174
IAN ROBINSON
particular care must be taken in defining which type of SST is needed, since
the atmosphere is in contact with the skin temperature rather than the upper
mixed layer temperature normally represented in the ocean model.
3.4
Ocean waves
The nowcasting and forecasting of ocean waves is an operational task
which benefits many different users of the sea and is essential for the safety
of mariners and for cost-effective navigation. Ocean wave forecasting
models depend on good wind forecasts from numerical weather prediction
models. Their performance can also be improved if good observations of
wave data can be assimilated in a timely way. In situ measurements from
wave buoys offer a means of testing and validating model predictions, but
they are isolated and too few to make much impact if assimilated into a
model. This requires the wide area coverage offered by satellites. Here we
consider two ways in which ocean wave data potentially useful for models
are measured from space.
3.4.1
Significant wave height measured by altimeters
When an altimeter measures the time for an emitted pulse to return, it
tracks in detail the shape of the leading edge of the echo, from which it is
possible to make a very good estimate of the significant wave height, H 1/3 ,
within the pulse-limited footprint illuminated by the altimeter. For a
perfectly flat calm surface the return echo has a very sharp edge. If there are
large waves, several metres in height from trough to crest, then the return
signal starts to rise earlier, as the first echoes are received from the crests,
but takes longer to reach its maximum, when the first echoes are received
from the wave troughs. The rising edge of the echo is modelled by a
function in terms of the root mean square ocean wave height, so that by
matching the observed shape to the model function it is easy to gain an
estimate of H 1/3 . This method has delivered robustly accurate measurements
of H 1/3 for more than twenty years from different altimeters (Cotton &
Carter, 1994) and comparison with buoys shows root mean square
differences of only 0.3 m (Gower, 1996), which is the limit of the buoy
accuracy.
It therefore provides an excellent source of information to be used in
wave models. At present there are at least two altimeters in operation (see
chapter 11 of Robinson (2004)), each crossing the Earth with 14-15 orbits
per day. However, the altimeter footprint measures waves only along a 10
km wide swath along the satellite ground track. At the Equator the tracks of
successive orbits are nearly 3000 km apart, and so in one day the available
samples are quite sparse and it is possible to miss altogether local regions of
high waves associated with a recent storm. Even with several altimeters in
IAN ROBINSON
particular care must be taken in defining which type of SST is needed, since
the atmosphere is in contact with the skin temperature rather than the upper
mixed layer temperature normally represented in the ocean model.
3.4
Ocean waves
The nowcasting and forecasting of ocean waves is an operational task
which benefits many different users of the sea and is essential for the safety
of mariners and for cost-effective navigation. Ocean wave forecasting
models depend on good wind forecasts from numerical weather prediction
models. Their performance can also be improved if good observations of
wave data can be assimilated in a timely way. In situ measurements from
wave buoys offer a means of testing and validating model predictions, but
they are isolated and too few to make much impact if assimilated into a
model. This requires the wide area coverage offered by satellites. Here we
consider two ways in which ocean wave data potentially useful for models
are measured from space.
3.4.1
Significant wave height measured by altimeters
When an altimeter measures the time for an emitted pulse to return, it
tracks in detail the shape of the leading edge of the echo, from which it is
possible to make a very good estimate of the significant wave height, H 1/3 ,
within the pulse-limited footprint illuminated by the altimeter. For a
perfectly flat calm surface the return echo has a very sharp edge. If there are
large waves, several metres in height from trough to crest, then the return
signal starts to rise earlier, as the first echoes are received from the crests,
but takes longer to reach its maximum, when the first echoes are received
from the wave troughs. The rising edge of the echo is modelled by a
function in terms of the root mean square ocean wave height, so that by
matching the observed shape to the model function it is easy to gain an
estimate of H 1/3 . This method has delivered robustly accurate measurements
of H 1/3 for more than twenty years from different altimeters (Cotton &
Carter, 1994) and comparison with buoys shows root mean square
differences of only 0.3 m (Gower, 1996), which is the limit of the buoy
accuracy.
It therefore provides an excellent source of information to be used in
wave models. At present there are at least two altimeters in operation (see
chapter 11 of Robinson (2004)), each crossing the Earth with 14-15 orbits
per day. However, the altimeter footprint measures waves only along a 10
km wide swath along the satellite ground track. At the Equator the tracks of
successive orbits are nearly 3000 km apart, and so in one day the available
samples are quite sparse and it is possible to miss altogether local regions of
high waves associated with a recent storm. Even with several altimeters in