HiRO-ACEv2: Simulating extreme weather at 3 km resolution for a fraction of the cost

What is HiRO-ACEv2

HiRO-ACEv2 is the newest update to the Allen Institute for AI (Ai2) High Resolution Output Ai2 Climate Emulator (HiRO-ACE) system, published October 14, 2026. It combines ACE, an AI model operating at a coarser 100 km resolution, with HiRO, a downscaling model that sharpens those coarse outputs into fine 3 km detail. The full system can generate global 3 km climate data at about 1/300th the cost of X-SHiELD, a high-resolution global atmospheric model that explicitly simulates storms and other small-scale weather.

Improved extreme event simulation

The biggest change in HiRO-ACEv2 is the new downscaling model, HiROv2, which can predict several high-resolution fields at once instead of just rainfall, capturing a much wider range of extreme weather. When given coarsened X-SHiELD data, it matches the target simulation's high-resolution statistics to within 3%, even for events so rare they occur only once in a million data points. It also accurately estimates the total energy of tropical cyclones, including storms it never saw during training, and generates hurricanes with realistic eyes and rain bands, though winds and pressure around the eye remain somewhat more lopsided than in X-SHiELD. HiROv2 additionally realistically generates strong winds flowing down mountain slopes.

Long-term simulations and cost savings

An updated version of ACE trained for this system produced a 10-year simulation whose hurricanes, average climate patterns, and overall distributions of rain, wind, and pressure all line up well with X-SHiELD. Putting HiROv2 and ACE together, Ai2 generated a full year of global 3 km data from a simulation ACE ran on its own rather than from real X-SHiELD data, and its statistics closely match. Across a 10-year run, the energy of HiRO-ACEv2's hurricanes is comparable to X-SHiELD's over the same period, and snapshots show a variety of other extreme events resembling those in X-SHiELD. Producing global high-resolution data costs roughly 300 times less than running X-SHiELD, and if only one region is needed, the savings can be more than a hundred times greater depending on the size of the area. Ai2 is also working on a streamlined HiROv2 version that could make it more than ten times faster still.

Limitations and future work

Because HiRO learned from X-SHiELD, it only works with the version of ACE tuned to match X-SHiELD, so it cannot yet produce high-resolution data for other climates, such as a warmer future or the historical record in datasets like ERA5. Ai2 plans to train on new X-SHiELD datasets, including one simulating a world 4 degrees Celsius warmer, so the system can handle different climates. The team also plans to add outputs like near-surface temperature and snow and ice, enabling uses such as snowpack prediction. To help researchers inspect results, Ai2 built a web demo showing pre-generated precipitation, wind, and sea-level pressure output from HiROv1 and HiROv2 side by side with X-SHiELD, along with tutorial notebooks and downloadable models for deeper analysis.

Sources: Hugging Face Blog