Journal of Quantitative Spectroscopy & Radiative Transfer 106 (2007) 417–436 Comparison between discrete dipole implementations and exact techniques Antti Penttila¨ a,Ã , Evgenij Zubko a,b,c , Kari Lumme a , Karri Muinonen a , Maxim A. Yurkin d,e , Bruce Draine f , Jussi Rahola g , Alfons G. Hoekstra d , Yuri Shkuratov b a Observatory, University of Helsinki, P.O. box 14, FI-00014 University of Helsinki, Finland b Astronomical Institute of Kharkov National University, 35 Sumskaya Street, Kharkov 61022, Ukraine c Institute of Low Temperature Science, Hokkaido University, Kita-ku North 19 West 8, Sapporo 060-0819, Japan d Faculty of Science, Section Computational Science of the University of Amsterdam, Kruislaan 403, 1098 SJ, Amsterdam, The Netherlands e Institute of Chemical Kinetics and Combustion, Siberian Branch of the Russian Academy of Sciences, Institutskaya Str. 3, 630090 Novosibirsk, Russia f Department of Astrophysical Sciences, Princeton University, 108 Peyton Hall, Princeton, NJ 08544-1001, USA g Simulintu Oy, Espoo, Finland Abstract The use of the discrete dipole approximation (DDA) method in wave optical scattering simulations is growing quite fast. This is due to the fact that the current computing resources allow to apply DDA to sufficiently large scattering systems. The advantage of DDA is that it is applicable to arbitrary particle shape and configuration of particles. There are several computer implementations of the DDA method, and in this article we will compare four of such implementations in terms of their accuracy, speed and usability. The accuracy is studied by comparing the DDA results against results from either Mie, T-matrix or cluster T-Matrix codes with suitable geometries. It is found that the relative accuracy for intensity is between 2% and 6% for ice and silicate type refractive indices and the absolute accuracy for linear polarization ratio is roughly from 1% to 3%. r 2007 Elsevier Ltd. All rights reserved. Keywords: Scattering; Discrete-dipole approximation 1. Introduction The importance of light scattering methods in studying the structure of various remotely sensed objects, e.g. in astronomy and in some technological applications has greatly increased in the last years. One obvious reason for this is undoubtedly the enormous and steady increase in the computer power and speed. The information content from the light scattering studies is very large because at its best all the 16 Mueller matrix ARTICLE IN PRESS www.elsevier.com/locate/jqsrt 0022-4073/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jqsrt.2007.01.026 Ã Corresponding author. Fax: +358 9 19122952. E-mail address: Antti.I.Penttila@helsinki.fi (A. Penttila¨).