The overall goals of this thesis are to examine quantitatively the controls that climate has on natural emissions of N2O and CH4 from the terrestrial biosphere to the atmosphere and to explore the feedbacks between climate and the N2O and CH4 cycles. A process-oriented global model for soil N2O emissions and a more empirically based global model for wetland CH4 emissions have been developed to address these goals. These emission models are capable of quantifying the natural emission changes due to climate change and the feedback of the natural emissions onto the climate system.

The global emission model for N2O, which focuses on soil biogenic N2O emissions, has a 2.5 degree x 2.5 degree spatial resolution. The model can predict daily emissions for N2O, N2, NH3 and CO2 and daily soil uptake of CH4. It is a process-oriented biogeochemical model including all those soil C and N dynamic processes for decomposition, nitrification, and denitrification in Li et al.'s (1992a, b) site model. The model takes into account the spatial and temporal variability of the driving variables, which include vegetation type, total soil organic carbon, soil texture, and climate paramters. Climatic influences, particularly temperature and precipitation, determine dynamic soil temperature and moisture profiles and shifts of aerobic-anaerobic conditions.

The methane emission model is developed specifically for wetlands and has a spatial resolution of 1 degree x 1 degree. There are three components for the global wetland methane emission model: high latitude wetlands, tropical wetlands and wet tundra. For high latitude wetlands (i.e. northern bogs), the emission model uses a two-layer hydrological model (Frolking, 1993) to predict the water table level and the bog soil temperature, which are then used in an empirical formula to predict methane emissions. For tropical wetlands (i.e. swamps and alluvial formations), a two-factor model (temperature and water availability) is used to model the methane flux by taking into account the temperature and moisture dependence of activity of methanogens. Methane emissions from wet tundra are calculated by assuming a constant small methane flux and an emission season defined by the time period when the surface temperature is about the freeezing point. The hydrological model and the two-factor model are driven by surface temperature and precipitation, which links methane emission with climate.

For present-day climate and soil data sets the N2O emission model predicts an annual flux of 11.3 Tg-N/year (17.8 Tg N2O/year). The spatial distribution and seasonal variation of the modeled current N2O emissions are similar to climate patterns, especially the precipitation pattern. Chemical transport model experiments using the modeled soil N2O emissions plus prescribed other (minor) emissions show good agreement with observations of trends of surface N2O missing ratios and the N2O interhemispheric gradient (Prinn et al., 1990). Sensitivity experiments suggest that soil organic carbon content, precipitation and surface temperature are the dominant factors in controlling global N2O emissions.

The global CH4 emission model predicts an annual flux of 127 Tg CH4/year for present-day climate and wetland conditions, which is in the middle fo the range of recent estimates for natural wetland emissions (Bartlett and Harriss, 1993; Reeburgh et al., 1993; IPCC, 1994). Global methane emissions have two strong latitudinal bands with one in the tropics and the other in the northern high latitudes. There are strong seasonal cycles for the high latitude CH4 emissions and hence for the global total emission amount.

The emission models for N2O and CH4 have been applied to two extreme climatic cases: that associated with doubling current CO2 levels and that during the last glacial maximum. While predicted equilibrium climates from three climate models (MIT 2D, GISS and GFDL GCMs) have been used in both cases, predicted soil organic carbon from terrestrial ecocystem model (TEM, Melillo et al., 1993) have been used in the "doubled-CO2" case and CLIMAP data (1981) have been used in the "ice age" case. Results indicate that equilibrium climate changes due to doubling CO2 would lead to a 34% increase in N2O emissions and a 54% increase in natural wetland CH4 emissions. Temperature increases seem to dominate the contribution to increases in N2O and CH4 emissions. Geographical coherence of predicted changes in surface temperature and precipitation is significant in determining the predicted changes in global emissions. Ice age soil N2O emissions and wetland CH4 emissions are predicted to be significantly smaller (about 50% of current emissions).

Finally, the emission models were coupled with 2D climate and chemistry models developed at MIT (Sokolov and Stone, 1995; Wang, Prinn and Sokolov, 1996). Model results indicate that changes in natural N2O and CH4 emissions corresponding to long term climate changes are significant. Predicted N2O and CH4 emissions indicate significant sensitivity to outputs from the climate (surface temperature and precipitation) and TEM (total soil organic carbon) models. Fully interactive runs show that there is a significant positive feedback between emissions and climate.