NO A A PROFESSIONAL PAPER 11 



the weather differences seem insufficient to account for 

 the differences in variance across all frequencies. The 

 mooring sites were about 27 km apart; the LT2 mooring 

 site (1976) is about 4 m shoaler than station 49. Some 

 amplification of tidal currents is expected in shoaler water. 

 In addition, the slightly stronger stratification occurring 

 in 1976 (ch. 5) may have enabled internal gravity waves 

 to be excited over a wider band of frequencies and more 

 vigorous internal tides to have been generated. Perhaps 

 more important, the LT2 bottom meter, being somewhat 

 nearer to the surface (23 compared to 27 m), may have 

 introduced more error into measurements in 1976. 



Intefpreting these data to address the cause of devel- 

 opment of anoxic conditions implies selection of a time 

 scale pertinent to the problem. For small (1-10 km) pol- 

 lution problems, the relatively rapidly varying currents 

 would be most important, and 1976 would likely be shown 

 by our data to be a relatively favorable year for pollutant 

 dispersal. Considerations of the oxygen demand on a static 

 environment (Segar and Berberian 1976) or of the ob- 

 served time and space scales of the D.O. distribution in 

 the New York Bight (ch. 2 and 5) suggest that 30 days, 

 or even a sequence of 30-day intervals, is appropriate for 

 this problem. Hence, the data are presented in 30-day 

 blocks. The vector mean flow over these intervals is prob- 

 ably the property of greatest significance here, but we also 

 discuss the total current variance or horizontal kinetic en- 

 ergy to demonstrate the point about variability made 

 above. 



TEMPORAL VARIATION OF CURRENTS 



Extremely low D.O. concentrations were not observed 

 in 1975. Data from stations PI 1 and 49 in the region where 

 very low D.O. concentrations were observed in summer 

 1976 are available for periods up to 90 days, starting from 

 Julian day (J.D.) 65 in 1975. This spans the period of 

 development of the pycnocline during spring (ch. 5). Re- 

 sults from meters located below or in the bottom of the 

 pycnocline during this time are summarized in PVD for- 

 mat (fig. 7-2). These PVDs reveal irregular intervals of 

 weaker or stronger flow, but they also show net monthly 

 displacements of 100 to 120 km to the southwest and south 

 at both stations (Pll and 49B). These data are part of the 

 same data set used by Beardsley et al. (1976) and Hansen 

 (1977) and are consistent in speed and direction with other 

 data acquired at that time from other parts of the shelf. 

 They also generally agree with drift bottle results acquired 

 over several years (Bumpus 1973). Bumpus also sum- 

 marized results from deployment of seabed drifters over 

 several years. Results from drifter studies necessarily are 

 biased toward shoreward movement, and Bumpus" (1973) 

 summary typically indicates a region of diffluence in the 



most critical area off New Jersey during the summer. Re- 

 gional patterns of confluence or diffluence are not well 

 determined from such data, however. 



In addition, the near-bottom direct current meter meas- 

 urements treated are 8 to 9 m above the bottom, and only 

 up to 90 -days of record are available for analysis. How- 

 ever, a considerable amount (nearly 12,000 hours) of data 

 are available from the LTM site (Mayer et al. 1979) 4 to 

 8 m above the bottom, although in water about 15 m 

 deeper and upshelf from the Hudson Shelf Valley. From 

 almost 18 months of these data, it has been estimated that 

 the average or normal flow at this level above the bottom 

 is 1 to 1.5 cm/s toward 220° T. Well within the bottom 

 boundary layer (1 m above the bottom), the picture is not 

 as clear, because current velocities are less than 1 cm/s 

 (with only about 6 months of data available) making it | 

 difficult to reasonably estimate either speed or direction. 

 With this limited data set, addition or deletion of several 

 weeks of data can reverse the direction of the mean, so 

 it is not meaningful to compare in detail our current meter 

 data (up to 90 days in 1975) with Bumpus" seabed drifter 

 results. 



The simple mean circulation described above is greatly 

 complicated by events (mostly of meteorologic origin), 

 some of which persist for as long as 3 months (Mayer et 

 al. 1979). Most events, however, lie within the 3- to 10- 

 day frequency band associated with energetic weather sys- 

 tems. These events are manifest in the many observed 

 upshelf velocities generated by an upshelf component of 

 wind stress. A conceptual model follows; "Upshelf winds , 

 cause offshore flow in the upper part of the water column, | 

 which presumably causes a drop of coastal sea level, pro- 

 viding a pressure gradient that drives a quasi-geostrophic 

 upshelf flow of deeper water"" (Mayer et al. 1979). 



By now it should be clear how variable the circulation 

 is on the Middle Atlantic shelf and how difficult it is to 

 make definitive statements about the mean or normal cir- I 

 culation. With regard to the background material con- 

 sisting of a limited (90 days or less) current meter data set 

 from 1975. we can say that it is consistent with what is 

 believed known about "normal"" circulation in the Middle 

 Atlantic Bight. 



A longer series of current meter observations in the 

 New York Bight, including station LT2 off New Jersey, 

 was obtained from October 1975 through summer 1976. 

 Ten months of continuous record from a current meter 

 located below the pycnocline during the stratified season | 

 was available for the analysis of anoxic conditions off New 

 Jersey. Figure 7-3 shows results from essentially the same 

 level as the 1975 data in figure 7-2 for corresponding time 

 periods. In October 1975 the displacement was about twice , 

 that observed during the preceding spring, but slowed ' 

 dramatically and reversed direction for about a month in 

 November. During the subsequent several months the 



152 



