Impact of Horizontal Resolution on Rapid

1
Impact of Horizontal Resolution on
Rapid Intensity Change Forecasts using
the Hurricane Research System
R. Rogers, S. Gopalakrishnan, K. Yeh, and X. Zhang
NOAA/AOML Hurricane Research Division Miami, FL
2
Motivation

• Advances in forecasts of tropical cyclone (TC)
  intensity, structure, and rainfall lag advances
  in TC track forecasts
• Multiscale nature of these processes major
  reason for this
• environmental - O(1000 km) - troughs, shear
• vortex - O(1-100 km) - symmetric/asymmetric
  dynamics, VRWs
• convective O(1 km) convective bursts,
  vortical hot towers
• turbulent O(1-100 m) - surface fluxes,
  entrainment/detrainment
• microscale O(1mm) -- hydrometeor production,
  latent heat release
• Numerical model forecasts of rapid intensity
  change (RI) should improve as resolution
  increases
• For some cases it does, in others it does not
  why?

3
Examples

• Two cases of RI shown here
• One that showed improvement with higher
  resolution, one that did not. All runs had
  similar (and good) track forecasts.
• Hurricane Katrina (2005)
• Underwent RI from Cat 3 to Cat 5 as it traversed
  Gulf of Mexico
• Both resolutions produced RI, but both delayed
• Good observational coverage (multiple P-3 flights
  with Doppler)
• Hurricane Emily (2005)
• Sustained period of RI, from tropical storm to
  Cat 4 in 48 h
• 3-km run produced RI during 1st 24 h, another
  intensification episode later than Best Track
• 9-km run did not intensify
• No airborne Doppler sampling

4
Track and Intensity Hurricane Katrina
track
intensity
279 km
RI
93 km
5
Vortex-scale fields - Hurricane Katrina
Doppler
9-3 km HRS
27-9 km HRS
70
65
60
Wind speed (m/s) at 2-km altitude
55
50
distance (km)
distance (km)
distance (km)
45
40
35
30
distance (km)
distance (km)
distance (km)
7
6
5
Vorticity (x 10-3 s-1) and w (m/s) at 4-km
altitude
4
3
distance (km)
2
distance (km)
distance (km)
1
0
-1
-2
distance (km)
distance (km)
distance (km)
2038-2231 UTC 8/28
0000 UTC 8/29
0000 UTC 8/29
6
Vortex-scale fields Hurricane Katrina
Radius-height axisymmetric (R-z mean) plots
Doppler
9-3 km HRS
27-9 km HRS
18
18
18
16
16
16
14
14
14
80
70
12
12
12
60
10
10
10
50
height (km)
40
8
8
8
Tangential and radial winds (m/s)
30
6
6
20
6
10
4
4
4
0
2
2
2
0
0
0
20
40
70
100
10
30
50
60
80
90
20
40
70
100
10
30
50
60
80
90
20
40
70
100
10
30
50
60
80
90
radius (km)
radius (km)
radius (km)
18
18
18
16
16
16
80
14
14
14
70
12
12
12
60
50
10
10
10
height (km)
40
8
8
8
Vorticity (x 10-4 s-1) and vertical motion (m/s)
30
20
6
6
6
10
4
4
4
0
2
2
2
20
40
70
100
10
30
50
60
80
90
20
40
70
100
10
30
50
60
80
90
20
40
70
100
10
30
50
60
80
90
radius (km)
radius (km)
radius (km)
2038-2231 UTC 8/28
0000 UTC 8/29
0000 UTC 8/29
7
Convective-scale statistics Hurricane Katrina
Contoured frequency by altitude diagrams (CFADs)
Variation of distribution with height
vertical motion
Doppler
9-3 km HRS
27-9 km HRS

