Journal of Student Research 2018

76 Journal Student Research can vary depending on length of the study, where the study is taking place, and what type of samples are being collected (Robertson & Roerish 1999; Harmel et al. 2003). Types of samples that can be collected include discrete or composite samples, and types of sampling strategies include time-interval or flow-interval (McFarland & Hauck 2001; King & Harmel 2003). The types of samples collected have advantages and disadvantages. Discrete sampling collects one sample per bottle either over time or flow interval, whereas composite sampling combines several samples into one bottle (King & Harmel 2003). Discrete sampling allows for pollutant distribution data to be used and reduces sampling error (McFarland & Hauck 2001; Harmel et al. 2003). However, discrete sampling requires the analysis of many individual samples in the lab, which can be time consuming (King & Harmel 2003). Composite sampling can be advantageous by decreasing the numbers of samples that need to be analyzed (McFarland & Hauck 2001). However, load estimate errors are more likely to occur (Harmel et al. 2003). Discrete and composite sampling both have benefits and drawbacks, and choosing what type of sample to collect can depend on economic or time constraints. The different types of sampling strategies also have advantages and disadvantages. For time interval sampling, samples can be collected at uniform or nonuniform time intervals (i.e. 15 min intervals). This sampling strategy is best when flow meters are unavailable and when sampling time intervals are small (Harmel et al. 2003; King & Harmel 2003). For flow interval sampling, samples are collected as a function of flow volume (i.e., every 100 gallons, Harmel et al. 2003). This sampling strategy can measure storm loads. However, flow-interval sampling requires flow meter monitoring to measure flow volume, and if a flow meter fails then no samples will be taken (Harmel et al. 2003). Harmel et al. (2003) designed a flow chart to help researches choose the best samples to collect and type of sampling strategy to use. The study site subwatershed (storm sewer 2) drains a 577-ac area containing an extensive parking lot that serves Luther Hospital (Fig. 3). The storm culvert is round (2 ft diameter) and drains into a riprap detention area to trap sediment before finally entering the lake. An automated storm sampler (ISCO 6700) equipped with a data logger (ISCO 750 module) and flow velocity probe were used to collect runoff samples. The flow velocity probe was equipped with a pressure transducer to measure water depth (ft or m) in the culvert and a Doppler sensor to measure flow velocity (ft/s or m/s). The Storm drain sampling approach on Half Moon Lake