In this paper we summarise a survey report on computational models for snow avalanche motion that was developed within the frame-work of the EU research project SAME (Snow Avalanche Modelling, Mapping and Warning in Europe).
An examination of existing models shows that: (1) there is not - and probably never will be - a single model that adequately describes all avalanche types; (2) in order to account for the extraordinary variability of avalanche motion in response to initial and boundary conditions, flow-regime transitions and the snow mass balance should be properly described in future models; (3) calibration and validation of these models will require a comprehensive measurement programme; (4) determination of realistic initial conditions is a serious problem. We suggest that using simple models to scan the relevant parameter space with more advanced models for detailed simulations of selected scenarios could improve this situation.
Finally, we discuss the needs for, and benefits of, a co-ordinated programme of aval-anche research. The main features of the SAME proposal for an extensive joint experi-mental programme are described. We suggest that international collaboration could produce high-quality models covering all essential practical needs. Increased interdiscipli-nary collaboration would be advantageous for model development and facilitate incor-poration of other scientific disciplines.