Development of Multiple Linear Regression Models for Predicting the Stormwater Quality of Urban Sub-Watersheds Amarpreet S. Arora • Akepati S. Reddy Received: 26 April 2013 / Accepted: 15 November 2013 / Published online: 24 November 2013 Ó Springer Science+Business Media New York 2013 Abstract Stormwater management at urban sub-water- shed level has been envisioned to include stormwater col- lection, treatment, and disposal of treated stormwater through groundwater recharging. Sizing, operation and control of the stormwater management systems require information on the quantities and characteristics of the stormwater generated. Stormwater characteristics depend upon dry spell between two successive rainfall events, intensity of rainfall and watershed characteristics. How- ever, sampling and analysis of stormwater, spanning only few rainfall events, provides insufficient information on the characteristics. An attempt has been made in the present study to assess the stormwater characteristics through regression modeling. Stormwater of five sub-watersheds of Patiala city were sampled and analyzed. The results obtained were related with the antecedent dry periods and with the intensity of the rainfall event through regression modeling. Obtained regression models were used to assess the stormwater quality for various antecedent dry periods and rainfall event intensities. Keywords Multiple regression analysis  Regression models  Stormwater characteristics  Urban stormwater management Stormwater runoff from urban areas can be highly polluted with various materials, indicating a significant non-point source (NPS) of pollution to receiving water bodies (Brezonik and Stadelmann 2002; Kim et al. 2005). Site and event parameters (total event rainfall, cumulative seasonal rainfall, drainage area, annual average daily traffic and antecedent dry period) were found to have significant influences on urban runoff. Irish et al. (1998) and Brezonik and Stadelmann (2002) determined that loads for each constituent are dependent upon a unique subset of vari- ables. Also, the processes responsible for the generation, accumulation, and wash-off of urban runoff pollutants are constituent-specific. Geographic and physical factors such as the type and intensity of urban land use, degree of imperviousness, tree cover, soil type and slope are also important parameters that impact the quality of urban runoff (Graves et al. 2004; Kayhanian et al. 2007). Due to the impacts on receiving waters and the expense involved in obtaining monitoring data on nonpoint source pollution data, interest has grown in analyzing existing/ measured data to develop estimation models for urban stormwater loads and concentrations (Thomson et al. 1997; Phillips and Thompson 2002). Such models will be very helpful in estimating concentrations for unmonitored watersheds. Consequently, parameters which are relatively easy to monitor can serve as indicators for other constituents reducing labour and time. Hence efforts have been made on similar lines to develop regression models for the few selected sub-watersheds of Patiala city, Punjab. Patiala (29°49 0 and 30°47 0 north latitude, 75°58 0 and 76°54 0 east longitude), city of Punjab (Northern India), does not have provisions for the stormwater drainage. As a result even modest rainfall events produce severe flooding in many parts of the city (Arora and Reddy 2012). Avoiding dis- charge of this polluted stormwater into the river Ghaggar, preventing flooding within the city and maintaining the groundwater table should be considered as utmost important while planning stormwater management system for the city. A. S. Arora (&)  A. S. Reddy School of Energy and Environment, Thapar University, Patiala 147004, India e-mail: enviro_amar@yahoo.com 123 Bull Environ Contam Toxicol (2014) 92:36–43 DOI 10.1007/s00128-013-1160-y