Transparency and Openness Promotion (TOP)
The overarching policy of the Journal of Fluid Mechanics is that research articles should contain sufficient information to allow others to understand, replicate and verify findings, and compare them with alternative studies. We therefore require that whenever possible articles demonstrate:
Understanding - Articles should be written and will be assessed for clarity, both of the execution of the research and for its outcomes and conclusions.
Replication - All information required to replicate the study must be provided, either within the body of the paper or in publicly accessible repositories. Examples of what is required include but are not limited by:
- for analytical studies, the mathematically complete set of equations and boundary conditions, any theorems relied upon, appropriately referenced;
- for numerical studies, the mathematically complete set of equations and boundary conditions, sufficient descriptions of the algorithms or packages used to solve them, appropriately referenced, and the resolution used with respect to the independent variables;
- for laboratory experiments, the dimensions and construction of any apparatus, the materials used including their relevant physical properties, the protocol adopted for the running of the experiments, the measurement tools used including their resolution and accuracy, including appropriate calibration standards;
- for field studies, the raw data collected or used, any protocols or tools used to access the data (e.g. data-mining tools) or to process it.
Verification - Most studies can be verified or falsified provided that sufficient detail is given for them to be replicated (see above). Where data is manipulated (for example, bringing together multiple data sets by scaling) the raw (dimensional) data relating to the primary measurements (laboratory) or outputs (numerical) should be provided together with the protocols or tools used to process them.
Comparison - All graphical information should be supplemented with numerical data or precise algorithms to reproduce it. For example, data points should be provided in a spreadsheet and curves should be defined either explicitly with an equation or as resulting from a precisely defined algorithm.