This web page contains links to documentation, source code, and input data for the stand-alone version release of CLM3.0. Links to the CAM3.0 (Community Atmosphere Model) stand-alone version release and coupled CCSM3.0 (Community Climate System Model) release can be found at the bottom of the page.
CLM3.0 is radically different from either CLM2.1 or CLM2.0, particularly from a software engineering perspective. The major difference is that CLM3.0 is now both scalar cache friendly as well as vector friendly and can perform well on both types of architectures. The subgrid hierarchy introduced in CLM2.1 has been maintained in CLM3.0. However, its implementation within the code has been completely modified in order to permit the code to behave acceptably on vector architectures such as the Earth Simulator or the Cray X1. The CLM3.0 code is in fact easier to read as well as being more accessible to the introduction of new parameterizations than either CLM2.1 or CLM2.0.
Several other software-related changes have also accompanied the new CLM3.0 data structures. A brief discussion of the new data structures as well as the resulting new decomposition algorithm is given in the CLM3.0 User's Guide. In addition, a completely new decomposition algorithm has been implemented for multi-tasked and/or multi-threaded model runs that result in a significant improvement in both the load balancing as well as scaling nature of the model. New interfaces have been implemented for the specification of history file fields and initial dataset fields. Furthermore, history output for RTM river output is now on the RTM grid rather than being interpolated to the model grid. We refer the reader to the CLM3.0 Developer's Guide for more comprehensive details of the coding implementations.
CLM3.0 contains several improvements to biogeophysical parameterizations to correct deficiencies in the coupled model climate using CLM2.1. In CLM2.1, the 2-m temperature frequently dropped below the atmospheric potential temperature during daytime heating in certain regions. Stability terms were added to the formulation for 2-m air temperature to correct this. In CLM2.1, there is a discontinuity in the equation that relates the bulk density of newly fallen snow to atmospheric temperature. The equation was modified to correct this problem. Aerodynamic resistance for heat/moisture transfer from ground to canopy does not vary with the density of the canopy in CLM2.1. This leads to high surface soil temperatures in regions with sparse canopies. A new formulation was implemented in CLM3.0 that provides for variable aerodynamic resistance with canopy density. The vertical distribution of lake layers was modified to allow for more accurate computation of ground heat flux. A fix was implemented for negative round-off level soil ice caused by sublimation. Competition between plant functional types (PFTs) for water, in which all PFTs share a single soil column, is the default mode of operation in CLM3.0. CLM2.1 accepts either rain or snow from the atmospheric model. If the precipitation is snow, a formulation based on atmospheric temperature determines the fraction of precipitation that is in liquid form. In CLM3.0, the atmospheric model (in cam and ccsm mode) delivers precipitation explicitly in liquid and/or solid form. In offline mode (uncoupled from an atmospheric model), the formulation based on atmospheric temperature is still used. A fix was implemented to correct roughness lengths for non-vegetated areas.