| In this dissertation I develop a mathematical model (RMBLMM) to investigate the response of the soil in a sub-alpine meadow to experimental warming. This work is part of an ongoing project at the Rocky Mountain Biological Laboratory (RMBL), in which we are experimentally warming meadow plots with overhead infrared radiators. I use RMBLMM in conjunction with field observations to explore the mechanisms underlying soil response to the experimental warming. I calibrate and validate the model with extensive, independent time-series data.; The model simulates fluxes of energy and water (liquid and vapor) within the soil and exchange with the atmosphere. Data from a nearby weather station drive the model. Vaporization of soil water and vapor transport within the soil are included in the model; these processes are usually ignored in land surface process models. I found that this vaporization scheme captured the soil temperature response to warming better than either an aerodynamic resistance or a Penman-Monteith combination equation for bare-soil surface evaporation. I also compared several other, mostly novel, schemes in RMBLMM with more-common approaches. In most cases my modifications were improvements, such as my formulation of downward infrared radiation and cloud cover effects, and my inclusion of rocks in the determination of bulk density, water content, and precipitation received.; I calibrated RMBLMM by comparing the control simulation with data from unheated field plots. Calibration and sensitivity analyses revealed that the most important factors governing the system's response to warming were hydrologic variables, evaporation, and bottom boundary conditions. The initial soil moisture and temperature had little influence on the model simulation. I tested the model's ability to predict response to the experimental warming against observations of treatment-control differences in soil temperature and moisture. The bare-soil model captured the temperature response reasonably well, but the moisture effect only partially. The failure of the model may be reasonably explained in part by the exclusion of plant processes, but not entirely. It is therefore possible that current evaporation formulations, including the vaporization scheme used here, are not well able to capture the dependence of temperature response on moisture differences resulting from experimental heating. |