2038-2231 UTC 8/28
0900Z 29 Aug
0900Z 29 Aug
8
Track and Intensity Hurricane Emily
track
intensity
279 km
RI
93 km
9
Environmental-scale fields - Hurricane Emily
850-200 hPa hodographs and inner-core vertical
shear
850
1200 UTC 7/13
200
1200 UTC 7/14
10
Vortex-scale fields Hurricane Emily
Time-radius Hovmollers of axisymmetric winds at
10-m height
279 km
93 km
00Z 16 JUL
00Z 16 JUL
12Z 15 JUL
12Z 15 JUL
00Z 15 JUL
00Z 15 JUL
30
20
25
15
20
12Z 14 JUL
12Z 14 JUL
10
15
10
5
00Z 14 JUL
00Z 14 JUL
5
12Z 13 JUL
12Z 13 JUL
00Z 13 JUL
00Z 13 JUL
100
150
200
50
100
150
200
50
radius (km)
radius (km)
11
Vortex-scale fields Hurricane Emily
R-z mean plots of tangential and radial winds
(m/s)
16
16
16
14
14
14
35
12
12
12
30
10
10
25
10
20
279 km
8
8
height (km)
8
15
10
6
6
6
5
4
4
4
0
2
2
2
0
0
0
25
50
75
100
25
50
75
100
25
50
75
100
radius (km)
radius (km)
radius (km)
16
16
16
14
14
14
35
12
12
12
30
25
10
10
10
20
height (km)
8
8
8
15
93 km
10
6
6
6
5
4
4
4
0
2
2
2
0
0
0
25
25
50
75
100
25
50
75
100
50
75
100
radius (km)
radius (km)
radius (km)
1200 UTC 7/14
1200 UTC 7/15
1200 UTC 7/13
12
Vortex-scale fields - Hurricane Emily
Vorticity (x 10-5 s-1) and vertical motion (m/s)
at 4-km altitude
300
300
300
300
200
250
250
250
150
100
200
200
200
75
279 km
50
distance (km)
150
150
150
25
10
100
100
100
-10
-25
50
50
50
150
200
150
200
150
200
50
100
250
300
50
100
250
300
50
100
250
300
distance (km)
distance (km)
distance (km)
300
300
300
300
200
250
250
250
150
100
200
200
200
75
distance (km)
50
93 km
150
150
150
25
10
100
100
-10
100
-25
50
50
50
50
100
150
200
250
300
50
100
150
200
250
300
150
200
50
100
250
300
distance (km)
distance (km)
distance (km)
1200 UTC 7/14
1200 UTC 7/15
1200 UTC 7/13
13
Convective-scale statistics Hurricane Emily
CFADs of vertical motion
18
18
16
16

14
14
12
12
279 km
10
10
8
8
6
6
4
4
2
2
0
0
18
18
16
16

14
14
12
12
10
10
8
8
93 km
6
6
4
4
2
2
0
0
1200 UTC 7/14
1200 UTC 7/15
1200 UTC 7/13
14
Summary

• Katrina case showed little improvement in
  predicting rapid intensity change at higher
  resolution, Emily case showed improvement
• Katrina
• Storm larger than observed in both runs
• Better eyewall slope, secondary circulation
  features in 3-km
• More realistic vorticity structures, stronger
  updrafts/downdrafts in 3-km
• Emily
• Shear, RMW same in both runs prior to 1200 UTC
  7/14
• Well-defined secondary circulation in 3-km,
  poorly-defined in 9-km HRS
• Inner-core vorticity stronger, more strong
  positive and negative vorticity cores in 3-km
• More strong (12-15 m/s) updrafts in 3-km

Questions

• What was difference between Katrina and Emily
  runs?
• Secondary circulation, RMW
• What is importance of convective-scale fields?
  Is ability to resolve them vital for improved
  rapid intensity change forecasts?
• What is robustness of physical parameterizations
  (e.g., surface fluxes, microphysics)?

15
Future work

• Perform more diagnostics
• Additional vortex-scale evaluations (e.g.,
  asymmetric fields)
• Additional convective-scale evaluations (e.g.,
  mean profiles and mass flux evolution)
• Finer-scale diagnostics (turbulent, microphysical
  scale)
• Extend evaluations to more cases, both with and
  without improvements at higher resolution
• Include 1-km HRS runs
• Evaluate runs with additional components (e.g.,
  air-sea coupling, inner-core data assimilation)
• Test alternate physical parameterizations,
  formulations
• Evaluate impact of resolution, parameterizations
  using observations

16
Supplemental slides
17
Vortex-scale fields - Hurricane Emily
Wind speed at 10-m height
800
800
800
700
700
700
40
35
600
600
600
30
500
500
500
25
20
distance (km)
400
400
400
279 km
15
300
300
300
10
5
200
200
200
100
100
100
distance (km)
distance (km)
distance (km)
600
600
600
500
40
500
500
35
400
30
400
400
25
distance (km)
20
300
300
300
15
93 km
10
200
200
200
5
100
100
100
100
200
500
600
300
400
100
200
300
400
500
600
100
200
300
400
500
600
distance (km)
distance (km)
distance (km)
1200 UTC 7/14
1200 UTC 7/15
1200 UTC 7/13
18
Vortex-scale fields Hurricane Emily
R-z mean plots of vorticity (x 10-4 s-1) and
vertical motion (m/s)
50
45
40
279 km
35
30
25
20
15
10
5
0
50
100
50
100
25
75
50
100
25
75
25
75
radius (km)
radius (km)
radius (km)
50
45
40
35
30
25
93 km
20
15
10
5
0
50
100
25
75
100
50
100
25
75
50
25
75
radius (km)
radius (km)
radius (km)
1200 UTC 7/14
1200 UTC 7/15
1200 UTC 7/13