Water-level measurements obtained offshore of Mattituck Inlet from

19 September to 8 October 2002 drove the model. It was found by numerical

experimentation that raising the water-surface elevation in this data set by 0.25 m

produced a successful model simulation that accurately represented the tidal

signature recorded at Goldsmith Pond, while not drying the channel. The authors

have never observed the inlet channel to dry, even during low tide. DYNLET

model water-level calculations are referenced to NAVD88. The results were

adjusted by subtracting 0.25 m to account for the datum shift of the input water-

surface elevation. A shift upward in the driving water level is functionally

equivalent to shifting the entire bathymetry grid down by the same amount.

DYNLET was calibrated by specifying larger values of the bottom friction

coefficient in the Goldsmith Inlet channel, where small rocks are present and can

protrude above the water surface, some of which may be remnants from jetty

construction. The default value of Manning's n of 0.025 m/sec1/3 was maintained

at most nodes, but in the channel where rocks and roiling water are seen, the

value was increased to 0.03 to 0.04. The time-step in the model was set to

30 sec.

In initial model runs, calculated water level at the pond gauge lagged the

measurements by 36 min. A lag between calculations and measurements is

expected, because Goldsmith Inlet lies 5 km east of Mattituck Inlet, and the tidal

wave travels from east to west. The tidal record offshore of Mattituck Inlet was

therefore adjusted forward 36 min to account for the time of tidal wave travel

between the location of the tidal record and the location of the input driving the

model. This adjustment implies that the tidal wave moves westward at about

0.23 m/sec along the shallow water of this portion of the north shore of Long

Island.

The input boundary condition (Node 1) and the first DYNLET water-level

calculation (Node 2) for the period of data collection (20 September 8 October

2002) are plotted in Figure 5-22a. Figure 5-22b shows these water levels for the

5-8 October 2002, spring tide. A comparison of water-level measurements at

Goldsmith Inlet for 20 September 8 October 2002 and the corresponding

DYNLET calculations (at Node 30) are shown in Figure 5-22c, and Figure 5-22d

gives this comparison for a period of spring tide.

Current velocity measurements taken within Goldsmith Inlet for a short

interval on 8 October are compared to corresponding DYNLET current velocity

calculations (at Nodes 13 and 14) in Figures 5-23a and 5-32b. The calculations

well reproduce the limited length of the measurements. The current velocity is

seen to be strong, exceeding 1 m/sec, and the calculated current is flood

dominant, meaning that the flood current has a higher peak velocity than ebb, but

shorter duration. This dominance has implications for sediment transport in the

inlet, discussed in the following paragraphs.

245

Chapter 5 Circulation Analysis

Integrated Publishing, Inc. |