Tornadoes in Michigan: 4 Surveyed Tracks, One Parent Storm, and a Demand for Answers
Tornadoes struck parts of West Michigan on March 6, 2026, in a concentrated outbreak that emergency managers and National Weather Service meteorologists spent the weekend mapping in damage surveys. Four touchdowns across Cass, St. Joseph, Branch and Calhoun counties were tied to the same parent supercell thunderstorm, a detail that is shaping how communities interpret both the rapid escalation of damage and the timeline of warnings. The human toll is already clear: Edwardsburg Public Schools is mourning 12-year-old Silas Anderson, killed when a tornado damaged his home on Conrad Road.
What the weekend damage surveys established
Damage surveys conducted by local and state emergency managers alongside National Weather Service meteorologists outlined the paths and intensity estimates for four tornadoes that touched down during the afternoon and early evening. While the surveys are technical by design—built from structural damage indicators, debris patterns, and track continuity—the results also serve as a public accounting of what happened on the ground, neighborhood by neighborhood.
One tornado first touched down northwest of Runkle Street and Conrad Road in Cass County, destroying a garage and damaging the front of a home. The Cass County Sheriff confirmed that a 12-year-old boy was killed at that location during the storm. From there, the track moved northeast, toppling trees and causing minor roof damage as it crossed Yankee Street east of Dailey Road. Survey details indicate the tornado widened and intensified along Harris Street between Hess Road and M-62, snapping or uprooting numerous trees and damaging pole barns and homes. It continued across M-62 and several local roads, passing near Curtis, Day and Chain lakes, then lifted east of South and Walnut roads at approximately 3: 35 p. m. ET. The survey information cited a track of more than 13 miles and a maximum wind estimate of 95 mph for that tornado.
A little more than 10 minutes later, a second tornado touched down in Three Rivers. Videos showed the circulation tracking just north of M-60 and causing extensive damage to several commercial buildings near West Broadway Street. A Menards store took a direct hit, with part of its roof and outer facade torn off. The tornado then moved east-northeast into the city, where it caused significant damage to other businesses and residential neighborhoods. As of Sunday night, National Weather Service personnel were still assessing the tornado’s starting and ending points; the rating and peak wind estimates remained preliminary and could be updated once final survey results are released. For residents and business owners, that “preliminary” label matters: it is the difference between an initial estimate and the final record that insurers, local planners, and public agencies often rely on.
The same damage survey reporting also referenced a Union City tornado path described as EF3, with the worst damage near Prairie Rose Lane and Tuttle Road on the northwest shore of Union Lake, weakening slightly as it moved northeast into town. In addition, survey notes referenced a Three Rivers tornado with a preliminary rating of at least EF2 and maximum winds of 130 mph, while emphasizing that a more detailed and finalized tornado summary is expected in the coming days. Taken together, the surveys depict a multi-county event where tornadoes were not isolated surprises, but components of a broader storm evolution.
Tornadoes, one parent supercell, and the risk of “compressed” impacts
One of the most consequential technical facts in the survey narrative is that all four tornadoes were produced by the same parent supercell thunderstorm. A supercell is characterized by a rotating updraft—an organizing feature that can sustain severe weather and, in some cases, produce multiple tornadoes in sequence.
That common origin helps explain why impacts can feel “stacked” across multiple counties in a matter of hours: the same storm structure can repeatedly generate damaging circulations along its track. It also complicates community response. When the damage spans rural roads, lakeside areas, and commercial corridors in quick succession, emergency resources can be pulled in multiple directions at once—tree removal, structural assessment, and immediate life-safety needs.
Separately, the event’s timing is a key part of its public significance. The survey description explicitly noted that while tornadoes can happen during any month, it is rare to see this many stronger tornadoes in early March. That rarity does not change what happened, but it does influence preparedness expectations—how people interpret the threat on a day when severe weather may not be front of mind.
Grief, institutional response, and the question of warnings
The outbreak is now inseparable from the death of Silas Anderson, 12, who was killed after a tornado damaged his home on Conrad Road Friday afternoon. Edwardsburg Public Schools issued a statement on Sunday saying the district would have support staff on hand Monday for students and staff struggling with grief. In a disaster, these measures can be as essential as debris cleanup—an acknowledgment that recovery is not only structural, but emotional and communal.
In parallel, the political and administrative stakes are rising. One of the central questions emerging from the aftermath is why a tornado watch was not issued—an issue referenced in the latest coverage as prompting a call for a probe from Governor Gretchen Whitmer’s office. The existence of that call, alongside the continuing work of National Weather Service meteorologists on survey finalization, signals a two-track process: technical reconstruction of the tornadoes and institutional scrutiny of the decision-making that framed public alerting.
It is important to separate confirmed facts from analysis at this stage. The surveys and timelines describe where tornadoes tracked and what they damaged; the inquiry focus is about warning process and whether it matched the evolving threat. Those are related, but not interchangeable. The next “finalized tornado summary” will likely sharpen the chronology of touchdowns and intensities, while any probe will examine how the warning posture aligned with storm evolution.
For Michigan communities in Cass and St. Joseph counties—and for the broader region that watched the same parent storm spawn repeated destruction—the unresolved issue is not simply how powerful the tornadoes were, but how the public warning framework functioned under pressure. With final survey results still pending and official scrutiny underway, the lingering question is whether the next fast-moving outbreak will meet a system better calibrated to act before tornadoes turn a normal afternoon into a multi-county emergency.
