Version 5 of the CENTURY model consists of monthly and daily time-step versions, which we call Century5 and DayCent5, respectively. Both share a large portion of their algorithms. DayCent5 includes maintanence respiration, daily soil hydrology and soil temperature submodels, and trace gas submodels.
Century5 and DayCent5 represent a substantial change in the technical aspects of the model implementation, along with some additional algorithms, and the ability to integrate into other models and the IRC gridded model, and run many simulations concurrently using the C5MPI. While most of the core algorithms of the model have not changed, the organization of the code, the programming language, and many of the file formats are new.
CENTURY 4 source code has been translated from Fortran 77 to ANSI standard C++. During the process of translation, the code was reorganized into C++ classes. Where possible, data management code and many algorithms were reorganized to take advantage of the object-oriented architecture available with C++. Most of these classes are mostly self-contained units that can be reused elsewhere, either as-is or extended for additional functionality using the inheritance capabilities of C++.
In addition to reorganization of the code, many of the files that CENTURY uses were reorganized using the netCDF file format and function library. These files are platform-independent; that is, they can be moved onto many different computer systems from various manufacturors with various operating systems, and be read by any netCDF software on that platform which wants to access those files.
We verified the C++ version by checking that it produced essentially the same results as the Fortran version, using a suite of CENTURY 4 site parameter sets and associated CENTURY 4 schedule files. The results of version 5 were compared against the output of version 4. Accounting for modifications, bug fixes, and enhancements in version 5, the results were very similar.
Since we did fix bugs, added new features to the simulation code, and reorganized many parts of the code, the results you will get from version 5 should not match exactly your version 4 results, and in some cases, may be quite different. Significant effects can result from, among the many modifications, the correct incorporation of runoff into the water model, and the addition of a layered physical soil structure.