Data sources and methodology

Every fact on Dump Nearby comes from a named public record. This page explains where the records come from, how they become listings, and the rules that decide what gets published.

Current dataset

• 5117 facility records: 4859 published, 258 held as drafts pending checks.
• 50 states with at least one published record; 95 city pages (cities with 5 or more facilities).
• 516 records with confirmed hours, fees, accepted materials or an official website; 580 records meet the bar for search indexing.

Sources

Records cite one or more of these sources. The count is how many published listings cite each host.

• www.epa.gov: 1212 listings
• www2.calrecycle.ca.gov: 855 listings
• data.ny.gov: 720 listings
• www.tceq.texas.gov: 615 listings
• ca.dep.state.fl.us: 546 listings
• enterprise.gisdata.mn.gov: 383 listings
• edocs.deq.nc.gov: 285 listings
• gisagoegle.state.mi.us: 223 listings

The EPA Landfill Methane Outreach Program (LMOP) database supplies ownership, acreage, waste-in-place and gas-collection facts for municipal solid waste landfills. State registries (for example CalRecycle's SWIS and the Texas Commission on Environmental Quality) supply permit numbers, permitted activities, throughput limits and operational status. Facility and county websites supply phone numbers, hours and fee schedules when they publish them.

From record to listing

1. Normalize. Each source row is mapped onto one schema (name, type, address, coordinates, attributes). Field names from different agencies are reconciled to a single label set.
2. Deduplicate. Rows describing the same physical site (same permit, or same name within a short distance) are merged. Alternative names are kept on the record but not shown.
3. Attribute provenance. Every field lists the source URL it came from, the date it was fetched, and which fields that source supports. The listing page prints this at the bottom.
4. Check. A record's "last checked" date is the date its sources were last compared with the live page. It is never set to the build date just to look fresh.

What gets indexed

A listing is submitted to search engines only when it answers a question someone would actually search for: confirmed hours, a fee schedule, an accepted-materials list, or an official website. A record with none of those is still indexed if it carries at least 6 verified attributes (registry IDs, permit numbers and "unknown" values do not count) and sits in a city that has its own page; sites that are not open to the public (restricted access, captive or industrial-only facilities, agency maintenance yards, street-sweeper dumps) are never indexed; otherwise it stays on the site, linked from its state page, with a noindex tag. State pages with fewer than 5 published facilities are also kept out of the index until more records are checked.

What is left out on purpose

• Ratings, reviews and "verified" badges of any kind.
• Hours, fees or phone numbers that no public source confirms. Missing means missing on the page and in the structured data.
• Generated descriptions of towns, weather or local history.
• Registry identifiers are shown under "More details" but do not count as facts about the facility.

Known limitations

Coordinates from some registries are centroid-level rather than gate-level; the directions link opens a map so you can confirm the entrance before driving. Agency datasets lag real-world changes by months, so a facility can close or change hours before the registry does. Call before you go.

Corrections

Submit changes through the correction form. We verify against the facility or its agency and publish the change with its source and a new "last checked" date. Operators can claim a listing through the same form.