Relative Backscatter Intensity


Seasonally averaged diel cycles of RB showed that there was a time in early- to mid-morning when vertical variation in RB was at a minimum and a time in mid- to late-afternoon when vertical variation in RB was at a maximum (fig. 3). Times of minimum (about 8:00–9:00 h) and maximum (about 15:00–19:00 h) RB near-surface occurred nearly simultaneously at the five ADCP sites (fig. 3), but varied widely near-bottom. At the deep site, ADCP1, the minimum RB near-bottom lagged the maximum near-surfaceby about 2 hours. At the other four ADCP sites, located in shallower water than site ADCP1, the minimum RB in the lower part of the water column preceded the maximum at the surface by several hours. The maximum RB at the surface at sites ADCP3, ADCP6, and ADCP7 appears to spread downward during the evening and early morning hours and displace the near-bottom minimum, moving it closer to the bottom (fig. 3).


In addition, there was good correspondence between the timing of diel cycles of winds and RB at the surface. Comparisons of diel wind and RB cycles at the five ADCP sites showed that RB preceded corresponding wind by about 2–4 hours. The average correlation coefficient, R, for the five sets of data was 0.92 (all p<0.0001). Timing of patterns was far more variable near-bottom where RB lagged winds by 2–10 hours (average R=0.88; all p<0.0001). The shortest lag was at site ADCP6; the longest lags were at sites ADCP1 and ADCP7.


In contrast, there was no consistent relation between the diel cycles of the horizontal currents and RB. The timing between RB and water currents near-surface was between a 10-hour lead at site ADCP5 and a 9-hour lag at site ADCP3 (average R=0.78; all p<0.0002). Similarly, the correspondence between water currents and RB also varied widely near-bottom, ranging from a 1-hour lead at site ADCP5 to a 9-hour lag at site ADCP7 (average R=0.65; all p<0.0005 except site ADCP6 where R=0.21 and p=0.14). 


The best correspondence and smallest phase difference occurred between the diel cycles of ΔT° and RB near-surface. Comparisons of diel cycles of RB and ΔT° showed little or no phase difference near-surface (RB lagged ΔT° by 0–1 hour; average R=0.93; all p<0.0001). Near-bottom, RB lagged ΔT° by 8–14 hours (average R=0.87; all p<0.0001).


The RB was most uniform through the water column in the morning, coincident with minimum temperature stratification, and most surface-intensified in the afternoon, coincident with the maximum temperature stratification. The diel cycle in RB was more closely aligned to the timing of ΔT° and, to a lesser extent, to diel wind cycles than to the timing of the maximum and minimum in currents. Daily surface intensification of RB was consistent with the accumulation of buoyant particles, but the diel cycle was about 12 hours out of phase with the theoretical light-driven movement of a buoyant cyanobacteria, in which colonies increasingly rise as darkness progresses and increasingly sink as light progresses. (Times of minimum and maximum RB near-surface were about 8:00–9:00 h and 15:00–19:00 h, respectively.)


At subseasonal time scales, correlations between filtered RB and water currents were positive through most of the water column at all but site ADCP3, the shallowest ADCP site (fig. 4). At sites ADCP1, ADCP3, ADCP5, and ADCP7, the correlations decreased away from the bottom and (except at ADCP5) became negative within the upper bins, indicating a dispersal of mass away from the surface to deeper in the water column and, possibly, increased resuspension of bed material with increased water current speeds. The negative correlations near-surface were consistent with a near-surface mixed layer of approximately 1 m at sites ADCP1 and ADCP7, and closer to 1.75 m at site ADCP3, in which mass decreased as water current speeds increased. The lack of negative correlations in upper bins at site ADCP5 may indicate a thin surface mixed layer within the zone that was unmeasured by the ADCP (< 0.5 m). The correlations at site ADCP6 were weak with little vertical variation, suggesting that the relation between RB and currents may be a function of advection rather than entrainment.


Correlations between filtered RB and air temperature at subseasonal time scales (fig. 5) were largely a reverse of the correlations with currents (fig. 4). To some extent, this results from negative correlations between water current speeds and air temperature (weather systems that are accompanied by lower than average air temperatures tend to be associated with stronger than average winds and resulting wind-driven currents). The correlations displayed in figure 5 also are consistent with the effect that air temperature has on water column mixing. Increased water temperature in the upper water column (from warmer air temperature) increases the resistance to mixing; colder air has the opposite effect. Negatively buoyant particles settle out of the water column and positively buoyant particles rise to the surface when mixing is suppressed. Correlations between RB and air temperature were negative through most of the water column at all sites except ADCP7, where the correlation was positive in the upper one-half of the water column (fig. 5). The negative correlation decreased moving away from the bottom and became positive nearer the surface at sites ADCP1, ADCP3, ADCP6, and ADCP7, suggesting the accumulation of mass in the upper water column as air temperature increased. This is consistent with the near-surface correlations between RB and water current speeds (fig. 4) at sites ADCP1, ADCP3, and ADCP7. 


First posted March 16, 2011