SHARM is a new 1-D radiative transfer (RT) code designed to compute monochromatic radiance/flux in the shortwave spectral region over a Lambertian or anisotropic surface. The atmospheric properties can change arbitrarily in the vertical dimension. The algorithm uses the method of spherical harmonics (MSH) [Muldashev et al., 1998; Lyapustin and Muldashev, 1999, 2000] with the boundary conditions in the form of Marshak [1947]. The code is rigorous in a sense that its solution converges to the true value at increase of the order of MSH. Two numerical "tricks" implemented in Sharm-1D make this code very fast. The first one uses special symmetry properties of the matrix of MSH system to reduce its size by a factor of two for the subsequent singular value decomposition (SVD), which was originally suggested for MSH by Karp et al. [1980]. The second one is an innovative "correction function" method of angular smoothing of solution developed by Dr. Muldashev. This method yields a high accuracy at relatively low orders of MSH [Muldashev et al., 1998] and is faster than the "source function method" [Dave and Armstrong, 1974]. A recently performed intercomparison among several 1-D codes [Lyapustin, 2002] showed that in the test cases Sharm-1D was as accurate as DISORT [Stamnes et al., 1988], yet faster in calculations with aerosol phase functions. On the other hand, the new code has yet to accumulate a long history of testing in the most "adverse" conditions that DISORT has. Therefore, it is advisable to the users to validate the results of Sharm-1D calculations against DISORT in the most difficult cases, at least initially.


Below you'll find documentation and software for SHARM. To download, Windows users right-click and Mac users control-click for a contextual menu and choose the appropriate selection to save to your hard drive.