JAWRA - Stormwater Runoff - Laserfiche WebLink

Home Browse Search

SEDAN AND CLAUSEN FIGURE 2. Traditional Watershed Subdivision for the Jordan Cove Project, Connecticut. FEET 0 50 100 300 METERS 10 50 10 — - -- — -- i Match line see Figure 2 FIGURE 3. LID Watershed Subdivision for the Jordan Cove Project, Connecticut. was used with reduced lawn areas, replaced by low - mow and no -mow areas. Deed restrictions prevented increased impervious surfaces, alteration of LID structures, rain garden filling, and changes to origi- nal driveway surfaces during the course of the study. Education sessions, informational packets, and soil testing were used to instruct homeowners on LID practices, such as lawn maintenance, rain garden care, and rain barrels for rainwater use. Study Design The paired watershed approach was used for this study (Clausen and Spooner, 1993). A primary strength of the paired watershed approach is that it corrects for year to year differences in precipitation, which can confound temporal studies. One control and two treatment watersheds were monitored during cal- ibration, construction, and postconstruction periods. Calibration occurred prior to any applied treatment in the watersheds, and regression relationships of paired runoff observations were established between the con- trol and the two treatment watersheds. The construc- tion of houses in the traditional and LID watersheds began the first treatment (construction), while the second treatment (postconstruction) began immedi- ately after construction was complete. Periods of calibration and treatment varied by date and length depending on the watershed and construc- tion schedules (Table 2). The length of the calibration period was determined by when construction began in the watershed and when a significant calibration equation existed between the paired watersheds. Since construction took longer in the LID watershed, the postconstruction period started later than in the traditional watershed. This paper presents only the data and results from the calibration and post - construction periods through June 2005. Monitoring Discharge from the control watershed was moni- tored with a combination rectangular/V -notch weir in a 76 cm stormwater pipe. A precalibrated 45.7 cm 11-flume was used to measure overland flow during the calibration period in the traditional watershed. During the postconstruction period, a precalibrated Palmer - Bowlus flume in a 38.1 cm stormwater pipe was used to measure discharge. Overland flow in the swales was monitored in the LID watershed using a 45.7 cm H -flume. Stormwater runoff from each watershed was moni- tored continuously using ISCO 4230 bubbler flow - meters (Isco, Inc., Lincoln, Nebraska). Fifteen - minute discharge was downloaded weekly from each ISCO flowmeter using a computer. An ISCO 2900 sampler (Isco, Inc.) collected flow - weighted, composite samples of stormwater runoff directly in a refrigerator. Sam- ples were collected in three plastic bottles. One bottle was preacidified with sulfuric acid for nutrient pres- ervation, a second bottle preserved metals with nitric acid. The third bottle was not acidified and was used for suspended solids analysis. Collected samples were retrieved weekly, and transported in a cooler with ice packs to the Univer- sity of Connecticut, Storrs and stored at 4 °C until analyzed. When discharge occurred during site visits, grab samples were collected from each monitoring station in sterile sampling bags for analyses of fecal coliform (FC) bacteria and 5 -day biochemical oxygen demand (BOD JAWRA 1000 JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION Match line see Figure 3