HR: 0800h
AN: A51A-0744    [Abstracts]
TI: Modeling daily and seasonal relationships between air and ground temperatures using a time-dependent thermal diffusivity in the shallow subsurface
AU: * Pollack, H N
EM: hpollack@umich.edu
AF: Geological Sciences Department University of Michigan, 2534 C. C. Little, Ann Arbor, MI 48109-1063 United States
AU: Smerdon, J E
EM: jsmerdon@umich.edu
AF: Applied Physics Program University of Michigan, 2534 C. C. Little, Ann Arbor, MI 48109-1063 United States
AU: van Keken, P E
EM: keken@umich.edu
AF: Geological Sciences Department University of Michigan, 2534 C. C. Little, Ann Arbor, MI 48109-1063 United States
AB: Subsurface temperatures have been inverted to reconstruct temperature histories at the ground surface on centennial time scales. These ground surface temperature (GST) histories in turn have been used to estimate SAT changes at times prior to instrumental records, assuming close coupling between SAT and GST at long time scales. This assumption has been the subject of some debate because ground temperatures in the upper few meters of the subsurface are influenced not only by SAT, but also by insulating effects of winter snow cover, latent heat effects associated with winter ground freezing and spring thawing, and summer evapotranspiration. Here we describe a simple way of accounting for these effects in terms of a time-dependent thermal diffusivity in the upper meter of the subsurface. The thermal diffusivity, defined as the ratio of the thermal conductivity to the volumetric heat capacity, is parameterized in terms of daily SAT, precipitation, and snow cover. The insulating effect of snow cover is equivalent to a reduction of the thermal conductivity, whereas latent heat effects are equivalent to an increase of the volumetric heat capacity; both lead to a reduction of the thermal diffusivity. We model subsurface temperatures by driving the subsurface with the SAT as the surface boundary condition, along with a variable diffusivity in the upper meter of the subsurface that changes according to daily precipitation, snow cover and SAT. We illustrate the method by comparing observed and calculated temperatures, using observational data from Fargo, North Dakota for 1981-82 and 1982-83, two very different meteorological years at this site. The time-dependent diffusivity model accurately simulates the observed temperatures in both of the years investigated, and performs significantly better than a model in which a single diffusivity is used for all depths and at all times. Because the time-dependent diffusivity model is parameterized in terms of archived meteorological variables, it shows promise as a means of investigating relationships between GST and SAT at long periods, the timescale relevant to the question of how well a GST history represents the SAT history.
DE: 3322 Land/atmosphere interactions
DE: 3337 Numerical modeling and data assimilation
DE: 3344 Paleoclimatology
DE: 1620 Climate dynamics (3309)
DE: 1645 Solid Earth
SC: Atmospheric Sciences [A]
MN: 2004 AGU Fall Meeting