A grassroots journal collects, assess, categorises and ranks all published articles, reports and preprints relevant for a specific topic/community. In this case the homogenisation community. It does not publish the articles themselves, but makes up to date open reviews.
This journal also has a page for further community activities, such as lists with upcoming conferences, homogenisation algorithms and validation datasets.
All current assessments are for articles about validation and uncertainty quantification. This way the page on this category can show the value of an article list that starts with the most important articles for each type of validation study as an entry into the literature, mimicking the function of a supervisor.
The assessment of the validation study by Stefanie Gubler et al. based on the high-quality Swiss network is complete with two independent reviews.
The example assessment of my own HOME benchmarking study using simulated data illustrates the general principles and the added value that it is now easy for an author to provide additional information. Authors naturally cannot write grading assessments.
The validation study by Zeke Hausfather et al. based on a reference network has one real and two example assessments, which partially disagree and show how the editors resolve this in their synthesis.
The article Break detection of annual Swiss temperature series in under review with one real review and two example reviews.
Join this experiment
If you are curious how this is organised you can read the page on the tasks of authors, reviewers and editors. For more general design considerations and ideas on how multiple journals could help and strengthen each other see my blog on grassroots scientific publishing.