Hydrologic processes in global climate change
Article first published online: 19 OCT 2006
©1991. American Geophysical Union. All Rights Reserved.
Eos, Transactions American Geophysical Union
Volume 72, Issue 10, page 114, 5 March 1991
How to Cite
1991), Hydrologic processes in global climate change, Eos Trans. AGU, 72(10), 114–114, doi:10.1029/90EO00095.(
- Issue published online: 19 OCT 2006
- Article first published online: 19 OCT 2006
Popular concerns over global resource issues have focused attention on the description of the dynamics of hydrologic processes at scales much larger than the traditional river basin. Improved, remotely sensed land surface data on a global scale and improved large-scale global circulation models (GCMs) have fueled the need to better integrate hydrologic processes with land surface features and atmospheric dynamics. Until recently, however, the interaction between climate modelers and hydrologists has been weak.
Given a prescribed geographic distribution of sea surface temperature, GCM models can approximately simulate long-term, large area mean precipitation, runoff, and other hydrologic variables. Runoff is the amount of precipitation leaving an area in surface drainage in contrast to that which infiltrates into the subsurface. Some hydrologists have been involved in so-called “effects” modeling, in which they attempt to use GCM output to assess the sensitivity of hydrologic systems to global warming. These efforts have made hydrologists painfully aware of the incompatibility of temporal and spatial scales between the GCMs and basin-scale hydrologic models in common use. GCM model simulations with and without mountains have been used to demonstrate the role of the mountains in controlling soil wetness in the North American mid-continent “Manabe, 1990”. Improvements in spatial resolution of the land surface does affect model performance, particularly in the influence of orographic processes and the scale-dependence of the representation of mountain-induced gravity wave drag. Simulation of runoff at the river basin scale using a GCM “pot model” (a lumped representation of hydrologic processes) has been compared with simple nonlinear soil moisture models. The results suggest that runoff predicted by the GCM has statistical characteristics much more similar to the rainfall input of the model itself than it does to actual observed runoff. Incorporation of a relatively simple representation of the soil moisture dependence of infiltration, along with a simple representation of base flow, leads to much better agreement with the statistical character of simulated runoff. Hydrologic characterization of large areas will improve as spatially distributed soil moisture modeling continues to evolve and remotely sensed data becomes available from the Earth Observing System [Wood, 1990].