Truace
Subscribe
60/100

Moderate evidence

GainPublished recordClimateGood Space

Full story
More
Export record JSONEmailLinkedInX

Skillful regional NWP forecasts of tropical cyclones driven by fully AI-generated boundary conditions

Truvace record
TRV-2026-2618raw data (JSON)
Version
v1Published
First published
Evidence type
Controlled or validated comparison
Sources
1 sourcePeer-reviewed
sha256
bd1ca4bd875e…check it

Abstract Forecasting tropical cyclones (TCs) remains one of the most enduring challenges in numerical weather prediction (NWP), particularly in terms of storm intensity and track at medium to extended lead times. Recent advances in artificial intelligence (AI)-based global models have demonstrated strong skill in capturing large-scale atmospheric dynamics; however, their coarse spatial resolution and limited variable suites constrain their ability to produce storm-resolving forecasts critical for high-impact eve…

The gain

In a study of 14 Northwest Pacific typhoons from 2024, an AI-driven regional system held four- to five-day track errors to about 320–340 km, versus 500–600 km for conventional regional models.

What this doesn’t fix

Intensity errors of 10–15 m/s persisted across multiple cases, rapid intensification and weakening remain a major limitation, and the evaluation covered only 14 storms from the 2024 Northwest Pacific season.

Sources

1 on file
  1. Environmental Research Lettersdoi:10.1088/1748-9326/ae7e0a
    Peer-reviewed2026-06-16

Cite this record

v1

Truvace Impact Record TRV-2026-2618, v1: “Skillful regional NWP forecasts of tropical cyclones driven by fully AI-generated boundary conditions.” Truvace, 2026-11-12. /record/TRV-2026-2618 (accessed on the day you copy it). sha256 bd1ca4bd875ea150…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv1bd1ca4bd875e…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-2618 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.