222 Allan R. Robinson and Jurgen Sellschopp
forecasts. Fig. 11.5b illustrates the upper thennocline dominant circulation mesoscale and sub-basin scale features present during RR97. The most energetic features
are the two anticyclonic eddies located in the central Ionian (Al and A2). They can
be easily observed in both altimeter and CTD (dynamic height) data. Between
these two anticyclones we observe two cyclonic eddies, CI and C2. The two eddies
CI and C2 occasionally appear as part of a larger cyclonic system that separates AI
and A2. Two smaller anticyclonic gyres are present in the same area, without being
visible in most ofthe SSH images (probably due to their size). The first one (A3) is
located south of the Gulf of Messina. At surf ace and subsurface layers (O-lOOm) it
appears to be connected with A2. The second anticyclone (A4) is located east of
A2 and C2. All four anticyclones (Al-A4) seem to be part ofthe same system, the
Ionian Anticyclone. In the eastern Ionian, along the coasts of Greece we observe
the Cretan cyclone (C3), the Pelops Anticyclone (AS), and two cyclonic gyres (C4
and C5). The Cretan Cyclone is a pennanent sub-basin scale feature of the East
Mediterranean upper thennocline.
The model (Fig. 11.9)seems to reproduce the three phases of the Ionian Anticyclone: the lst phase when the large scale circulation dominates and the anticyclones are part of the same system; the 2nd phase with i) the intensification of
cyclones CI and C3, ii) the weakening ofthe large scale circulation, iii) the intensification of A2; and the 3rd phase when Al and A2 seem to merge again in large
scale circulation. Finally, it is interesting to note that the model simulates the
cyclonic circulation in the NE Ionian (C4 - C5) but the Cretan cyclone (C3) in the
SE Ionian is not well represented at least during the first half of the period (up to
Sept. 8).
11.6.5 Accomplishments and implications
The time evolution and thus the predictive capability (recall section 11.5) of
regional circulation systems is driven by internal dynamical processes, inflow and
outflow boundary conditions and surface atmospheric fluxes. Our knowledge of
intemal dynamical processes and boundary conditions is affected by the type, density, accuracy and collection strategy of environmental data and the calibration and
tuning of the physical dynamical model. Rapid Response forecasts have proven to
be usefully accurate. The predictability of the individual regions is limited: in
weather prediction, to a few days by atmospheric internal non-linear dynamics,
and, within the ocean, to a week or two by similar processes.
Each of the Rapid Response exercises provided new and unique opportunities.
While many aspects of the REA process remained consistent between exercises,
the variation in domains and dynamical processes contributed to differences
between each operation. RR96 was our first experience in REA. Dialogs were initiated with naval officers and products tailored for specific warfare activities. The
NRV Alliance served as a shipboard forecast center. HOPS utilized a novel synoptic initialization from wanned historical synoptic data. Aircraft adaptive sampling
was initiated both subjectively and based on the first real time ESSE error forecasts.
forecasts. Fig. 11.5b illustrates the upper thennocline dominant circulation mesoscale and sub-basin scale features present during RR97. The most energetic features
are the two anticyclonic eddies located in the central Ionian (Al and A2). They can
be easily observed in both altimeter and CTD (dynamic height) data. Between
these two anticyclones we observe two cyclonic eddies, CI and C2. The two eddies
CI and C2 occasionally appear as part of a larger cyclonic system that separates AI
and A2. Two smaller anticyclonic gyres are present in the same area, without being
visible in most ofthe SSH images (probably due to their size). The first one (A3) is
located south of the Gulf of Messina. At surf ace and subsurface layers (O-lOOm) it
appears to be connected with A2. The second anticyclone (A4) is located east of
A2 and C2. All four anticyclones (Al-A4) seem to be part ofthe same system, the
Ionian Anticyclone. In the eastern Ionian, along the coasts of Greece we observe
the Cretan cyclone (C3), the Pelops Anticyclone (AS), and two cyclonic gyres (C4
and C5). The Cretan Cyclone is a pennanent sub-basin scale feature of the East
Mediterranean upper thennocline.
The model (Fig. 11.9)seems to reproduce the three phases of the Ionian Anticyclone: the lst phase when the large scale circulation dominates and the anticyclones are part of the same system; the 2nd phase with i) the intensification of
cyclones CI and C3, ii) the weakening ofthe large scale circulation, iii) the intensification of A2; and the 3rd phase when Al and A2 seem to merge again in large
scale circulation. Finally, it is interesting to note that the model simulates the
cyclonic circulation in the NE Ionian (C4 - C5) but the Cretan cyclone (C3) in the
SE Ionian is not well represented at least during the first half of the period (up to
Sept. 8).
11.6.5 Accomplishments and implications
The time evolution and thus the predictive capability (recall section 11.5) of
regional circulation systems is driven by internal dynamical processes, inflow and
outflow boundary conditions and surface atmospheric fluxes. Our knowledge of
intemal dynamical processes and boundary conditions is affected by the type, density, accuracy and collection strategy of environmental data and the calibration and
tuning of the physical dynamical model. Rapid Response forecasts have proven to
be usefully accurate. The predictability of the individual regions is limited: in
weather prediction, to a few days by atmospheric internal non-linear dynamics,
and, within the ocean, to a week or two by similar processes.
Each of the Rapid Response exercises provided new and unique opportunities.
While many aspects of the REA process remained consistent between exercises,
the variation in domains and dynamical processes contributed to differences
between each operation. RR96 was our first experience in REA. Dialogs were initiated with naval officers and products tailored for specific warfare activities. The
NRV Alliance served as a shipboard forecast center. HOPS utilized a novel synoptic initialization from wanned historical synoptic data. Aircraft adaptive sampling
was initiated both subjectively and based on the first real time ESSE error forecasts.