JERICHO: Coastal applications of satellite altimeter data

INTRODUCTION:
As part of the JERICHO project, it was necessary to develop new applications of satellite altimeter wave data at locations close to the coast. Historically it has been easier to measure waves and estimate wave climate at near-shore sites than in the open ocean, but using the altimeter, the reverse is true. The satellite has problems in making measurements within a few kilometres of the coast, and the wave climate has much greater spatial variability in coastal waters than in the open ocean.

COASTAL PROBLEMS:
If there is any land within the radar footprint then altimeter wind and wave measurements are not possible. This means no data are obtained within about 5 - 10 km of the coast, depending on wave height (if the waves are higher, the altimeter signal has a larger footprint at the sea surface). In addition, if the satellite is travelling from land to sea, then the altimeter has to ‘home in’ on the sea surface. This can take 3 - 6 seconds, during which the altimeter travels 20-40 km. The problem is illustrated in Figure 1 which shows the location of TOPEX records off Lyme Bay. The track approaching the shore from the SW gets within a few km of the coast, while the first location with good data from the track coming off the land from the NW is about 38 km from the shore.

Figure 1. Locations of TOPEX data off Lyme Bay

VALIDATION:
Even where altimeter data are available, it is important to be certain that the measurements, in the form of significant wave height (Hs) , surface wind speed (U) and zero upcrossing wave period(Tz), are reliable. After a comparison against in situ data around the UK, the altimeter Hs measurement was found to be robust and accurate (with a residual root mean square, rrms, of less than 0.45 m). The new algorithm used to estimate Tz measurement was also found to be reliable (rrms of less than 0.8 s). However, in an important new finding, the altimeter wind speed was found to underestimate (by up to 30%) in more sheltered sea areas where there is likely to be less swell.

SMALL SCALE VARIABILITY:
Where along-track altimeter data are available, they can provide valuable information about small scale variability in waves and winds. Figure 2 illustrates how the mean wave height decreases northwards into Lyme Bay. A similar analysis of altimeter data in the Bristol Channel showed a sheltering effect in the lee of Lundy Island (to the north-east).

Figure 2. Mean wave height from TOPEX tracks across Lyme Bay; *: into Lyme Bay, diamond: out of Lyme Bay

Model Boundary Conditions:
Altimeter data were used to generate boundary conditions for the shallow water models. The technique developed was to generate time series (for the STORM wave model), or representative mean and extreme values (for the SWAN wave model), of the boundary parameters required. The models require a full 2D wave spectrum at the boundary, so it was necessary to assume a particular form for the wave spectrum (JONSWAP) and fit this to altimeter derived parameters. The models also need input at each boundary point (~200m in JERICHO), which is produced by interpolation. The extrapolation from the nearest altimeter observation to the model boundary is site-specific, depending on the coastal configuration, the bathymetry, the proximity of the satellite track, and the availability of in situ data. Figure 3 illustrates the locations of satellite data and model boundary at Carmarthen Bay. The satellite data cannot provide wave direction. This must be inferred from in situ wind or wave data, a larger scale wave model, or other external data source.

Figure 3 The contours show the SWAN model area for Carmarthen Bay, with model output points as red crosses. The circles are the altimeter tracks (dark/light blue TOPEX, magenta/green ERS-2)

.
Estimating Nearshore Extremes:
Offshore, all data from a large area can be used to estimate the parameters of the Hs distribution. Nearshore, because of the large changes over a few km, this approach is not possible. Here, estimates of the Hs distribution can be obtained by analysing altimeter values at individual data-set locations. For TOPEX data this would give only one measurement every ten days (except near a crossover point), generating ~200 values in 6 years. However, these are independent values, unlike 3 hourly buoy values, and contain between-year variability. Figure 4 shows an FT-1 distribution fitted to TOPEX data near Lyme Bay, indicating a 100 year return value of about 10 m.

Figure 4. Distribution of Hs from TOPEX data at the STORM wave model boundary in Lyme Bay.

JONSWAP A wave spectrum for wind driven seas, based on results of the Joint North Sea Wave Project

STORM Halcrow's ray tracing wave model.

SWAN A public domain third generation fully spectral gridded wave model for shallow water. The latest version can be downloaded from http://swan.ct.tudelft.nl/

TOPEX A US/French altimeter satellite, launched in 1992.

THE PARTNERSHIP (see Contact) : The Environment Agency was the customer for JERICHO, it wished to develop its long term strategy for the protection of the English and Welsh coastline. The Centre for Coastal and Marine Science — Proudman Laboratory, and Halcrow Maritime provided expertise in shallow water wave modelling. Southampton Oceanography Centre undertook analyses of large scale wave climate variability and provided computing support, and Satellite Observing Systems were project managers and carried out analyses of satellite and in situ data. JERICHO was supported by the British National Space Centre under the Earth Observation LINK programme.