Heat and salt balances over the New England continental shelf, August 1996 to June 1997

Abstract

Heat and salt balances over the New England shelf are examined using 10 month time series of currents, temperature, and salinity from a four element moored array and surface heat and freshwater fluxes from a meteorological buoy. A principal result is closure of the heat budget to 10 W m−2. The seasonal variation in depth‐average temperature, from 14°C in September to 5°C in March, was primarily due to the seasonal variation in surface heat flux and a heat loss in winter caused by along‐shelf advection of colder water from the northeast. Conductivity sensor drifts precluded closing the salt balance on time scales of months or longer. For time scales of days to weeks, depth‐average temperature and salinity variability were primarily due to advection. Advective heat and salt flux divergences were strongest and most complex in winter, when there were large cross‐shelf temperature and salinity gradients at the site due to the shelf‐slope front that separates cooler, fresher shelf water from warmer, saltier slope water. Onshore flow of warm, salty slope water near the bottom and offshore flow of cooler, fresher shelf water due to persistent eastward (upwelling‐favorable) winds caused a temperature increase of nearly 3°C and a salinity increase of 0.8 in winter. Along‐shelf barotropic tidal currents caused a temperature decrease of 1.5°C and a salinity decrease of 0.7. Wave‐driven Stokes drift caused a temperature increase of 0.5°C and a salinity increase of 0.4 from mid December to January when there were large waves and large near‐surface cross‐shelf temperature and salinity gradients.

Seasonal warming of the Middle A tlantic B ight C old P ool

“…The minimum temperature in the Cold Pool increases steadily from April to September at a rate of 1°C month −1 over the New England shelf. The Cold Pool disappears in fall (October–November) as vertical mixing from increasing storm activity and decreasing surface heat flux destroys the seasonal thermocline [ Lentz et al ., ].…”

“…The dependence of warming rate on Cold Pool thickness is consistent with (2) (black dashed lines in Figure ) with larger heat fluxes through the thermocline over Georges Bank (∼100 W m −2 ) and the New England shelf than in the New York Bight and southern MAB (∼30 W m −2 ). The inferred heat flux through the thermocline over Georges Bank and the New England shelf is a substantial fraction of the net surface heat flux through the sea surface which ranges from 100 W m −2 in spring to a maximum of 200 W m −2 in July and August [ Beardsley et al ., ; Lentz et al ., ; Lentz , ; Fewings and Lentz , ; Georgas et al ., ]. The inferred heat flux through the thermocline in the southern half of the MAB is a smaller fraction of the net surface heat flux.…”

“…throughout the year and consequently is always warmer than the near-bottom shelf water. The Cold Pool begins to emerge in April, as decreasing storm activity and increasing surface heat flux lead to the development of the seasonal thermocline [e.g., Lentz et al, 2010]. The thermocline strengthens from May to August.…”

Mapping the Mid-Atlantic Cold Pool evolution and variability with ocean gliders and numerical models

“…The term Cold Pool continues to be used to describe this dynamic feature [13] [4] [2] [3], which is characterized by southwestward advection with an accompanying gradual increase in the heat and salt fluxes through the surface and lateral boundaries [12] [14] [15]. It has been shown that the Cold Pool affects phytoplankton productivity [16] [9] and the behavior and recruitment of pelagic and demersal fish on the shelf [21] [24].…”

Optimal Asset Distribution for Environmental Assessment and Forecasting Based on Observations, Adaptive Sampling, and Numerical Prediction

“…The models produce dynamically consistent output on a regular grid, which greatly facilitates the computation of the mean and eddy fluxes. The basic methodology being used is similar to that used during the CODE experiment [Lentz, 1987;Lentz et al, 2010]. The surface flux computations are complete (see below).…”