APPENDIX A
Physical Characterization
Appendix A: Physical Characterization December 2007
A. PHYSICAL CHARACTERIZATION
A.1 Physical Oceanography of Massachusetts and Cape Cod Bays
A.1.a General Summary of Processes Massachusetts and Cape Cod Bays are subject to the combined influence of atmospheric forcing (wind stress, heat flux, and precipitation), river inflows (both direct and remote), and boundary forcing of tidal flows, storm surges, and currents of the Gulf of Maine—in particular the Western Maine Coastal Current (WMCC) (Brooks 1985; Brown and Irish 1992; Geyer et al. 2004). Temperature variations are mainly due to surface heating and cooling, following the seasonal cycle of the air temperature. Salinity is mainly influenced by the river inflows, particularly the Merrimack and the Charles Rivers. The water properties in Massachusetts Bay are also influenced by the conditions in the Gulf of Maine—in fact it is instructive to think of Massachusetts Bay as a small “arm” of the Gulf of Maine rather than a distinct water body. This is illustrated in particular by the dissolved oxygen variations in Massachusetts Bay, which very closely track the variations in dissolved oxygen of the adjacent waters of the Gulf of Maine (Geyer et al. 2002).
Currents within Massachusetts Bay are generally on the order of 10 cm/s (or 8 km/day) (Butman, 1978), with stronger currents near the mouth, particularly in the vicinity of Race Point to the south and Cape Ann to the north. There is a general counter-clockwise circulation in Massachusetts Bay (Geyer et al. 1992), although the mean flow becomes weak in western Massachusetts Bay, and most of the flow there is due to tidal and fluctuating, wind-forced motions. The predominant wind-forced motions are upwelling and downwelling currents. Upwelling is caused by southerly winds, most typically during summer months. The surface currents are directed offshore due to the Coriolis effect acting on the wind-induced motions (Ekman transport; Csanady 1982). This causes the warm surface waters to be advected offshore and replaced by cooler waters that have upwelled from below the thermocline. Downwelling is the other important type of wind-forced motions. It is most strongly driven by northeasterly winds, as it sets up an along-coast flow between Cape Ann and Boston. During the spring, northeasterly winds may advect low-salinity water from the GMCC (Butman 1972), enhancing the circulation in Massachusetts Bay and potentially advecting harmful algal blooms into the bay (Anderson et al. 2005a). Downwelling is also associated with strong vertical mixing. Both upwelling and downwelling may contribute to increased productivity by bringing nutrients into the surface layer, either by advection (in the case of upwelling) or mixing (in the case of downwelling).
A.1.b Physical Dynamics and the Outfall The fate of effluent from the outfall depends on the stratification conditions and the regional current pattern. Stratification persists from May through October—this causes the trapping of the outfall plume below the pycnocline. During the unstratified winter months, the outfall plume mixes through the whole water column, with roughly twice the initial dilution as during the stratified months. The transport and dispersion of the effluent away from the outfall occurs due to a complex combination of tidal, wind-driven and density-driven motions. The dispersion is relatively rapid, rendering the effluent signal indistinguishable from ambient water within 10-15 km from the outfall site. No particular conditions have been identified that would significantly increase the residence time of the effluent. Thus, the main importance of the physical forcing is to affect the physical and biological environment of the receiving waters.
Appendix A: Physical Characterization December 2007
A.2 Forcing Conditions
A.2.a Freshwater run-off River discharge influences salinity, stratification, and strength of the coastal circulation. The Charles River is the largest river feeding directly into Massachusetts Bay, and its discharge is correlated with surface salinity at the outfall site. The Merrimack River enters the Gulf of Maine just north of Massachusetts Bay, but it is a much larger source of fresh water than the Charles River. Its variation is correlated with both surface and bottom salinity in the nearfield. The year 2006 was another wet year, like 2005. In fact, there was near-record discharge on the Merrimack River around May 15 and it was the largest spring discharge on the Merrimack for the entire measurement period (Figures A-1 and A-2; Table A-1). The Charles River was wetter than normal, but not to the extreme extent of the Merrimack River. Although the flood of May 15 was devastating to the riparian towns in the lower Merrimack Valley, it only had a minor influence on currents in Massachusetts Bay, unlike the major event in 2005. This is apparently related to the timing of the winds relative to the freshwater inflow, as discussed in Section A.6. Wet conditions continued through the rest of the year, although not in as dramatic fashion as the spring.
