The fact that the system that alerts millions of Americans to tornadoes, flash floods, and potentially fatal storms was initially developed at the same time that people were still learning how to use dial-up internet is particularly ironic. Since the 1990s, the Advanced Weather Interactive Processing System, or AWIPS, has served as the foundation of the National Weather Service. In general, it has been successful. However, being adequate and working are two different things, and for the forecasters who depend on it on a daily basis, the difference has gotten bigger every year.
That is now changing. Contracts to completely replace AWIPS have been awarded by NOAA‘s National Weather Service, which will move the entire operation into the cloud and introduce tools designed for this decade rather than the previous one. Accenture Federal Services is one of the companies chosen to contribute to the development of the next two platforms, which appear to be a real advancement at least on paper. The first, known as NWS HIVE, is intended to serve as the primary workspace where forecasters conduct data analysis, create forecasts, and issue alerts. The second, NWS CIRRUS, is a cloud-based data repository that serves as a single, central location to store and retrieve data that is owned by NWS as well as data that is received from partners in the private sector. It’s the kind of structural change that subtly affects nearly everything but receives little public notice.
Observing this develop, it is striking how much of the constraints of the previous system were related to geography. Forecasters were essentially confined to their desks under AWIPS, with hardwired workstations in designated offices that downloaded data to nearby computers before they could start working. There was no simple way to enlist the help of a coworker who lived two states away if a significant storm struck and a nearby office was understaffed. That is precisely what the new system is meant to address. In theory, forecasters will be able to work from laptops, log in remotely, and instantly provide real-time support to other offices. It’s unclear how well that flexibility will work in practice, particularly in high-stress, high-chaos scenarios where coordination tends to deteriorate most quickly.
Although ambitious, the two-phase rollout is carefully planned. Phase 1, which aims to be finished by early 2027, focuses on setting up the fundamental infrastructure, including data storage, tools for visualization, and the capacity to provide standard public forecasts and warnings. The complete transition takes place in Phase 2, which is scheduled for early 2028. Once HIVE and CIRRUS are handling the entire operational load, AWIPS will be formally decommissioned. The agency in charge of 122 local forecast offices, 13 river forecast centers, and numerous other facilities across the nation has a limited amount of time to replace its central nervous system.

It’s possible that the human aspect of the transition—training forecasters who have worked in one system for their entire careers to function well in a totally different one—will prove to be more difficult than the technology itself. The administrator of NOAA, Neil Jacobs, has been quite straightforward about the daily friction caused by the previous structure, characterizing the time required to simply download data to workstations as a significant hindrance to operational capacity. Many forecasters seem to feel that this upgrade will be more of a relief than a disruption, though that will likely depend on how well the migration actually goes.
The larger effort is contained within a framework known as “Ken’s 10,” a list of transformation priorities that NWS Director Ken Graham created when he assumed the position in 2022. One component of it is cloud infrastructure. It reads like a long-overdue reckoning with decades of postponed modernization—next-generation radars, AI-driven forecast modeling, enhanced emergency alert systems. The entire endeavor has been framed by Secretary of Commerce Howard Lutnick in terms of public safety and national competitiveness, which is probably the correct frame. It’s not just a technical issue when the system that warns people to seek shelter is based on architecture that is thirty years old. It’s open to the public.
At a time when the stakes are truly high, Accenture was awarded the contract to update NOAA’s National Weather Service. Extreme weather events are more common, climate patterns are less predictable, and the window of opportunity for delayed or inaccurate warnings has shrunk significantly. It remains to be seen if HIVE and CIRRUS fulfill their promise, but it seems noteworthy that the replacement process has officially begun.
