A new 240 m resolution Wash tide-surge model, nested within the Holderness grid, was set up for use in collaborative projects (LOIS Special Topic 146, with the NERC Environmental Systems Science Centre (ESSC) and Inter-Tidal Digital Elevaion Models Using Satellite Data (INDUS), a BNSC "LINK" project with ESSC and National Remote Sensing Centre Ltd. (NRSC)). The aim is to construct digital elevation models (DEMs) describing the topography of the inter-tidal zone. The model provides water levels along geo-located shorelines defined by image analysis from SAR scenes at different tidal states. Interpolation from the "pseudo contours" produces the DEM. The accuracy of the first Wash DEM (Figure 1a) is about 20 cm. This depends on both model heights and precision of shoreline location. Differences between inter tidal area in the DEM and in the Wash model (Figure 1b) suggest that the DEM can improve the initial model bathymetry and, with the finer resolution, reduce water level errors.

account for the tidal variations in Hs observed at N2 as interactions were not included.

A scheme was established to nest SWAN, a nearshore wave model being developed by Delft University of Technology, into WAM. The work contributes to PROMISE and to the aims of the "Waves in Shallow Environments" (WISE) group.

Sediment transport modelling progressed with application of a single point vertical (1DV) turbulent energy model with k-e closure, developed in PROMISE, to predict accurately the vertical structure of currents, turbulence and concentrations of suspended particulate matter (SPM), and a 2D model representing advection, diffusion, erosion and deposition to predict the horizontal distribution of fine grained SPM (< 200 µm) on the 1.2 km grid, contributing to LOIS, PROMISE and CAMELOT. Both models link with the 2D hydrodynamic model. The 1DV model reproduced near-bed currents observed about 10 km offshore, and described the balance of terms in the turbulent kinetic energy (tke) equation, varying with height above the bed (Figure 3). At 3 m (d) production and dissipation dominate, with production exceeding dissipation, whereas at 13 m (c) dissipation and produciton are O(10) times smaller, with dissipation exceeding production. The difference is supplied by upward diffusion of tkefrom the bed.

The 2D SPM model may be either coupled directly with the hydrodynamic model or, more efficiently, driven by stored hourly data from it. Input of monthly mean fresh water flows from the Ouse and Trent provided more realistic conditions in the Humber Estuary for a re-run simulation of October 1994 - March 1995 with hydrodynamic and SPM models.