Table S1. Correlation (based on Scully (2010b) method,) of July hypoxic volume (DO<1 mg/l) with May-July wind and winter-spring flow for data in the full period (1950-2007 or 1950-2001) and two split periods (1951-1984 and 1985-2001) .

Period of data used / r-values for the variables correlated with hypoxic volume / Note
N-Load / Flow / SE wind / S wind / W wind / Wspd
1950-2007 (n=58) / 0.44 / 0.24 / -0.48 / -0.34 / 0.71 / -0.04 / From Scully (2010b), Table 1.
1950-2001 (n=52) / N/A / 0.27 / -0.51 / -0.08 / 0.58 / -0.09 / Using July hypoxic volume from Hagy et al. (2004), and May-July wind.
1950-1984 (n=35) / N/A / 0.43 / -0.59 / -0.43 / 0.36 / 0.08
1985-2001 (n=17) / N/A / 0.60 / -0.002 / 0.26 / 0.25 / 0.05

Table S2. Correlation of Susquehanna flow with stratification strengths in fixed lag-days, using values in specific dates that had Bay-wide monitoring events (1985-2012) (n ~ 500) *

SS / Stratification strength (SS), in lag days after flow
No lag / 1-day / 2-day / 4-day / 30-day / 60-day / 90-day / 120-day / 150-day
0.217 / 0.228 / 0.240 / 0.272 / 0.329 / 0.348 / 0.178 / 0.217 / 0.002

* Note: There are about 500 measurements of stratification at CB4.1C in 1985-2012. Flows are selected on the days matching the stratification measuring dates to perform correlation using day-to-day values. When applying lag time, the dates for flow were shifted ahead with the lag days.

Table S3. Correlation of Susquehanna flow with stratification strengths in fixed lag-months, using monthly averaged flow and stratification strength (1985-2012) (n=28 x 12)

SS / Stratification strength (SS), in lag time after flow
No lag / 1-mon / 2-mon / 3-mon / 4-mon / 5-mon
0.288 / 0.496 / 0.461 / 0.408 / 0.292 / 0.056

Note: please refer to the note of Table 3. Here, the data are monthly average from the data used in Table S2.

Table S4 Correlation between average winter-spring (Jan-May) flow from Susquehanna and the stratification strengths in the subsequent months in the years (1985-2012) (n=28)

Stratification strength in individual months
April / May / June / July / August / September
0.288 / 0.411 / 0.470 / 0.282 / 0.055 / 0.147

Table S5. Correlations of anoxic volume (AV) with its influencing constituents based on 1991-2005 data (n=15)

Jan-May average / July or Summer average
ENP / SNP / flow / Wspd / Cwd4 / SS / Wtmp / Atmp
AV_
_Obs / Summer / 0.83 / 0.89 / 0.89 / -0.37 / -0.42 / 0.50 / -0.15 / 0.33
July / 0.77 / 0.71 / 0.82 / 0.20 / -0.36 / 0.42 / -0.20 / 0.17
AV_
_Mod / Summer / 0.81 / 0.77 / 0.78 / -0.53 / -0.48 / 0.39 / -0.29 / 0.36
July / 0.85 / 0.78 / 0.85 / -0.66 / -0.54 / 0.40 / -0.24 / 0.34

* Based on observed data, the correlation of winter-spring nutrient load (SNP) with July anoxic volume is weaker than the correlation of SNP with summer anoxic volume. This is mainly due to more variability of July wind’s effect (comparing to summer average) in the scattered monitoring events. If we had DO monitoring in everyday, it is expected the correlation of SNP with the averaged July anoxic volume or summer anoxic volume would be similar. The modeled anoxic volume is averaged from every day, the aforementioned variability is reduced, thus July anoxia and summer anoxia have a similar correlation with SNP. This table is used to emphasize that the apparent low correlation of July hypoxia with load is due to the spared observation frequency. This conclusion would just be a guess if without the model support.

Fig. S1. An example of high-frequency DO observations (mg/l, middle panel), and the associated wind speed (m/s, bottom panel) and salinity (psu, top panel). Modified from Wang and Wang (2012).

Fig. S2. Another example of high-frequency DO observations (circles, in mg/l), and the associated wind speed (dots, in m/s). Modified from Wang and Wang (2012)

Fig. S3. Daily flow from the Susquehanna River to the Bay for 1985-2012;

Fig. S4. Stratification strength (the N2 value) in CB4.1C in 1985-2012 on the days having monthly or semi-monthly monitoring events.

1