UNC-CH SABSOON NOPP PROJECT

Participants: Cisco Werner, Brian Blanton, Alfredo Aretxabaleta

PROVISIONAL FREQUENCY-DOMAIN (STEADY) RESULTS

Date: 25 July, 2000

Initial Tests of MICOM TS fields in FUNDY5sp

Temperature and salinity fields from an isopycnal coordinate general ocean model are used to provide a sigma-t 0 of FUNDY5sp. The Miami Isopycnal Coordinate Model (MICOM) configured for the Atlantic ocean at 1/12th degree resolution provides TS fields for the MICOM domain for each of years 16-20 of a 20-year MICOM integration. We convert from the vertical MICOM isopycnal coordinates to z-coordinates using a code provided by MICOM (Dr. Z. Garraffo).

Next, we optimally interpolate (using OAX5.0) temperature and salinity from the (x,y,z) coordinates onto constant-z levels required for FUNDY5sp. The levels are

{0,10,20,30,40,50,60,70,80,90,100,150,200,250,300,350,400,500,
                 600,700,800,900,1000,2000,3000,4000,5000} meters.

Horizontal correlation scales are set stronger along-isobath than cross-isobath, and vertical correlations vary with total water-column depth.

Better versions of the following plots will be forthcoming. You may need to make the netscape window as big as possible.

Initial Comparison Plots

First, the MICOM Surf Sal and Temp, from the 16-20 year average, are shown in the left column below, with the corresponding OAXed Sal and Temp onto the SABRE1 FEM domain in the right column. Interesting points: the temperature "tongue" at about 31deg lat and 20degC (this feature is presumably present in the monthly COADS climatology used to force MICOM), and the cooler water in the lower Mid-Atlantic Bight.

The upper layers of the oaxed sigma-t field (computed from the oaxed t,s using the SEAWATER UNESCO eqn of state, at a reference level of 1000db):

FUNDY5sp Results

This plot shows the F5sp surface velocity response to the above described sigma_t field generation. The south boundary (upstream relative to the Gulf Stream or downstream relative to Kelvin wave propagation) condition is geostrophic. The elevation is prescribed along the east and north open boundaries. This elevation is set as per the "Mary Joe" compensation method. The vector scale is 100 cm/sec. More aggressive local smoothing in OAX might solve some of this. The surface elevation is in general consistent with the MICOM 16-20 year averaged surface topography, shown on the right. This indicates to some extent that this initial process of taking the MICOM layer-averaged TS fields to the level surfaces needed by FUNDY is not completely unreasonable. The range of the colormap for the MICOM surface topography is (blue,red) -> (-1.1 , .72) meters.

Caveats:

All of the above results are preliminary; we really just wanted to see what would happen. Nonetheless, these are quite interesting. Clearly, the sensitivity of the optimally interpolated fields to correlation scales needs to be investigated, along with examination of the lower level F5sp response, which really means looking closely at the interpolation between the MICOM layers and level surfaces. Our procedure is not optimal, efficient, or (probably) entirely correct.