HELIOGate, a Portal for the Heliophysics Community - Science Gateways for Distributed Computing Infrastructures - page 191

propagation models. These workflows are also used to find the list of events to be

analyzed by statistical propagation models.
VOTable parser workflows offer parsing capabilities for VO tables. They are
used in combination with the metadata query workflow when the capabilities of the
HELIO services cannot meet the

filtering and sorting requirements. Helper work-
flows offer simple value manipulation; they are mainly used to modify the values
extracted from the catalogues to adapt them to be used by the propagation models.
Commonly they create ranges from single values or collect lists of results for
parametric propagation models. Propagator workflows that execute the propagation
model with the given parameters.
If a user wants to develop his/her own modi ed propagation model in a non-

native workflow language such as TAVERNA, it is possible to remotely invoke
external non-native workflows through the SHIWA interoperability platform.
This approach has been tested to investigate high-speed coronal mass ejections
(Pierantoni 2013) that required an extension of the model.

14.4 HELIOGate Portlets

Depending on the type of users, HELIOGate offers different usage modalities to the
propagation models. For users who do not want to deal directly with the workflows,
HELIOGate offers an Advanced Propagation Model portlet that executes a com-
plete statistical assisted and validated parametric propagation Model. The advanced
propagation model portlet is based on different simpli cations (a single catalogue

can be used to find events, only a single catalogue can be used to validate the

results, and the only parameter set as range is the speed), but it is useful as a general
tool for a first-draft analysis of multiple phenomena. It is also useful in under-

standing the concepts behind the different models. The user is led through the steps
shown in Fig. 14.2 .
The

first steps of assisted propagation models where
catalogues of events are queried. The user selects a time range that he wants to
investigate and an event catalogue that has to be queried. The advanced propagation
model portlet then returns a list of all the events that were spotted on the surface of
the Sun in that period, and the propagation model will be executed for all the events
that happened in that period.
In the second step, the user de

first step is the same as the

nes the parameter values that were not de ned by

the first step and starts the execution. This step happens in a parametric propagation

model, where the model is executed for a range of values leading as multiple
outputs the Expected Times of Arrivals at a variety of targets for each of the set of
parameters within the ranges.
As the execution of a parametric model results in a set of results and not just one,
the last step of the advanced propagation model portlet is the same as a validated
propagation model where the results of the model are checked against catalogues of