME PAGE   | CV MAIN PAGE  | CV  PUBLICATION INDEX  |ECOLOGICAL RESEARCH OVERVIEW  |  CLIMATE RESEARCH OVERVIEW