Saltzman B. (editor) Anomalous Atmospheric Flows and Blocking

Подождите немного. Документ загружается.

384

CARLOS R. MECHOSO

AL.

DAYS

6-10

DAYS

6-10

FIG.

Fig.

except

for

the

low-resolution 9-layer forecast from

January.

4.2.

Forecast

Jrom

January

The initial condition for the geopotential field at

500

mbar is in Fig.

The

5-day forecasts with the four versions of the model are shown in Fig.

The

omega-shaped pattern over the North Atlantic Ocean, consisting of a high

over the southern tip of Greenland flanked by two lows at lower latitudes, is

visible in all forecasts, especially in those with the high resolution. On the

other hand, the pronounced trough over the Rocky Mountains is underpre-

dieted

all forecasts. The forecasts for high latitudes show some lack of

detail. The accuracy

the forecasts decreases considerably after day

closer

look

the evolution

zonal wavenumber

in geopotential

height at

500

mbar,

60"N,

is obtained from the harmonic dial shown in Fig.

DAYS

6-10

NUMERICAL FORECASTS OF EVENTS

385

DAYS

6-10

DAYS

6-10

DAYS

6-10

FIG.

As in Fig.

except for the low-resolution 15-layer forecast.

10. In this plot the distance from the origin represents the amplitude in

meters, the right half of the horizontal axis corresponds to a ridge over the

Greenwich Meridian, and a counterclockwise progression corresponds to an

eastward translation of the ridge. Squares and circles on the lines indicate

values at

-day intervals. Wavenumber

decayed during this period while

slowly moving westward (thick solid line). The forecasts show retrogression

for the first few days, with the largest westward translation obtained from the

low-resolution 9-layer forecast (dashed-dotted line), and the smallest from

the high-resolution

5-layer forecast (dashed line). The predicted amplitude

of wavenumber

decreases during the first half of all forecasts, but remains

considerably higher than observed during the second half except for the

low-resolution

5-layer forecast (thin solid line).

DAY

9-LAYER

4x5

DAY

15-LAYER

4x5

DAY 5 9-LAYER

2.4X3

DAY

15-LAYER

2.4X3

FIG.

9. Geopotential field (meters) at 500 mbar

for

January (top), and

forecasts

from

January with the low-resolution 9-layer version (middle left), low-resolution 15-layer version

(middle right), high-resolution 9-layer version (bottom left), and high-resolution 15-layer ver-

sion

(bottom right).

NUMERICAL FORECASTS OF EVENTS

387

QOOE

180'

QOOW

FIG.

10.

Harmonic dial for zonal wavenumber

observed

500-mbar popotential

height at

60"N

during the period

January to

January (thick solid lines) and forecasts with the

low-resolution 9-layer version (dasheddotted lines), low-resolution 15-layer version (thin solid

lines), high-resolution 9-layer version (dotted lines), and high-resolution 15-layer version

(dashed lines). Other notations

are

explained in the text.

Whatever success the forecasts from

January have is largely due to the

reasonably good forecast of the decay of zonal wavenumber

since the

forecast of remaining ultralong waves are poor after

days. This is seen

clearly in Fig.

11,

which shows the histories at

500

mbar, 60"N, of the

superposition of waves

2-4

in geopotential field. The observed field shows

an eastward-propagating wave train which seems to originate early in the

period east of the Rocky Mountains. Such a wave train results in an intensi-

fication of the omega-shaped pattern over the North Atlantic Ocean and

seems to

related to an intensification of a ridge west

North America.

(Inspection of other levels reveals that this wavetrain has little tilt with

height.) All forecasts performed poorly in this respect.

4.3.

Forecast

fiom

February

The initial geopotential field at

500

mbar is shown in Fig.

12.

There are

pronounced ridges over the eastern North Atlantic Ocean and over the

Bering Strait. There is also an intense, cyclonic circulation at very high

388

CARLOS

MECHOSO

ETAL

LONG

TUDE

-*I

ODnJQ)

m'vw

o'v

,I=

FIG.

1 1.

Contribution to the geopotential field (meters) at

60"N.

