Development of Stormwater Control Targets to Reduce Flashiness in Urban Streams

Urbanization without proper stormwater controls has been shown to change the flow regime of streams and rivers. The frequency and magnitude of wet weather flow rates increase while the baseflow rates in dry weather decrease. Consequently, the available habitat for fish is impacted and reduced.

The Rouge River Watershed Management Plan was updated in 2009. The Rouge River watershed is highly developed and covers a large part of Metropolitan Detroit, Michigan. The fish communities in the river are limited by the flow regime. Also, water quality deficiencies exist from pollutant loadings from stormwater, sanitary sewer overflows and combined sewer overflows.

The plan considered the stormwater controls needed to protect and restore fish habitat. The measures to reduce the “flashiness” and increase the baseflow rates involve adding controls to existing stormwater basins and outlets, and changing drainage standards for roadways, development and re-development.

The USGS stream gage data for the Rouge River watershed were evaluated to develop targets for Best Management Practice (BMP) stormwater controls that may be implemented. The USGS stream flow data available included annual peak, annual mean, and daily mean flow rate data. Since 1990, 15 minute flow rate data also are available. There are seven (7) long-term stations and four (4) shorter-term stations in the watershed. The long-term stations had from 47 to 76 years of data. The shorter-term stations had 5 to 8 years of data.

Flood frequency analyses for the long-term stations were conducted for the periods of record using the annual series data. For the most recent 10 years, the 15 minute data also were obtained and partial duration series analyses were conducted. The annual series curves were combined with the partial duration series data points and plotted for each station. The overbank, 1 year, 1 month, and 15 day flood frequencies were determined.

The overbank elevations at the gage stations were surveyed. The overbank levels were used with the rating curves to determine the overbank flood flow rates. The return frequencies of the overbank floods ranged from 1.2 to 21.2 years, depending on station.

For the most recent 10 years of data, the peak flood flow rates were correlated to the daily mean flow rate for the smaller flood events. The correlation equations were used to estimate the frequency of the smaller flood flow rates over the entire period of record for the long-term stations. Clear trends were not observed in the overbank and 1 month flood frequency over the years of record. However, clear increasing trends were observed for the 15 day flood event.

The 15 minute flow data were used for the most recent 10 years of data for each station. The flood volumes and peak rates for each event above a baseflow level were determined and plotted. The events were sorted by peak flow rates that exceeded the overbank, 1 month, and 15 day flood flow rates. These flood volume plots were used to calculate targets for increasing stormwater storage and the required release rates to increase baseflow rates. These targets were used in developing a set of recommended BMPs.