A.2.b Wind Forcing The most important aspect of the wind forcing is the average north-south component of wind stress, which determines the preponderance of upwelling or downwelling conditions in western Massachusetts Bay. Upwelling provides flushing of bottom waters and causes colder water temperatures, which usually leads to higher near-bottom dissolved oxygen. The upwelling index is shown in Table A-2 and Figure A-3. The most notable features of the 2006 wind forcing were strong downwelling conditions in the winter and spring and strong upwelling in July. On a seasonal basis, winds speeds were close to the long term averages in 2006 (Table A-3).
May is normally transitional between winter downwelling and summer upwelling favorable conditions, and so the net north-south wind stress is typically close to zero. In 2006, there was strong downwelling during the month of May, although not as strong as 2005. The spring downwelling was associated with several late-season northeasterlies. As in 2005, they influenced the inflow from the Gulf of Maine importing fresh water and potentially harmful algal blooms from the Gulf of Maine (Anderson et al. 2005a). Northeasterlies also cause the largest waves and potentially the most significant wave-induced bottom resuspension. In contrast to the strong downwelling conditions in May, the July conditions were favorable for upwelling and may have provided an additional nutrient supply to the euphotic zone in July to support the summer diatom bloom that was observed during survey WN069 (see Appendices B and C). An analysis of the influence of prevailing downwelling and upwelling conditions on the currents and physical structure of the water column are discussed in Section A.6.
A.2.c Air Temperature Air temperature has a significant effect on water properties during the winter, when it sets the minimum water temperature. Table A-4 shows the wintertime air temperature for the period of the monitoring program. The winter of 2005-2006 was warmer than average, although not extreme (Figure A-4). May and June were colder than average, due to the northeasterly storms.
Appendix A: Physical Characterization December 2007
Table A-1. Seasonal river discharge (m3/s) summary for the Charles and Merrimack Rivers. (measured at Waltham and Lowell, respectively, by USGS) Year Jan-Mar Apr-Jun Jul-Sep Oct-Dec Charles River Discharge 1990 13 13 7 13 1991 13 7 3 10 1992 10 8 2 9 1993 15 15 1 5 1994 15 11 3 7 1995 11 5 1 7 1996 16 12 4 16 1997 12 13 1 4 1998 21 21 8 7 1999 18 7 4 9 2000 13 16 4 7 2001 14 14 4 2 2002 6 10 1 9 2003 13 17 5 10 2004 9 16 4 10 2005 15 14 3 19 2006 17 18 4 10 mean 14 13 3 9 Merrimack River Discharge 1990 333 366 164 331 1991 289 237 117 295 1992 254 266 100 174 1993 200 393 51 198 1994 253 380 74 164 1995 295 154 45 292 1996 409 487 127 401 1997 296 404 70 123 1998 401 454 122 116 1999 328 175 103 180 2000 292 410 104 160 2001 196 392 55 58 2002 121 307 42 146 2003 235 384 82 366 2004 182 382 128 128 2005 272 517 108 564 2006 395 525 135 342 mean 279 367 96 238
Appendix A: Physical Characterization December 2007
Table A-2. Southerly (upwelling) wind stress. Estimated seasonally averaged stress in Pa x103 at the Boston Buoy Year Jan-Mar Apr-Jun Jul-Sep Oct-Dec 1990 -0.0 1.4 0.8 0.1 1991 -1.6 -0.2 1.0 -4.2 1992 -3.8 -0.4 1.0 -3.4 1993 -4.5 -0.0 1.3 -1.3 1994 -3.5 1.0 0.4 -1.7 1995 -0.1 0.0 -0.0 -0.9 1996 -2.8 0.5 -0.2 -1.3 1997 -0.1 -0.8 0.5 -2.2 1998 -4.3 -0.8 0.9 -0.5 1999 -2.1 -0.2 0.7 -0.9 2000 -3.3 0.0 -0.1 -2.6 2001 -4.6 -0.3 0.6 -0.1 2002 0.5 0.2 -0.3 -2.7 2003 -2.2 -1.7 1.2 -1.4 2004 -4.4 -0.6 -0.1 -2.9 2005 -5.1 -1.8 0.5 -2.6 2006 -3.8 -1.2 0.6 -1.2 mean -2.7 -0.2 0.5 -1.8