500

mbar by wavenumbers

-4

during the

period

January

January. From top to bottom: observed, low-resolution

9-layer version, low-resolution 15-layer version, high-resolution 9-layer version, and high-reso-

lution 15-layer version. The contour interval is

Values

greater than

120

m are lightly

shaded, and values less than

-120

are

heady shaded.

NUMERICAL FORECASTS OF EVENTS

389

0000

GMT

FEB

1979

FIG.

12.

Geopotential field (meters) at

500

mbar

for

February.

latitudes with a clear wavenumber

pattern. The intense flows at very high

latitudes over the prime meridian and the dateline remain

days after, as

shown in Fig. 13. The ridge over the eastern North Atlantic Ocean is more

intense than on 15 February and extends over northern Europe, poleward of

a trough over the Mediterranean Sea. The ridge over the North Pacific Ocean

(the major episode over the Pacific Ocean sector discussed in Section

also more intense than on 15 February. There is a deep low over northeastern

North America and a cyclone over northeastern Asia. Inspection of Fig. 13

reveals that, in general, the high-resolution forecasts are considerably more

accurate than the low-resolution forecasts, particularly over the North At-

lantic and Pacific Oceans. All forecasts show a much weaker flow over the

polar region than observed.

The 5-day means from 20 February to

February of the observed and

predicted geopotential field at

500

mbar, shown in Fig. 14, illustrate the

higher accuracy of the high-resolution forecasts in the prediction of quasi-

stationary features of the circulation during the second half of the period.

The ridges over northwestern Europe and over the Bering Sea and the lows

over northern North America and Siberia are well predicted, reflecting a

good forecast of the evolution of wavenumber

which was very important

390

CARLOS R. MECHOSO

0000

GMT

FEB

1979

DAY

15-LAYER

4x5

DAY

9-LAYER

2.4X3

DAY

15-LAYER

2.4X3

FIG.

13.

Fig.

except

for

February and

forecasts

from

February.

NUMERICAL FORECASTS OF EVENTS

0000

GMT

20-25

FEB 1979

DAYS 6-10 9-LAYER

4x5

DAYS 6-10 15-LAYER

DAYS 6-10 9-LAYER

2.4X3

DAYS

6-10

15-LAYER

2.4X3

FIG.

14.

As in Fig.

13,

except

for

the

5-day

means

from

February

February.

392

CARLOS

MECHOSO

OOOOGMT

FEB

1979

FIG.

IS.

Geopotential field (meters) at

500

mbar for

February.

during that period. On the other hand, the predicted midlatitude circulation

over North America shows a trough (ridge) over the eastern (western) part of

the continent, in disagreement with the observed situation.

4.4.

Forecast

jrom

February

The geopotential field at

500

mbar for

February is shown in Fig.

15,

and

the 5-day forecasts for that level are shown in Fig.

16.

The ridge over the

eastern North Atlantic is predicted with generally weaker amplitudes.

common failure

all forecasts, particularly of those with the low-resolution

versions, is on the northern Pacific Ocean region where the ridge over the

Bering Strait is greatly underpredicted. The accuracy of all forecasts de-

creases notably after day

STRATOSPHERIC FORECASTS

These are discussed in detail in Mechoso

et al.

(1985).

We summarize

some of the results in what follows.

NUMERICAL FORECASTS

EVENTS

393

0000

GMT

22 FEB

1979

DAY 5 9-LAYER 4x5

DAY

15-LAYER 4x5

FIG.

16.

Geopotential field (meters) at 500 mbar for

February (top left), and forecasts

from

February with the low-resolution 9-layer version (top

right),

low-resolution 15-layer

version (bottom left), and high-resolution 15-layer version (bottom right).

5.1.

Forecastsfrom

January and

January

The

0-mbar geopotential field for 16 January is shown in Fig.

17.

The 5-

and 10-day low-resolution 15-layer forecasts from 16 January are shown in

the same figure. The transition that occurred at this level, from a circumpo-

lar cyclonic vortex to a two-vortex system, consisting of a cyclonic vortex