Today is a special day. Finally, I have DO results from LiveOcean!

Recall that this test implements WWTPs as tiny rivers. There is a Baseline condition which models existing nutrients conditions, and there is a No WWTP DIN condition that has zero nitrate and ammonium in WWTP effluent.

When I first started working with the model output this week, the No WWTP DIN condition had only finished running through September 15. Thus, the comparison plots below only show difference timeseries through September 15. The colormaps show average bottom DO from September 1 - September 15. Note that all data were extracted on a daily frequency (hour 20 of each day).

Figure 1 presents the original plot I had sketched up for the poster outline.


Fig 1. DO comparison plot. Panel (a) presents a bottom DO timeseries in Penn Cove from the Baseline. Panel (b) shows the difference in bottom DO at Penn Cove between the No WWTP DIN condition and the Baseline run. Panels (c) and (d) are analagous to (a) and (b), respectively, except for Lynch Cove. Panel (e) shows the average bottom DO in Puget Sound between September 1 - September 15. The locations of the Penn Cove and Lynch Cove timeseries are highlighted. Panel (f) shows the difference in average bottom DO between the No WWTP DIN condition and the Baseline condition.


There are several observations that stand out to me. First, the only standout hypoxic zones I see in the Baseline run (Fig. 1e) are in Lynch Cove and Holmes Harbor (South Whidbey Basin). Several of the inlets in South Puget appear rather well oxygenated. Based on the Penn Cove (Fig. 1a - 1b) and Lynch Cove (Fig. 1c - 1d) timeseries, there does not seem to be a large difference in bottom DO between the No WWTP DIN run and the Baseline run. At these locations, the bottom DO difference varies between -0.05 and 0.05 mg/L. From Figure 1f I see that there are not many regions where bottom DO differs by much. The two locations that stand out most are Dabob Bay and Port Orchard Harbor. In Figure 2 below, I have re-made Figure 1, except with Dabob Bay and Port Orchard Harbor timeseries.


Fig 2. DO comparison plot. Panel (a) presents a bottom DO timeseries in Dabob Bay from the Baseline. Panel (b) shows the difference in bottom DO at Dabob Bay between the No WWTP DIN condition and the Baseline run. Panels (c) and (d) are analagous to (a) and (b), respectively, except for Port Orchard Bay. Panel (e) shows the average bottom DO in Puget Sound between September 1 - September 15. The locations of the Dabob Bay and Port Orchard Bay timeseries are highlighted. Panel (f) shows the difference in average bottom DO between the No WWTP DIN condition and the Baseline condition.


Based on Figure 2a - 2b, it seems like the No WWTP DIN condition tends to have lower bottom DO in Dabob Bay than the Baseline condition. This difference is the opposite of what I would expect. In contrast, the No WWTP DIN conditions tends to always have higher bottom DO in Port Orchard Bay than the Baseline condition. It’s strange to me that the difference between the two runs at Port Orchard Bay is so much larger than the difference anywhere else. One suspicion I have is that the LiveOcean grid does not appear to include Agate Passage, which connects Port Orcahrd Bay northwards to Port Madison. Could Agate Passage help with flushing in Port Orchard Bay? Or perhaps these results suggest that Port Orchard Bay is indeed more sensitive to the presence of WWTP DIN. It’s also weird that in Figure 2d, there appear to be times in which the test conditon and the basline condition have the same bottom DO, and these times occur with regular periodicity.

This analysis is still new. I am looking forward to exploring more question, working with a full year of model output, and analyzing DO on different spatial and temporal scales.