DDSCAT.5a

DDSCAT.5a differs from previous versions in five major respects:

1. Use of the new Generalized Prime Factor Algorithm (GPFA) developed by Clive Temperton (1992) for FFT calculations. The GPFA algorithm is generally faster than the previous algorithms, yet requires no more memory than the algorithm of Brenner (1969). See §12.
2. Availability of a new algorithm for iterative solution of the system of complex linear equations. This algorithm is often faster than the algorithm of Petravic and Kuo-Petravic (1979) which was used through DDSCAT.4b (and which remains available as an option in DDSCAT.5a). See §13.
3. Automatic calculation of the transverse components of the radiative force on the target. See §14.
4. Capability to compute the electromagnetic torque on the target, due to absorption and scattering of light from the incident beam, as described by Draine and Weingartner (1996). See §14
5. Automatic calculation of the 44 Mueller scattering matrix. See §22.
6. Improved output handling, including FORTRAN unformatted binary write option as well as a netCDF interface and IDL postprocessing code.
7. Finally, this much-improved and expanded User Guide!

We also call users' attention to a minor but possibly confusing change: DDSCAT.5a uses a different convention for specifying the target axes and for certain target choices (RCTNGL, ELLIPS, ...). We think the new convention is more straightforward. See the discussion in §18 below for details.