Forecast Seed is a forensic meteorology of weather modification. Cloud seeding sits in a strange public position: it is a real, permitted, documented activity with a fifty-year federal paper trail — and simultaneously the subject of claims wild enough that three states have banned it by law. The operators publish optimistic numbers; the critics publish outrage; the actual record — filings, permits, flight tracks, radar archives, snow chemistry — mostly goes unread. This site reads it.

It is part of the same independent network as weather.city, Forecast Eagle and Forecast dBZ — places, models, radar — and it uses the same instruments: the models Eagle reads and the MRMS radar dBZ serves are the same feeds this site's seedability and forensics work is built on.

Every quantitative claim on this site carries one of four chips, and the difference between them is the entire editorial policy:

RECORD — a filed document: a NOAA Form 17-4, a state permit, a statutory annual report. It proves what was declared, not what worked.
MEASURED — an observation: radar, a gauge, a transponder track, ICP-MS chemistry on a snow sample, a randomized experiment's result. The strongest chip, and the rarest.
MODELED — a simulation, with the model named. Most seasonal seeding estimates live here, including the good ones.
CLAIMED — an operator's or government's own number, not independently verified. Not an insult; just a label that does not wash off.

Not a warning source. For watches and warnings, go to weather.gov. Not an operator. This site neither conducts nor sells cloud seeding, and takes no money from anyone who does. Not a debunking site, and not a believing one. When someone asks "was there seeding here?", the answer is whatever the five layers of evidence say — which is sometimes "yes, 70 grams of it, two days earlier and 150 miles away," and sometimes "no, and the atmosphere was unseedable anyway." And not about chemtrails. Persistent contrails are ice from jet exhaust; cloud seeding is a permitted activity with a paper trail. Conflating them is how a checkable subject became an uncheckable one, and keeping them separate is half this site's reason to exist.

The site is written from primary sources: the NOAA weather modification filings and state permit systems (see Records), program annual reports, the peer-reviewed literature from the 1981 first radar detection through SNOWIE and the modern detectability theory, and public data feeds — NEXRAD Level II and MRMS archives, ADS-B flight history, SNOTEL and gauge networks, GOES cloud-top phase, and operational model fields. Each analytical page states its method and its blind spots inline, because a forensic site that hides its error bars is just another claim.