Versions:

2.7.4
2.7.3
2.7.2
2.7.1
2.7.0
2.6.5
2.6.4
2.6.3
2.6.1
2.6.0
2.5.8
2.5.7
2.5.6
2.5.4
2.5.3
2.5.2
2.5.1
2.5.0
1.3.1

ArWen, published by Fahrenheit Research and currently at version 2.7.4 (one of 19 recorded releases), is an independent, GPU-native implementation of a WRF-ARW-class regional atmospheric model, verified against WRF v4.6.1. It is not affiliated with or endorsed by NCAR or UCAR, and the name is a wordmark for "the ARW solver, GPU-native"; the Python package itself is currently named gpuwm. The software falls into the regional numerical weather prediction and atmospheric modeling category, with its purpose being to bring kilometer-scale limited-area forecasting to a single consumer GPU rather than requiring the institutional clusters that convective-scale modeling has traditionally demanded. Its core use case centers on accessibility: a user can pick a point, size a nest ladder to match their graphics card, pull public analysis data, and run a 1 km or even 500 m simulation of their own area on hardware they already own. This is especially relevant in regions where no national convection-permitting model exists, and where most of the world currently relies on global-model guidance at 10–25 km resolution, leaving local-scale weather detail unresolved. ArWen thus targets researchers, educators, and enthusiasts who want verified, high-resolution limited-area modeling without access to supercomputing infrastructure. The publisher is explicit about the tool's boundaries: ArWen is a research and educational tool, and it is never a substitute for official forecasts and warnings issued by a national meteorological service. It should not be used to make safety decisions. This safety framing is integral to the product's positioning, reinforcing that its verified, kilometer-scale output is meant to support learning and experimentation in numerical weather prediction rather than operational forecasting. With 19 versions released and the current version at 2.7.4, the project presents an actively developed option in the landscape of GPU-accelerated atmospheric models, aiming to democratize convective-scale regional modeling for users worldwide who lack institutional computing resources.

Tags: