10 Strategy for Regional Seasonal
Forecasts
KIKURO MIY AKODA
George Mason University, COLA, Maryland, USA
10.1 Numerical predictions in general
The topic is numerical weather prediction, in particular air-sea coupled forecasts
over the Mediterranean Sea and the surrounding areas. At this stage, it is not certain whether or not this venture will be successful. The key questions are: (Question 1) Are three month forecasts over the Mediterranean Sea possible ? and
(Question 2) Are forecast studies of the ocean state or ecological conditions
useful?
Global-to-mesoscale
The first issue is how small a scale is predictable for a season. In the 1970's,
when mesosca1e weather forecasts became a focus of research, various definitions
were proposed for mesoscale disturbances.
Or1anski's (1975) nomenclature for the scale of atmospheric disturbances is as
follows:
1) macro a (L>10 4 km);
2) macro ~ (l04km>L>2 10 3 km)
3) meso a (2 10 3 km>L>2 10 2 km)
4) meso /3 (200 krn>L>20 krn)
5) meso ' Y (20 krn>L>2 krn)
6) micro a
7) micro ~
where L is the disturbance wavelength.
This list may lead one to speculate that for seasona1 time scales, meso ~ sca1es
may not be predictable, and perhaps even the meso a sca1es are questionable. But
this question itself is a major research issue.
For the ocean the dominant time scales are much longer than those for the atmosphere because of the greater thermal and kinetic inertia. In addition, the horizontal
length scales of the relevant motions in the ocean are approximately one tenth of
those in the atmosphere.
One measure of the relevant horizontallength scale is the radius of deformation,
L defined as the ratio ofthe phase speed of gravity waves, C g to the Coriols parameter, f:
Forecasts
KIKURO MIY AKODA
George Mason University, COLA, Maryland, USA
10.1 Numerical predictions in general
The topic is numerical weather prediction, in particular air-sea coupled forecasts
over the Mediterranean Sea and the surrounding areas. At this stage, it is not certain whether or not this venture will be successful. The key questions are: (Question 1) Are three month forecasts over the Mediterranean Sea possible ? and
(Question 2) Are forecast studies of the ocean state or ecological conditions
useful?
Global-to-mesoscale
The first issue is how small a scale is predictable for a season. In the 1970's,
when mesosca1e weather forecasts became a focus of research, various definitions
were proposed for mesoscale disturbances.
Or1anski's (1975) nomenclature for the scale of atmospheric disturbances is as
follows:
1) macro a (L>10 4 km);
2) macro ~ (l04km>L>2 10 3 km)
3) meso a (2 10 3 km>L>2 10 2 km)
4) meso /3 (200 krn>L>20 krn)
5) meso ' Y (20 krn>L>2 krn)
6) micro a
7) micro ~
where L is the disturbance wavelength.
This list may lead one to speculate that for seasona1 time scales, meso ~ sca1es
may not be predictable, and perhaps even the meso a sca1es are questionable. But
this question itself is a major research issue.
For the ocean the dominant time scales are much longer than those for the atmosphere because of the greater thermal and kinetic inertia. In addition, the horizontal
length scales of the relevant motions in the ocean are approximately one tenth of
those in the atmosphere.
One measure of the relevant horizontallength scale is the radius of deformation,
L defined as the ratio ofthe phase speed of gravity waves, C g to the Coriols parameter, f: