Flooding that occurred in southwest Iowa during June 15–17, 1998, was the worst flood ever recorded on the Nishnabotna River, exceeding the theoretical 500-year flood calculated from peak-flow records (1922 to present). This flood was a direct consequence of severe thunderstorm activity that caused more than 4 inches of rain to fall over a large part of the Nishnabotna River Basin. In fact, a new official State record for 24-hour total rainfall (13.18 inches) was set by this storm. The peak streamflow of the Nishnabotna River near Hamburg, Iowa, was 65,100 cubic feet per second, about 20 percent more than any previous recorded peak streamflow at this site.

To determine the concentrations of selected contaminants that might be present in this record flooding, water-quality samples were collected within hours of the flood peak. The results from these samples documented the presence of numerous herbicide compounds (11 parent compounds and 12 herbicide degradates). The highest herbicide concentration was 5.06 micrograms per liter (µg/L) for atrazine, followed by metolachlor (1.16 µg/L), metolachlor ESA (1.04 µg/L), acetochlor OA (0.99 µg/L), and acetochlor ESA (0.95 µg/L). The total herbicide concentration (summation of the 23 herbicide compounds detected) was 15.6 µg/L. The timing of the severe thunderstorm activity and flooding, which occurred shortly after chemical application associated with planting of crops, was the principal reason for the large number and concentrations of herbicide compounds found in the flood water.

At the time the water-quality samples were collected, the Nishnabotna River was transporting about 6,000 pounds of suspended sediment, 18 pounds of nitrogen, 3 pounds of phosphorus, and 0.02 pound of atrazine each second. These loads were about 10 to 150 times greater than those during a previous runoff event, and about 260 to 4,600 times greater than those during a previous base-flow condition.

This sampling demonstrates the importance of collecting both water-quantity and water-quality data during flood events to estimate contaminant loads. Potential environmental effects of a flood can only be understood when both components are measured.