Table A-3. Seasonally averaged wind speed (m s-1) at the Boston Buoy
Year Jan-Mar Apr-Jun Jul-Sep Oct-Dec 1990 7.0 5.8 4.4 7.9 1991 7.6 5.8 5.3 7.5 1992 7.9 5.8 5.1 7.0 1993 7.7 5.8 4.9 6.9 1994 7.4 5.9 5.6 6.8 1995 6.6 4.6 4.6 7.2 1996 7.3 5.1 4.5 6.6 1997 7.6 5.3 5.1 6.6 1998 6.9 4.6 3.9 6.8 1999 7.3 4.5 4.3 6.8 2000 7.3 5.4 4.6 7.2 2001 7.1 4.5 4.2 6.4 2002 6.9 5.4 4.6 7.8 2003 7.5 4.8 4.0 7.1 2004 7.4 4.8 4.2 7.0 2005 7.0 4.9 4.2 7.2 2006 7.5 5.3 4.4 6.7 mean 7.3 5.2 4.6 7.0
Appendix A: Physical Characterization December 2007
Table A-4. Average winter air temperature (°C) at the Boston Buoy, 1993-2006.
Year Dec 1 - Feb 28 1992-1993 -0.4 1993-1994 -1.4 1994-1995 1.7 1995-1996 -0.4 1996-1997 2.3 1997-1998 2.6 1998-1999 2.2 1999-2000 0.8 2000-2001 0.0 2001-2002 3.6 2002-2003 -0.9 2003-2004 -0.8 2004-2005 0.6 2005-2006 1.6 mean 0.8
A.3 Water Temperature Structure and Variability The continuous time series of near-surface water temperature in the vicinity of the bay outfall for 2006 (Figure A-5) show warmer than normal conditions in the winter, but colder than normal in the spring and early summer. The cold temperatures and wind-induced mixing during the northeasterlies kept the surface from warming as fast during May and June. Upwelling during the early summer also resulted in episodic cooling.
A combined plot of wind stress and surface temperature for the late spring and summer (Figure A-6) indicates cooling events due both to wind-induced mixing and to upwelling. The first cooling event indicated (June 15) is due to a late-season northeasterly, which resulted in cooling due to vertical mixing and surface heat loss. The other two events (June 29 and July 3) were due to upwelling, which brought cold water to the surface via vertical and lateral advection rather than mixing. This persistent upwelling may have provided nutrient fluxes for an anomalous plankton bloom in July, 2006.
Figure A-7 shows the near-surface and near-bottom temperature data obtained through the entire monitoring program from the shipboard surveys. In 2006, the water column remained mixed for longer than normal, due to the spring northeasterly mixing events and summertime surface waters were cooler than average, due to the upwelling-favorable conditions.
A.4 Salinity Structure and Variability The salinity data in 2006 showed the influence of the high runoff conditions in both near-surface and near-bottom salinity (Figure A-8). Surface salinities reached the lowest values since 1998. Bottom salinities were lower than normal, but not as low as 2005. This might be because there was less mixing during the large freshwater flow events in 2006 compared to 2005 (see discussion in A.6). The low salinity during the spring caused stronger than average stratification in May and June 2006, as shown in Figure A-9. The stratification was approximately 50% stronger than normal during Appendix A: Physical Characterization December 2007
May and June, due to the increased freshwater inflow. The stratification was weaker than normal during July and August, due to the prevalence of upwelling.
A.5 Dissolved Oxygen Structure and Variability The near-bottom dissolved oxygen concentrations were lower than normal between April and September, nearly the lowest of the outfall monitoring program (Figure A-10). The drop in DO during April seems to be the preconditioning event for 2006 that causes the low DO levels, because for the rest of the summer it follows a typical decrease. Ventilation during late September and early October causes the DO levels to return to average levels by the time of the usual annual minimum.
