Research Catalog: CGD's Terrestrial Sciences

	Figure 1. Comparison for 12 river basins of two estimates of change in water storage derived from the GRACE satellite mission with that simulated by CLM3 and CLM3 with hydrologic modifications. High resolution figure.
	Figure 2. Differences in precipitation rate (mm/day) between a simulation with increased soil wetness (MOIST) and the control (CTRL) for October-December (left) and January-March (right). High resolution figure.
	Figure 3. Difference in 2-m air temperature for the period March-May between climate model simulations with the new snow cover parameterization and the control. Top: prescribed vegetation. Bottom: Dynamic vegetation. High resolution figure.

CLM hydrology

Keith Oleson, David Lawrence, and Gordon Bonan led a project within the CCSM land model working group to improve the surface hydrology of CLM. Numerous university colleagues contributed to this project including, but not limited to, Robert Dickinson (Georgia Institute of Technology), Zong-Liang Yang and Guo-Yue Niu (University of Texas), Reto Stockli (Colorado State University), and Peter Lawrence (University of Colorado). A basic deficiency in version 3 of CLM is the partitioning of latent heat flux into evaporation of water intercepted by plant canopies, transpiration, and soil evaporation. The model has too much interception and soil evaporation and too little transpiration, in part due to dry soils. New parameterizations of hydrologic processes in CLM greatly improve the simulated hydrologic cycle. This improvement is seen in comparisons for several river basins of simulated soil water storage to that derived from the GRACE satellite. (Figure 1.)

Ben Cook (University of Virginia) collaborated with Sam Levis and Gordon Bonan to quantify hydrologic feedbacks using CAM3 and CLM3. One set of simulations investigated the effects of increased soil moisture on wet season (October through March) precipitation in southern Africa. Increased soil wetness results in decreased precipitation over the region of perturbed soil moisture compared to a control simulation. The increased soil moisture alters the surface energy balance, resulting in a shift from sensible to latent heating. The shift from sensible to latent heating cools the surface, causing a higher surface pressure, a reduced boundary layer height, and a positive vertical gradient in equivalent potential temperature – indicative of increased atmospheric stability. The surface changes induce anomalous surface divergence and increased subsidence. This causes a reduction in cloud cover and specific humidity above 700 hPa and results in a net decrease of column integrated precipitable water. (Figure 2.)

In a second project, Cook, Levis, and Bonan examined interactions between snow and vegetation in the Arctic. Climate simulations using CAM3 and CLM3 were performed with prescribed vegetation cover and with vegetation simulated by CLM3’s dynamic vegetation model. In the control simulation, snow cover fraction changes gradually with snow depth; an alternative, yet equally plausible parameterization, results in greater snow cover fraction for a given snow depth. In simulations where the vegetation was prescribed, the choice of snow cover parameterization results in a very limited response of the model. With dynamic vegetation, however, the change is much more dramatic, where slight initial increases in snow cover fraction with the new parameterization lead to large scale retreat of boreal vegetation, widespread cooling, and persistent snow cover over much of the boreal region during the boreal summer. (Figure 3.)

Additional Information

Print version (.pdf)

CGD Sectional Research

CAS | CCR | CDP | CMS | CSEG | OS | TSS