At each site we installed an autonomous logger (miniDOT; PME) for dissolved oxygen and temperature measurements and a submersible fluorescence sensor (Cyclops 7-F; Turner Designs) calibrated for the detection of suspended Chlorophyll-a concentration [Chl-a]. We installed both sensors at UPSite and DWSite on March-2019 and June-2019, respectively, and set them up to record at a 5-min interval. Fluorometers were periodically cleaned and recalibrated in-situ (usually every 2 weeks). We occasionally collected a 1 L water sample to analyze acetone-extractable [Chl-a] in the laboratory following standard methods (and using an AU-10 fluorometer (Turner Designs) in the laboratory.
For each event, Chl-a and DO data were paired with discharge and turbidity data from corresponding USGS sites at 15-min intervals. We estimated hysteresis and flushing indices for both Chl-a and turbidity data. First, concentration (as Chl-a or turbidity) values, i, for each stormflow event were normalized (xi - xmin/(xmax -xmin) to allow comparisons among sites and dates, interpolated by linear regression at 1% intervals, j, of normalized discharge using two adjacent measurements. The hysteresis index (HI) at each discharge interval is determined by simple subtraction of corresponding interpolated values in the falling (Ci,fall) and rising (Ci,rise) limb. Stormflow HI represents the average of HIj values and varies from -1 to 1. Then, we calculated the flushing index (FI) as the slope between concentration (as Chl-a or turbidity) prior the storm and at the peak flow.
