 |
|
 |
||||||
|
||||||
 |
Home
| Mission
|
about SciDAC
|
Contact Us |
|||||
 |
 |
 |
||||
|
Alumni ProjectModeling Dynamic Vegetation for Decadal to Century Climate Change StudiesAndrew D. Friend, Columbia UniversitySummaryVegetation is an important component of the global climate system through its control of energy partitioning over substantial portions of the land surface, but is only simply treated in most global climate models (GCMs). The aim of this project is to develop, evaluate, utilize, and make available a model of vegetation dynamics to improve the ability of GCMs to make predictions of future climate change. The model is being developed by combining a treatment of carbon and nitrogen fluxes, vegetation structural dynamics, and competition with the existing land surface model of the NASA Goddard Institute for Space Studies GCM. Climate simulations with global climate models (GCMs) typically employ a fixed vegetation map with simple invariant physical properties. Disadvantages of this approach are that vegetation structure, physiology and distribution do not respond to climate changes on short or long time scales, and surface fluxes of water and energy can be overestimated or underestimated compared to reality. Therefore this project aims to develop, evaluate, utilize, and make available a model of global vegetation dynamics that can be embedded in GCMs to improve the realism of climate simulations. Vegetation is an important component of the global climate system through its control of energy fluxes over substantial portions of the land surface. In turn, vegetation distribution, structure, and physiological state are largely determined by climate. Therefore the vegetation-climate feedback needs to be included in climate change studies. For example, increased leaf area tends to increase absorption of sunlight, thereby increasing atmospheric instability and precipitation in tropical and sub-tropical regions, resulting in a positive feedback. Deforestation can cause a corresponding negative feedback. Shorter-term physiological responses also result in important feedbacks. For example, the width of stomatal pores on leaf surfaces is a major determinant of the rate of evaporation of soil water, but these pores are themselves strongly influenced by soil and atmospheric humidity, as well as atmospheric CO2 (thereby affecting the response of the hydrological cycle and climate to changed CO2). In addition, vegetation plays a major role in the global carbon cycle. The natural relationship between vegetation and climate is being increasingly disturbed by human land use and inadvertent effects such as fertilization by increasing atmospheric CO2. The model of vegetation dynamics being developed in this project describes the responses of vegetation physiology, structure, and distribution to climate, atmospheric CO2, nitrogen deposition, and land use change on timescales of minutes to centuries. The model also treats the effects of natural disturbances such as fire and includes a treatment of soil carbon and nitrogen dynamics. The model is being developed by combining the existing land surface model of the Goddard Institute for Space Studies (GISS) GCM with a process-based treatment of carbon and nitrogen fluxes, structural dynamics (such as leaf area and rooting depth changes), and competition. The model is being written as a separate module from the GCM, capable of running outside the GCM or being coupled to another climate model. A major portion of the work so far has been the development and parameterization of the canopy physiology component of the vegetation dynamics model from site-level data on water, carbon, and energy fluxes. This component of the vegetation dynamics model is now fully coupled to the GISS GCM. Significant improvements in simulated climate occur as a result of more realistic responses of evaporation to vegetation type, atmospheric humidity, light, and temperature. For example, increased evaporation over the Amazon rainforest reduces mean surface air temperature by at least 1 °C where the model was previously too warm, and reduced evaporation causes summertime eastern US temperatures to be increased by up to 2 °C where the model was previously too cold. The effect of the new parameterization on the response of climate to increased CO2 is also significant, with surface temperatures up to 2 °C higher in many vegetated regions in response to the closure of stomatal pores, reducing evaporation and increasing the amount of energy as heat in the atmosphere (Figure 1). Global CO2 uptake by vegetation through photosynthesis within the GCM is increased by 48% for doubled CO2 conditions, with influences from both climate change and CO2 fertilization. The effects of changes in leaf area, timing of leaf display in response to temperature and soil water, and vegetation distribution will be considered next by including the structural dynamics component of the vegetation model within the GCM. This model is currently being extensively tested off-line.
The GISS GCM is being parallelized in Fortran 90 for use on new powerful computing platforms such as a Compaq Alpha Server SC45 with 96 processors, available to the project. In addition, the code is being modularized to enable easier modification and assessment of alternative parameterizations, as well as use of modules by other groups. For further information on this subject contact:Dr. Andrew D. FriendColumbia University 2880 Broadway New York, NY 10025 Phone: 212-678-5587 afriend@giss.nasa.gov
|
||||
Home | ASCR | Contact Us | DOE disclaimer |
|
|
