Development of an Antecedent Precipitation Index model to detect seasonal changes in fecal coliform levels during storm events.

An Antecedent Precipitation Index (API) model was created for Janes Creek in Arcata, California. The API model was used to predict and quantify periods of poor surface water quality in connection with precipitation events and land use practices. Janes Creek currently does not have a streamflow gage; therefore the API model was used to generate streamflow data. Streamflow data allowed fecal coliform monitoring to be performed in context of hydrological conditions. Additionally, the API model accounts for soil saturation levels, which is important in determining runoff response. Janes Creek was sampled for fecal coliforms from November of 2003, through February of 2005. Three sampling sites included one reference site above cattle grazing activities and two below. Over the sampling period, regular monthly samples were analyzed along with samples from a series of storms that occurred during the storm season of 2004-2005. Fecal coliform concentrations increased 100 fold from background levels during the first major storm of 2004, which registered over 0.5 inches of precipitation. Fecal coliform levels were significantly higher (p 0.05) and associated with storm events at the two sites below cattle grazing activities. Based on findings in this watershed, increased precipitation and streamflow will increase enteric bacterial concentrations in surface waters affected by catchments used for grazing dairy cattle. Forecasted precipitation amounts applied in the API model allowed both sampling timing and pollution events to be anticipated ahead of time.