Investigation of the sensitivity of the land‐surface parameterization of the NCAR Community Climate Model in regions of tundra vegetation

Open Access
Publisher Website
Google Scholar
Abstract
A series of sensitivity experiments has been conducted using a version of the NCAR Community Climate Model (CCM) and a complementary zero‐dimensional land‐surface model in order to invesligate the sensitivity of the models to surface process parameterizations in northern high latitudes. The study was motivated by anomalously high surface temperatures at these latitudes in a control simulation with the CCM. The effects of perturbing the maximum vegetation cover, distribution of the roots within the active soil layer, the depth of the active soil layer, soil albedo, and of maintaining a fully saturated upper and active soil layer were explored. Little response occurred in either model when the depth of the soil or the root distribution was altered. Increasing the percentage of ground covered by vegetation increased the canopy air temperature but generally reduced soil temperatures in the zero‐dimensional model. However, in the CCM integration the surface air temperature increased over most of the region for this change. This response seems to be due to the high level of precipitation during the three‐dimensional integration and is similar to one of the zero‐dimensional computations in which a large rainfall event was prescribed on the second day of the 10 day integration. Maintaining the moisture contents of both soil layers at their capacity caused a decrease in soil temperatures at most locations in the CCM and in the zero‐dimensional model once the soil in the control began to dry out. Increasing the soil albedo also produced soil and surface air cooling in both models. All the integrations underlined the considerable sensitivity of the land‐surface parameterization scheme to the instantaneous and preceding soil moisture conditions. Reduction in surface temperatures achieved was significantly less than the original discrepancy between CCM temperatures and observations. Part of the discrepancy could be caused by unrepresentative observational data and a model bias toward high levels of net surface radiation. In addition, preliminary calculations suggest that explicit inclusion of permafrost may be necessary before a fully satisfactory representation of the tundra land‐surface climate can be achieved. This conclusion may be important for CO2‐induced climate‐change studies.