7 Dead in Indiana as White River Hits Record Crest After Six Days of Storms

Seven is the number that matters in Indiana. Seven people dead after six days of storms and flooding that started Aug. 11, 2026, and pushed the White River to a crest around Sunday, Aug. 16, as the rain finally eased ^23671011. It is not a large number until you sit with who it covers: a 4-year-old boy, three men and three women, ages stretching from 4 to 86 ³⁴⁵⁹. That span is the whole story in miniature, a disaster that did not pick a neighborhood or an age, but found people where they lived, drove, sheltered and worked through nearly a week of water.

The duration is what turned rain into record flooding. The storms ran "from Tuesday through Sunday" ⁴⁹, beginning Aug. 11 and not releasing Indianapolis and much of central Indiana until the weekend crest ^23671011. Six days does not sound dramatic in the way a single tornado does, but for flooding it is the most dangerous kind of calendar. Ground saturates, creeks stay high, rivers have nowhere to put the next inch, and each additional band of rain stacks on top of the last instead of draining away. By the time floodwaters crested in Indianapolis' White River around Aug. 16, the system had wrung out almost everything it had ^23671011.

What officials have confirmed about the dead is spare but clarifying. The seven included a small child and older adults, three men and three women in all, with the youngest 4 and the oldest 86 ³⁴⁵⁹. That accounting, repeated across multiple wire-based reports, is the firmest fact in the file ^23671011. It tells you this was not one isolated tragedy, not a single car swept off a single road, but a statewide stretch of lethal conditions in which different people faced different risks on different days and still ended up in the same toll.

One of those deaths shows how wide the danger was. Paula Plebanski, 58, a homeless woman in Gary, was killed around 11:30 a.m. Tuesday when roofing blew off a store and struck her during a storm ⁴⁹. Her death was not a drowning in the White River and did not happen at the crest. It happened early, in wind, in a city far from the Indianapolis neighborhoods that would later dominate flood images. That matters because it widens the definition of this event. The research file treats Aug. 11 onward as storms and flooding together, not flooding alone, and Plebanski's death is a reminder that the first hazard was overhead before it was underfoot ⁴⁹.

In Indianapolis, the damage settled closest to the river. On the northside, where 37-year-old furniture maker Frank Gounder assessed damage on his birthday, homes hit by flooding sat less than a mile from the White River ⁶⁷. That detail, less than a mile, carries more weight than any crest chart. It places the water inside ordinary residential life, not on a distant floodplain, and it explains why recovery began as personal inventory — floors, walls, tools, birthdays interrupted — even while officials were still warning that receding water remained dangerous ²⁶⁷. Gounder's birthday becomes a marker for how disasters overwrite private time.

Hundreds of homes remained evacuated and widespread power outages persisted as floodwaters receded, with officials warning of continued dangerous conditions ². That sentence should stop anyone tempted to treat a crest as an ending. A crest is a peak on a gauge. For people, the worst phase often starts after, when water sits in walls and basements, when power stays off, when roads that looked passable hide washouts, and when cleanup brings its own injuries and contamination. The warning that conditions remained dangerous during the recession is standard emergency language, but it is also precise ². Floodwater does not leave cleanly.

The White River did reach record-setting levels, according to headlines describing the toll, and one secondary account says floodwaters crested at record levels in multiple cities citing AP ^235671011. The plain reading is that Indianapolis and other points saw water higher than they had seen before in the modern record period, enough for editors to call it record-setting without hesitation ^23671011. That consensus across several AP-based reports is strong on the fact of records, even when the visible excerpts do not print gauge heights ^23671011.

What the file does not support is the more specific history. The claim that the river shattered a 1913 record, and the associated figure of near 25 feet at Anderson and Noblesville after a rapid multi-foot rise, do not appear in the primary wire texts provided ^23671011. A secondary headline echoes the 1913 comparison ⁵, but the underlying AP reports available here do not state that detail in their visible text, so the 1913 comparison must be treated as unverified in the sources provided ⁵. The honest position is this: record flooding is well supported ^23671011, a particular century-old benchmark and a particular crest height are not, and repeating them as settled would mislead a reader who never presses play.

The day after the crest did not bring relief everywhere. Water rescues and stranded vehicles were reported in Morgan and Johnson counties on Aug. 17 ⁸. Those counties sit south and southwest of Indianapolis, downstream in both geography and timing, where water that has already peaked upstream keeps moving and keeps finding cars. A stranded vehicle report a day after a crest is not a footnote. It is evidence that the hazard migrates, that people resume driving while ditches and low crossings are still running high, and that rescue crews stay busy after national attention starts to drift ⁸.

A reader who watched early video updates will have heard larger, more cinematic numbers: inches of rain by the foot, tens of thousands without power, dozens of rescues, bridges lost, tornadoes confirmed, emergency declarations and federal aid. None of those specifics can be confirmed from the article texts provided here. The file is explicit that the visible source text does not verify a statewide disaster emergency covering more than 12 counties, specific rescue counts, a 150,000 to 170,000 power-outage figure, six confirmed tornadoes, a Hazel Bridge washout or sheared parkway bridge, an Aug. 16 federal aid request and its approval status, or rainfall totals of around 12 inches ^234567891011.

That absence is not proof those things did not happen. It is a limit on what a careful article can assert. When sources disagree or stay silent on a distance, a unit name or a tonnage, the rule here is to say they disagree or are silent, not to pick a winner. The same applies to counts that conflict for rescues, displacement and outages, and to whether federal emergency aid was intended or approved. Without corroborating text in the reporting file, those video details remain uncorroborated claims attributed to Indiana emergency officials and National Weather Service data in a single broadcast, not independently verified facts for print ^234567891011. A responsible write-up holds them out of the confirmed toll and puts them where they belong: in the watch list.

Recovery, as described in the days after the crest, was underway but not over ⁶⁷. That phrase covers a lot of unglamorous work: pumping basements less than a mile from the White River ⁶⁷, checking homes that remained evacuated ², restoring widespread outages ², and answering water calls in counties still seeing rescues on Aug. 17 ⁸. The toll held at at least seven ^23671011, the ages held at 4 to 86 ³⁴⁵⁹, and the calendar held at Tuesday through Sunday into a Sunday crest ⁴⁹. Everything else — how much rain, how many rescued, which bridges, which aid — waits for text that actually says it.