Data and Reproducibility Policy

Scientific conclusions must be supported by traceable data and sufficiently detailed descriptions of methods. This policy establishes requirements for materials needed to verify research results, their deposit, citation and preservation, and procedures for justified restrictions on access.

01. Purpose and scope

The policy of Technosphere Safety applies to raw and processed data, software code, computational models, protocols, images, maps and other materials supporting the conclusions of published articles. It covers original and reused data, including materials from special issues and supplementary files.

Reproducibility in this policy means the ability to verify how the reported results were obtained using the described data, methods and computational procedures. Repeating a study with new observations constitutes an independent verification and does not replace providing materials from the original research. Authors must identify limitations on exact computational reproduction, such as algorithmic randomness or the unavailability of a historical version of an external service.

Research articles must provide a minimum sufficient set of materials to verify their conclusions. As a general rule, this set is made openly available by publication; justified restrictions are considered under this policy. Applying the FAIR principles means making materials findable, accessible, interoperable and reusable; it does not require unconditional disclosure of protected information.

02. Data and documentation requirements

The minimum sufficient set includes data necessary to verify the principal calculations, tables, figures and conclusions, together with information linking them to original observations. Final averages or graph images alone are insufficient if they do not permit verification of the analysis. Authors are not required to deposit unrelated materials from the entire research project.

  • For field, industrial and laboratory research: observations for individual experimental units, identification of sites or objects and replicates, dates, measurement conditions, units, and information about instruments and calibration necessary for quality assessment.
  • For geospatial and time-series data: source, coverage, coordinate reference system, spatial and temporal resolution, time zone where relevant, treatment of missing values and transformations. Sensitive coordinates are protected under restricted-access rules.
  • For computational studies: input data used, significant output results, parameters and configuration files allowing the calculations to be linked to published conclusions.
  • For systematic reviews and quantitative synthesis: search strategy and date, selection criteria, list of included sources, extracted values and calculations; third-party full-text publications must not be redistributed without the necessary rights.

Files must be accompanied by documentation describing their structure, variables, codes, units, missing-value indicators, processing steps and relationship to the article. Common machine-readable formats are preferred; when a specialised format is used, the required software and reading procedure must be identified. Tables must not be supplied solely as scans when the underlying numerical data are available.

03. Description of methods and analytical verifiability

Methods must describe the study design, experimental conditions, inclusion and exclusion criteria for observations, sample size, independent replicates, treatment of missing values and outliers, calculation formulas, statistical procedures and uncertainty assessment. Technical repeated measurements must not be presented as independent experimental units. Decisions made after examining results must be identified when they affect interpretation.

For models of industrial risks, hazardous substance dispersion, fires, accident processes and other engineering systems, authors must specify initial and boundary conditions, parameters, calibration, validation and significant assumptions. Authors explain the separation of calibration and validation data and the limits of transferring conclusions to other facilities, operating conditions and regimes. Key calculations must be traceable, while rounding and permissible computational differences must be understandable.

Where a preregistered protocol or analysis plan exists, a reference and any deviations must be reported. Registration performed later must not be described as prospective preregistration. Negative findings and checks relevant to the reported conclusion must not be concealed to strengthen that conclusion. These requirements do not impose mandatory preregistration on every engineering study.

04. Deposit, versions and availability periods

At submission, data and analytical materials must be accessible to editors and reviewers through a secure method. By publication, the openly shareable portion of the minimum dataset must be deposited in a suitable disciplinary, institutional or general-purpose repository. Authors must select a service with clear preservation arrangements, metadata, version control and a persistent identifier, preferably a DOI. A repository's name does not replace verification of its actual capabilities.

Links must lead to the specific dataset and version used, rather than merely the repository homepage. For changing external databases, the version or extraction date and exact criteria used to obtain the data must be stated. Where rights permit, the data snapshot used must be preserved. A code development repository, personal website or temporary cloud link does not, by itself, replace an archived version of research materials.

The deposit plan and permissible exceptions must be agreed upon before article acceptance. A temporary private link may be used during peer review and replaced with a permanent link before publication. Embargoes are permitted only for a specific justified reason, with an agreed end date and published access conditions. An intention to prepare additional articles is not, by itself, sufficient justification.

Small datasets may be provided as supplementary files only with editorial agreement, provided they are available alongside the article, unambiguously identified and suitable for reanalysis. Repository deposit remains the preferred approach. Uploading a file to OJS solely for peer review does not constitute open deposit.

05. Data and code availability statement

Each article must contain a separate "Data and Software Code Availability" statement, provided at submission and updated before publication. It must identify the main categories of materials, their location, version or identifier, terms of use and all restrictions. The data availability statement must be consistent with the methods section and the declaration of AI use.

  • Open materials: "Data [description] and code [description, version] are deposited in [repository] at [DOI or persistent link]; terms of use: [licence]."
  • Third-party materials: "[Dataset, provider, version/date] were used. Access is available through [source and conditions]. Derived materials and code are available at [location] or restricted for [specific reason]."
  • Restricted access: "Materials [description] are not openly available because of [reason]. The custodian is [institution/department]; requests may be submitted through [contact or webpage]. Request review conditions are [criteria, procedure and response time]. The open portion is available at [location, if applicable]."
  • Materials fully included in the article: "All data necessary to verify the analysis presented are included in the article and [specifically identified appendices]." This statement is permissible only when the materials are genuinely complete and the arrangement has been agreed upon under the preceding section.
  • No data: "No new data were generated and no existing datasets were analysed in this work. Data availability is not applicable." This option must not be used where the study reanalyses third-party data or generates modelling results.

Code availability, or the absence of research software code, must be stated separately. Examples must be completed according to the actual circumstances; unverified DOIs and promises of future deposit must not be presented as existing access. The statement "available from the author upon request" is insufficient without explaining the restriction and access procedure.

