Sign inRegisterCheck connection

Building a Scientific Bibliography for a Personal Weather Site

Build a scientific bibliography for a personal weather site: primary sources versus blogs, outbound-link labels, and keeping a resource notebook from rotting.

Back to Personal weather station notebooks

A personal weather site (PWS) needs a bibliography the way a station needs metadata: so a visitor can tell an NWS product from a blog, and so the operator can find the definition of a knot two years later. This notebook is a method for building that bibliography. It is not a dump of outbound links. Links rot; classification does not have to.

Historical context

The historical page at this URL was a resource-and-navigation strip: “links to other weather resources,” mixed with leftover dashboard chrome and a pointer at a PHP calculate notebook. The useful intent was a teaching list. The failure mode of every such list is the same: unlabeled URLs, dead hosts, and a blog post standing in for a WMO guide. Rebuild the list as a small catalog with roles, not as a blogroll.

Input

Each outbound resource is a catalog record:

| Field | Meaning | |---|---| | Title | The document’s own title | | Publisher | NWS, NCEI, WMO, NHC, vendor, journal, or named person | | Class | primary, vendor, handbook, software-doc, commentary | | Role | What a PWS operator uses it for | | Permalink | Prefer office or /data/ paths over campaign URLs | | Checked | Date you last loaded it | | Status | live, moved, retired |

You also need a scope sentence for the notebook itself: this site’s bibliography supports this station’s instruments, units, and climate context. A Mesa-era site did not need every WFO’s CLI. A tropical-methods page needs NHC, not a random radar aesthetic.

Primary sources versus everything else

Primary, for weather and climate

Primary, for the logger you actually run

  • Vendor manuals and description files (Weather Display, Cumulus, Davis, and so on). These govern packet layout. They do not govern climate normals.

Handbooks and peer-reviewed papers

Use when you need a method (psychrometrics, siting). Cite the paper or the NWS technical attachment, not a screenshot of someone else’s calculator.

Commentary

Blogs, forums, personal notebooks (including this legacy cluster), and news articles. Allowed as pointers and as historical color. Never as the authority for a number you print as fact. If a forum post is the only place you saw a ClientRaw index, verify against the vendor description file.

A practical test: if the link died, could you recover the claim from an office or a DOI? If not, the claim should not have been on the station site.

Process: labeling outbound links

Every visible link needs a label that survives skimming.

Pattern: [Publisher — document type] short title (what you will get)

Examples:

  • NWS — live hazards: your local forecast office
  • NHC — Tropical Cyclone Report: Katrina 2005 (PDF, dated)
  • NCEI — U.S. Climate Normals 1991–2020
  • Weather Display — vendor site and clientrawdescription.txt (on disk)
  • Commentary — historical TNET notebook on ClientRaw parsing

Do not use “click here,” “great site,” or a bare domain. Do not use official logos. Do not imply NOAA endorsement of the PWS.

On the page, group by class (Official observations and warnings, Climate archives, Station software, Methods notebooks, Commentary). Mixing those groups in one alphabetized list is how a blog inherits the visual authority of weather.gov.

Mark time-sensitive targets:

  • Live — NWS hazards, NHC active cyclones. No archive date; the URL is a door, not a document.
  • Dated — TCR PDFs, normals sets, service assessments. Show the document date.
  • Versioned — software docs. Show the software version the note was checked against.

Process: keeping the notebook from rotting

Link rot is certain. Editorial rot is optional.

  1. Cap the list. A working bibliography for a PWS is dozens of records, not hundreds. If you cannot review it, you do not own it.
  2. Review on a calendar. Official live portals: quarterly HEAD/GET. Dated PDFs: yearly, or when you cite a number. Vendor docs: at each software upgrade.
  3. Store checked and status. When a URL 404s, search the publisher first. If a replacement exists, update the permalink and keep the old URL as moved. If not, retired and remove the public link. Do not leave “coming soon.”
  4. Prefer stable paths. nhc.noaa.gov/data/tcr/ and ncei.noaa.gov/products/… outlive decorative landing pages.
  5. Do not hotlink images or live products into your template. Cite and send the reader. Hotlinked radar is both brittle and easy to misread as your own observation.
  6. Separate the fact file from the bibliography. Rotating educational cards have their own sourcing rules (weather facts notebook). The bibliography is the shelf those cards point at.
  7. Retire commentary aggressively. A 2005 PHP tricks page can remain in this legacy cluster as history. It should not remain in the “official methods” group.

When you add a new instrument, add its vendor doc the same day. When you add a climate claim, add the NCEI or NWS table the same day. A bibliography that lags the homepage is how folklore starts.

Output

A maintainable PWS bibliography page shows:

  • Scope sentence
  • Grouped, labeled links
  • Document dates where they exist
  • A “last reviewed” line for the page
  • A clear statement that the site is not NWS

That is enough. A wall of unlabeled URLs is less than enough.

Troubleshooting

Half the links are dead. You stored bookmarks, not records. Rebuild from publisher homepages; do not scrape old sidebars.

Readers think your commentary is a warning. Move NWS/NHC to the top. Label this site’s notebooks as methods.

A vendor PDF moved. Link the vendor root and the on-disk filename, not a third-party copy.

You want to list every WFO. Don’t. List the offices that cover your station and the national hubs. Geography is part of metadata.

SEO temptation. Do not add high-traffic keywords that are not your subject. A bibliography earns trust by classification.

Modern relevance

TNET’s public research library is organized the same way this notebook asks a PWS bibliography to be organized: sources and quality rules in one place, methods in another, no claim that a research brief is an NWS product. Start with data sources and methodology. Neighboring methods: notebooks hub, psychrometric calculations, NWS statement literacy.

Sources