A regression model was developed (Geyer et al. 2002) to relate the interannual variation of the dissolved oxygen minimum in the bottom water of the nearfield to the variations in temperature and salinity (Figure A-11). The model explains about 60% of the variance of dissolved oxygen, using the regression coefficients established from the data of 2002 and earlier. The main purpose of the model is to identify the variations that result from the natural variability of the environment in order to detect deviations that may be due to the outfall. The 2006 observations indicate that the near- bottom DO levels were slightly above average, whereas the model predicts slightly below-average levels, based on slightly warmer temperatures and higher near-bottom salinities than average.
A.6 Summary of 2006 Physical Conditions
A.6.a The Influence of Meteorological Events on Physical Oceanographic Conditions Both 2005 and 2006 were notable with respect to the frequency of northeasterly storms during the late spring months. In both years, unusually strong southeastward currents were observed during the northeasterly events. The early May events are particularly important because they transport Alexandrium cells into Massachusetts Bay (Anderson et al. 2005a and 2005c). Northeasterlies typically occur during the winter and early spring and their occurrence during the late spring (and summer in the case of 2006) is unusual. Plots of the wind forcing and near-surface current response at the Gulf of Maine Ocean Observing System (GoMOOS)-A Buoy (Figure A-12) indicates the strong southeastward currents during the events in both 2005 and 2006. The current magnitudes of more than 80 cm/s are much stronger than typical wind-forced motions of the western Gulf of Maine, which are more typically 20 cm/s and rarely 40 cm/s. The reason for these unusually strong currents is that the strong northeasterlies occurred late enough in the year that there was significant stratification, mostly due to the freshwater inflow. The stratification suppresses the turbulence with the underlying water, resulting in greater current speeds for a given wind stress than would occur with unstratified or weakly stratified conditions.
The incidence of these late-season northeasterlies is related to the increased freshwater flow of 2005 and 2006, as there was significant precipitation associated with these storms. It was mainly rain, not snow-melt, which caused the high discharge rates in 2005 and 2006. May of 2005 and 2006 both experienced strong negative fluctuations of the North Atlantic Oscillation (based on NOAA’s Climate Prediction Center data). This may be fortuitous—a more in-depth analysis would be required to determine whether the occurrence of late-season northeasterlies is correlated with the NAO.
The oceanographic response of Massachusetts Bay to this anomalous forcing is evident in the salinity, temperature and currents. In both 2005 and 2006, the salinities were significantly lower than average, due to the high river discharge levels. The 2005 conditions differed from 2006 in that the salinity anomaly was mostly confined to the surface waters in 2006, whereas it showed up Appendix A: Physical Characterization December 2007
strongly in both bottom and surface waters in 2005. This difference is probably related to the timing of freshwater flow events relative to the timing of northeasterlies. During 2005, the major event around May 23 occurred in combination with a peak in river outflow, and there was deep mixing of the fresh water, as described in the 2005 Water Column Report. In 2006, the very large discharge event around May 15 occurred with nearly easterly winds (Figure A-12), which did not result in large currents at GoMOOS-A. This is probably because the winds were more perpendicular than parallel to the coast, so the Merrimack River plume was not being accelerated down-coast by the winds during that event. Thus, the particular conditions that caused the deep mixing in 2005 did not occur in conjunction with the strong river outflow in 2006.
A.6.b Regional Dissolved Oxygen Response The GoMOOS time series dissolved oxygen data were compared to the nearfield DO data, to examine the relationship between the farfield and the nearfield DO conditions. Unfortunately, the GoMOOS mooring is not subject to frequent calibration, so the actual observations had to be adjusted based on the farfield data at station F22, and it was not possible to come up with a reliable indication of the interannual variations of DO to compare with the nearfield data. However, the GoMOOS data are particularly useful for documenting the seasonal trends and the influence of short- term forcing processes.
