• Depositing the data, and any new code or models, in a public repository with a persistent identifier.
• Providing supplementary files such as raw measurements, input files, scripts or extended results.

Experimental and Materials Research

• Materials and chemicals, with supplier, grade or purity.
• Specimen preparation, dimensions and conditioning.
• Equipment and instruments, with manufacturer, model and key settings.
• Calibration procedures and reference standards used.
• Test standards followed, for example ASTM, ISO or SNI, with any deviations explained.
• Number of specimens or repeat tests, and how results were averaged.
• Measurement uncertainty or variability, such as standard deviation or error bars, and how it was calculated.

Numerical Modelling and Simulation

• Software name and version, or the source of any in-house code.
• Governing equations, constitutive models and key assumptions.
• Geometry, boundary and initial conditions, and material parameters.
• Mesh or grid details and evidence of mesh or time-step convergence.
• Solver settings and convergence criteria.
• Validation against experimental, field or benchmark data.

Field, Geotechnical and Earth-Surface Studies

• Location of sites or survey points, with coordinates and datum where this can be disclosed.
• Dates and duration of observations.
• Instruments, sampling methods and sampling intervals.
• Data processing steps, corrections and filtering.
• Sources of secondary data such as maps, remote-sensing imagery or monitoring networks.

Machine Learning and Data-Driven Methods

• Data sources, size, and pre-processing steps.
• How data were split into training, validation and test sets, and how leakage between them was prevented.
• Model architecture, hyperparameters and how they were selected.
• Random seeds, hardware and software environment where results depend on them.
• Performance metrics, and comparison with appropriate baselines.

Industrial Case Studies and Surveys

• How the case, site or respondents were selected.
• The period and conditions under which data were collected.
• The questionnaire or interview guide, which can go in supplementary material.
• How confidential information was anonymised.

In these cases the authors must explain the restriction in the Data Availability Statement. Where possible, they should share what they can, such as aggregated or normalised values, a representative subset, or synthetic data with the same properties, and name a route through which qualified researchers can request access.

The editors may ask to see the data confidentially during review to confirm that the results are supported.

For code, a GitHub or GitLab repository is useful for development, but a versioned, archived copy with a DOI, for example via Zenodo, should be cited in the article.

Authors must keep the original data, laboratory notebooks, instrument output files, unprocessed images and code on which the article is based for at least 5 years after publication, or longer if required by their institution or funder.

If a reader is unable to obtain data described as available, or raises credible doubts about whether the results can be reproduced, the editors will contact the authors and ask them to address the concern.

Depending on the outcome, the journal may publish a correction to the Data Availability Statement or the methods, an expression of concern or, if the findings are shown to be unreliable, a retraction.