![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
The RainCloud project: Harnessing Cloud Computing for a meteorological application at the Tyrolean Avalanche Service |
VerfasserIn |
Felix Schueller, Simon Ostermann, Matthias Janetschek, Radu Prodan, Georg Mayr |
Konferenz |
EGU General Assembly 2013
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 15 (2013) |
Datensatznummer |
250080786
|
|
|
|
Zusammenfassung |
Precipitation in mountainous regions is an essential process in meteorological research
for its strong impact on the hydrological cycle and its effects on avalanche danger. Our
meteorological aim is to investigate and simulate precipitation in mountainous regions
with a simple numerical model. This model, due to its mathematical nature, is called the
"`linear model of orographic precipitation"' (short: LM). As it is a very simple and basic
model, it can be easily run in a large number of iterations for parameter studies, i.e.
the same setup with variations in certain input parameters.
We are using the LM in 3 different "`flavours"' within the same ASKALON workflow: an
operational type, and two research types with varying idealization of the meteorological
setup.semi-idealized and an ideal type. The operational type comprises the repeated and
regular run of a certain version of the workflow providing the Tyrolean
Avalanche Service ("`Lawinenwarndienst Tirol"') with a spatially detailed (500 m
resolution) probabilistic precipitation forecast to help them forecast avalanche danger.
Using Cloud Computing allows us to be more flexible and cheaper than dedicated hardware.
Our workflow invocations can differ substantially in requirements to the computational
infrastructure, both in computing power and data size.
We present the design of a meteorological application for the Tyrolean Avalanche Service
using the ASKALON environment comprising graphical workflow modelling and execution in a
Cloud computing environment. We demonstrate with the Amazon EC2 Cloud that within the
limits of Amdahl's law our workflow can gain important speedup when executed in a
visualized Cloud environment with important operational cost reductions.
Results from the research type workflow implementations show the usefulness of our model
for determining precipitation distribution in the case of two field campaigns over Norway. |
|
|
|
|
|