Automatic Triggering of Scientific Models For Avalanche Research

The Swiss Federal Institute for Snow and Avalanche Research in Davos (Switzerland) is in charge of issuing avalanche forecasting bulletins for the whole Swiss Alps. In order to do so, several automatic weather stations networks (~180 stations for Switzerland) collect data that are then automatically sent to a snow pack simulation code, Snowpack. This code is also at the foundation of the Alpine3D model of alpine surface processes. Within the “Swiss Experiment” project (data acquisition and sharing infrastructure between various Swiss institutions) as well as within the AAA/Switch grid computing inititiative, several issues are in need of improvements. Among them, a new architecture allowing simulations to be automatically triggered by the availability of new data has to be defined and implemented.

This internship is about transforming the structure of the code so that a runaway simulation can be detected and stopped, without preventing the timely delivery of simulated data to the avalanche warning service. Development is in C on Linux. The code written during this internship will be used in a mission critical environement (24/7, on supercomputers for simulations lasting hundreds of hours). The candidate might have to travel to Davos a few times.

Goal:      Practical experience in developing large scale scientific applications.

Benefits

• Becoming familiar with scientific computing.
• Practical experience on large scale scientific application.
• Gaining experience in mission critical and high performance computing.
• Working in an interactive and interdisciplinary research environment.

Prerequisites

• beginner to intermediate level experience in using Linux.
• Good C skills.
• Good software engineering skills.
• Being interested in showing leadership in a multidisciplinary environment.

Additional Information

• Second project contact:  Mathias Bavay (bavay-at-slf.ch)