06. Justified restrictions and controlled access

Valid grounds may include protection of research participants and personal data, specific risks to safety or protected facilities, third-party rights, contractual obligations and other applicable restrictions. The justification must be submitted to the editor at the time of submission. The word "confidential" alone is insufficient; supporting documents may be reviewed privately where necessary.

Authors should provide a safe, openly available portion wherever possible, including anonymised or aggregated data, metadata, descriptions of variables and code. For precise coordinates of hazardous industrial facilities, critical infrastructure or locations where potentially dangerous substances are stored, the adequacy of a generalised location must be assessed. Generalisation must not misrepresent modified data as original observations or conceal important analytical limitations. Applicable rights and agreed governance arrangements must be respected for community data.

For controlled access, the custodian, permitted purposes, applicant requirements, decision-making procedure, expected response time and grounds for refusal must be established in advance. Conditions must not depend arbitrarily on an applicant's agreement with the article's conclusions; co-authorship must not be a condition of access. If disclosure is entirely impossible, this must be stated explicitly rather than making an unfulfillable promise.

The editor assesses whether sufficient independent verification is possible, including permissible secure inspection or assessment by a qualified independent specialist. Each exception is recorded with its justification and available alternative. A manuscript may be rejected if its central findings cannot be verified reliably; the journal does not require violations of law, participant consent or the data owner's rights.

07. Software code, models and AI

Authors must provide their own code, scripts, spreadsheet formulas and settings that materially affect the results. Software and dependency versions, execution procedures, input and output files and, where necessary, computational environments, random seeds and hardware requirements must be specified. Documentation must allow a specialist to understand the sequence of analysis without personal explanations from the authors.

The use of commercial or proprietary software is not prohibited in itself. Authors must specify the exact version, available settings, lawful means of obtaining the software and verification limitations; their own scripts and input data must be shared within the limits of their rights. Where a key algorithm is unavailable, the editor evaluates the adequacy of alternative verification. Full reproducibility must not be claimed where significant components remain inaccessible.

For machine learning, significant information concerning training, validation and the model version used must be preserved; where necessary and legally permissible, this may include model weights, prompts and response-processing rules. Measured, reconstructed and synthetic values must be identified separately. Detailed requirements for disclosure and permissible uses of AI are established in a separate policy.

08. Citation and licensing

Datasets, models and software products used in the research must be cited in the text and reference list as distinct research materials. Where possible, citations include the creator, year, title, version, repository or provider, and DOI or another persistent identifier. Citing an article describing a dataset complements but does not replace identification of the data actually used.

Clear reuse conditions must be specified for each deposited object. Authors should select an open licence for their own data where they hold the necessary rights; an appropriate software licence applies to code. The article's CC BY 4.0 licence does not automatically alter rights concerning a separately deposited dataset, program or third-party material. Restrictions and exceptions must be clearly identified.

Authors must not grant rights they do not hold or remove source attribution when processing data. The existence of a licence does not replace citation. Detailed provisions concerning copyright and reuse of published articles are set out in the Copyright and Licensing Policy.

09. Verification during manuscript review

The editorial office checks for the availability statement and verifies that the stated access routes function. Reviewers assess whether the data and descriptions are sufficient to verify the main conclusions and may request specific files, explanations or control calculations through the editor. Verification does not require the editorial office to rerun every calculation and does not guarantee the absence of all errors.

For double-anonymous peer review, anonymised copies or suitable private links without author-identifying metadata must be used; information on provenance and the responsible custodian remains with the editorial office. If complete anonymity is impossible because of an existing open dataset, the editor must be informed. Scientific verifiability must not be sacrificed solely to remove information that is essential to the assessment.

Non-public materials supplied for verification are used solely for the editorial task and must not be shared with other persons or external generative AI services. Access rights are limited to necessary participants. Before a final decision, authors must resolve or adequately explain significant discrepancies among the article, dataset and code. Requirements concerning the independence and confidentiality of peer review apply in full.

10. Preservation, updates and enquiries

Authors must preserve original research materials, documentation and code used for at least 5 years after publication; a longer period applies where required by an institution, contract or other applicable rules. This is the journal's minimum retention period, not a universal requirement imposed by indexing databases. For information subject to mandatory retention restrictions, permissible periods and safely preserved documentation must be agreed upon in advance. Materials relevant to unresolved concerns must not be destroyed within the limits of lawful retention while the matter remains under consideration.

A responsible custodian must be designated; a change in an author's employment must not leave data without an accountable person. Authors must arrange backups, access protection and transfer of responsibilities. Expiry of the minimum period does not mean that openly deposited datasets must be deleted from repositories. Preservation of research materials is distinct from archiving published issues, which is governed by a separate journal policy.

Corrections to data must be issued as a new version with a description of changes and a link to the version used in the article. Files that alter the meaning of published results must not be silently replaced. Authors must notify the editorial office of significant errors, loss of access or changes in access conditions; their impact on the article is considered under the Corrections and Retractions Policy.

Problems with access or reproduction should be reported through OJS or to technosphere@tiiame.uz, identifying the article, version of the materials and problem encountered. A disagreement or unsuccessful replication attempt does not, in itself, establish misconduct; the procedures for handling allegations and appeals apply. The actual dates of policy approval and publication are recorded. New formal requirements are not retrospectively attributed to earlier publications, although concerns about their reliability may be investigated.

Methodological references

This policy takes account of the DOAJ principles of transparency, FAIR principles, AGU recommendations concerning data and software, and FORCE11 data citation principles. The specific requirements, exceptions and retention period below are established by the journal; they do not constitute a claim that the journal operates its own data repository.

Back to top


ISSN 2181-3981 (Online)