TARGO is a plugin for targeted, object-based diagnostics in the ICON atmospheric model. It is implemented as a COMIN plugin and enables feature-based sampling during model runtime, as well as an offline replay mode for development and debugging. It was developed at the TROPOS institute. 

Object-based studies provide valuable insights into cloud life cycles in atmospheric circulation models. Cloud tracking tools are therefore widely applied but are often limited by data availability and temporal resolution, ranging from minutes to several hours depending on the model application. To address these limitations, TARGO (Targeted Output) was developed, an approach that enables cloud detection during the atmospheric model runtime.

TARGO is implemented as a plugin for the ICON model and coupled via the ComIn interface. It employs the TOBAC tracking algorithm for cloud detection and temporal linking at every model time step and allows the targeted output of arbitrary model variables at cloud positions. Alongside cloud number, lifetime, and location, the targeted output enables the analysis of cloud-centred variables such as vertical profiles of mixing ratios, temperature, and wind velocity.
The capabilities of TARGO are illustrated using s set of realistic limited-area simulations of deep convective development, highlighting the benefits of cloud and object tracking on the model time-step level for analysing cloud evolution and sensitivities in ICON.

The software is open-source (see license).

If you want to get started, the software is available on GitLab with documentation and test cases.