Abeshu Hydrosystems Intelligence Lab Handbook

Research Workflows

Practical habits for organizing projects, data, analyses, and reproducible outputs.

Policy note. This handbook is a living document and does not replace official NMSU, College of Engineering, Department of Civil & Environmental Engineering, Graduate School, funding-agency, immigration, employment, or university policies. When conflicts arise, official policies and written funding/appointment letters take precedence.

Project Organization

Each project should have a clear home for code, data notes, documentation, and outputs. Keep raw data separate from processed data and document every step needed to reproduce figures and tables.

Recommended top-level structure:

project-name/
  data/
  docs/
  notebooks/
  outputs/
  scripts/
  src/
  README.md

Project Work Plan

At the start of a new project, use the project work plan template to convert a broad research direction into trackable work. The template is especially useful for advisor-student meetings because it keeps tasks, sub-tasks, data needs, blockers, meeting dates, and status in one place.

Download the project work plan template

Suggested status labels:

  • START NEXT: ready to begin.
  • IN-PROGRESS: actively being worked on.
  • ON-HOLD: blocked, waiting, or intentionally paused.
  • COMPLETE: done for the current project stage.

Reproducibility

Every project should move toward reproducible workflows: documented code, clear data provenance, version-controlled scripts when appropriate, and enough documentation that another lab member can understand the analysis. This does not mean every exploratory notebook must be perfect, but it does mean the final path from data to claims should be inspectable.

  • Write a README before the project becomes complicated.
  • Track code in Git.
  • Record package dependencies in environment.yml, requirements.txt, or another project-appropriate file.
  • Use clear file and folder names that future you and future collaborators can understand.
  • Keep raw data unchanged.
  • Document data sources, licenses, spatial coverage, temporal coverage, and processing assumptions.
  • Save scripts used to generate figures and tables, or document exactly how final figures and tables were produced.
  • Record random seeds, model versions, configuration files, and compute environment details when they affect results.
  • Back up important project files in an approved storage location rather than relying on one laptop or one local drive.

Meetings

For recurring project meetings, keep a lightweight running agenda with:

  • Updates since the last meeting
  • Decisions needed
  • Blockers
  • Action items
  • Next milestone

Archiving

Before a paper, report, or dataset is finalized, confirm that the analysis can be rerun from documented inputs and that the final outputs are stored in a stable location.

Project README Minimum

Every active project should eventually include:

  • Research question
  • Repository owner or point of contact
  • Data sources and access notes
  • Environment setup
  • Reproduction steps
  • Output locations
  • Known limitations