Figure A-13 shows a comparison of the GoMOOS-A 50-m DO data to the nearfield near-bottom observations for the July-December period for 2002 through 2006. The notable feature of the comparison is that the offshore waters show a very similar decrease in DO to the nearfield—a nearly constant decrease of approximately 1 mg L-1 per month. The interannual variations are mainly due to differences in the initial DO levels and the time of reventilation, when the DO level increases sharply. There are slight year-to-year differences in the slope of the nearfield DO to the GoMOOS data; for example in 2002, the nearfield decreased more rapidly than offshore, whereas in 2006 the nearfield showed a lower rate of decrease than offshore. The most likely explanation for a lower rate of decrease, as in 2006, is that there was some reventilation in the shallower waters of the nearfield (in September or October) that did not penetrate to 50-m depth measured at the GoMOOS-A buoy. The higher nearfield rates during other years may be associated with warmer water or possibly with higher consumption rates closer to shore.
The GoMOOS time series illustrate that the reventilation is a sudden process, occurring when fall cooling (due typically to a frontal passage) causes the overlying water to reach the density of the near-bottom water (see the next section for more information about destratification). The temporal resolution of the nearfield surveys was not adequate to determine whether the timing of reventilation was significantly earlier in the nearfield—it would be expected to be somewhat earlier due to the shallower water.
In summary, these data confirm the observation noted by Geyer et al. (2002) and reported in previous water column reports (e.g. Libby et al. 2005a and 2005b) that the dissolved oxygen variation is mostly associated with regional variability rather than local variations associated with the conditions at the outfall. There are slight variations at the outfall from the regional patterns, but the interannual variations are mostly associated with the regional signal.
A.6.c Stratification Variations and Consequences The unusually strong stratification observed in 2006 provides motivation for taking a closer look at the seasonal variations of stratification in Massachusetts Bay, to see whether it may shed light on the variability of other quantities. In order to quantify the variations of stratification, the continuous CTD cast data from the water column surveys were contoured as a function of time (Figures A-14 Appendix A: Physical Characterization December 2007
and A-15). These data provide 0.5-m vertical resolution, yielding much more detail about the vertical structure than the discrete-depth data. The temporal separation of the surveys does not fully resolve the variability (as shown by the GoMOOS data), but the seasonal trends and even some of the shorter term variability are resolved.
The variations of the nearfield (station N16) density and dissolved oxygen are shown for the years 2002-2006 in Figure A-14 and the conditions in the farfield (station F22, near the GoMOOS-A buoy) are shown in Figure A-15. The nearfield data indicate that the onset of stratification varies slightly from year to year, but it typically is established by around the first of April. A period of intense stratification usually commences around the first of June. The near-surface stratification was particularly strong in 2005 and 2006, due to the large freshwater inputs during those years. As the season progressed (June-August), the stratification remained strong in the near-surface waters (0-15- m depth), but there was weaker stratification in the near-bottom waters than in previous years. this shallow stratification may have implications for the plankton productivity, as suggested by the dissolved oxygen data.
The dissolved oxygen shows strong sub-surface maxima during both late May/early June 2005 and July 2006 (Figure A-14). These maxima are a likely indication of high rates of photosynthesis in the sub-surface chlorophyll maximum. The relatively shallow pycnocline during these periods allows the bottom water nutrients to extend higher in the water column, so that although they do not reach the surface layer, they reach a level with sufficient light to support growth. This shallow pycnocline may provide the explanation for the diatom bloom observed in the summer of 2006.
The late-season development of the low-dissolved oxygen zone is well resolved by the timeseries contour plots. The lowest DO occurs around the beginning of October, just before the vertical ventilation associated with fall cooling and mixing events. In 2005 and 2006, the first fall storms did not penetrate to the bottom, so the low DO conditions persisted into November at station N16.
The offshore data at station F22 show a similar pattern as the nearfield (Figure A-15), with some interesting differences. Surprisingly, the onset of stratification tends to be earlier in the offshore waters, perhaps due to the influence of the Merrimack River plume. Also, the stratification tends to be stronger near the surface, again because of the stronger riverine influence. The dissolved oxygen data show a persistent sub-surface maximum, probably as a consequence of the shallow pycnocline. This was expressed most intensely during the summer of 2005 (although note that DO data are not available for the ARRS surveys conducted during the same period of 2006). The high productivity in 2005 cannot be solely explained by the stratification in this case, because it was not too different from 2003 and 2004. It is possible that the strong mixing that occurred during May 2005 provided the required nutrient flux to support the subsequent productivity.
The subsurface dissolved oxygen maximum is a potentially valuable tool for diagnosing seasonal and interannual variations in plankton productivity. Further comparisons are warranted with the chlorophyll and plankton data.
Appendix A: Physical Characterization December 2007
Figure A-1. River discharge at the Merrimack River (Lowell gauge) and the Charles River (at Waltham), from 1992 through 2006 (data from USGS). Red lines indicate three-month moving averages.
85th th st th percentile 97 91 79
st th th 85 91 th 68 percentile 74
Figure A-2. Comparison of the 2006 discharge of the Charles and Merrimack Rivers (red curve) with the observations of the past 16 years (light blue lines). Appendix A: Physical Characterization December 2007
Figure A-3. Monthly average N-S wind stress at Boston Buoy for 2006 (red line) compared with the previous 12 years of observations (1994-2005; light blue lines). Positive values indicate northward-directed, upwelling-favorable wind stress.
Figure A-4. Hourly air temperature (°C) for 2006 at the Boston Buoy (Black) superimposed on the data from the previous 17 years (1989-2005; light blue). Appendix A: Physical Characterization December 2007
Figure A-5. Hourly near-surface temperature (°C) for 2006 at the Boston Buoy (black) superimposed on the data from the previous 17 years (1989-2005; light blue).
Figure A-6. N-S wind stress for late spring and summer (top panel), and near-surface temperature at the Boston Buoy (bottom panel). Three cooling events are indicated with vertical lines: the first is a downwelling event that causes cooling due to vertical mixing; the other two are upwelling events that cause cooling by vertical and lateral advection.
Appendix A: Physical Characterization December 2007
Figure A-7. Timeseries of near-surface (blue) and near-bottom (green) temperature in the vicinity of the bay outfall (averaging the data from nearfield stations N16, N18 and N20).
Figure A-8. Timeseries of near-surface (blue) and near-bottom (green) salinity in the vicinity of the bay outfall (averaging the data from nearfield stations N16, N18 and N20). Appendix A: Physical Characterization December 2007
Figure A-9. Stratification near the outfall site (nearfield stations N16, N18 and N20) for 2006 (dark blue) compared with the previous 14 years of observations (1992-2005; light blue lines).
Figure A-10. Dissolved oxygen at near-surface and near-bottom near-field stations (N16, N18 and N20) for 2006 (dark blue) compared with the previous 14 years of observations (1992-2005; light blue lines). Appendix A: Physical Characterization December 2007
Figure A-11. Upper panel: Average near-bottom dissolved oxygen in nearfield (stations N16, N18 and N20) during September-October, compared with linear regression model based on temperature and salinity variation. Lower panel: The bar plot shows the individual contributions due to temperature and salinity for each of the years.
Appendix A: Physical Characterization December 2007
Figure A-12. N-S wind stress (at the Boston Buoy) and near-surface currents at the GOMOOS-A buoy during May and June of 2005 and 2006.
Appendix A: Physical Characterization December 2007
Figure A-13. Timeseries of dissolved oxygen measured at the GoMOOS-A buoy, 50-m depth (thin, continuous lines), compared with the nearfield, near-bottom DO (average of stations N16, N18 and N20), for the July-December period from 2002 to 2006. The GoMOOS data were adjusted based on the farfield station F22 to remove sensor offsets and drift (typically 1-2 mg/l errors). Appendix A: Physical Characterization December 2007
Figure A-14. Contours of density (left panels; kg/m3) and dissolved oxygen (right panels; mg/l) vs. time and depth for the last 6 years of the Outfall Monitoring Program at nearfield station N16. The times of profiles are indicated by the ticks at the tops of the figures.
Appendix A: Physical Characterization December 2007
Figure A-15. Contours of density (left panels; kg/m3) and dissolved oxygen (right panels; mg/l) vs. time and depth for the last 6 years of the Outfall Monitoring Program at the farfield station F22. The times of profiles are indicated by the ticks at the tops of the figures.