Doppler Radar Wind Analysis

Doppler Radar wind analysis
Kwang-Ho Kim
Radar analysis division of Weather Radar Center
21 June 2022
[email protected]
Instruments to
measure winds
21 June 2022 Radar wind analysis

Strong wind
Wind shear
Typhoon 3

Instruments to measure winds
Anemometer Radio sonde

Instruments to measure winds
Doppler weather radar
Wind profiler radar

Doppler velocity

Doppler velocity (Radial velocity)
 Measurement

 Consideration of radial velocity

True Velocity Equals

Velocity Vector to Doppler Velocity
(True Velocity)
Positive
Doppler Zero Doppler
Velocity Velocity
Negative Doppler
Velocity
Doppler radar

Analysis of Doppler velocity
[STEP 01] Find the line of radial velocity of 0m/s (i.e. zero isodop)
[STEP 02] Draw a virtual line between radar and the certain point over zero isodop
[STEP 03] Draw a perpendicular line to the virtual line
[STEP 04] Add the arrow head to the perpendicular line
(from inbound (negative sign) to outbound (positive sign))
10

11

 Divergence
 Convergence
12

 Cold advection
 Warm advection
13

 Cold front
14

 Doppler velocity of KMA’s weather radar
15

Wind fields
from Doppler velocity
16

Wind fields from Doppler velocity
Doppler velocity Wind fields
17

레이더 바람
 Doppler velocity equation
𝑥 𝑦 𝑧
𝑉 𝑟 =𝑢 ∙ +𝑣 ∙ +(𝑤 +𝑉 𝑡 ) ∙
𝑟 𝑟 𝑟
18

레이더 바람
Traditional analysis WRC
Single Doppler radar (VAD, VVP)
a
Multiple Doppler radar

e (Miller and Anderson, 1991)
(Lhemitte & Atlas, 1961)
Dual Doppler radar (Amijo, 1969)
19

Single Doppler Radar analysis
 Vertical wind profile

a
20

Single Doppler Radar analysis
 Time series of vertical wind profile
21

Dual Doppler Radar analysis
For M radars, the linear system of equations to be solved at time t.
uâ m  vˆbm  Wˆ c m  d m ; m  1, 2, 3, ...... m Where, dm is the term involving radial velocity.
For 2 radars, the linear system of equations to be solved at time t.

uÂ1  vˆB1  Wˆ C1  D1
Precipitation particle (x,y,z)
z
uÂ2  vˆB2  Wˆ C2  D2 V1
The solutions are: y

D1 B2  D2 B1 ˆ B1C2  B2C1 V
uˆ  W  u   uWˆ
D D
D A D A A C  AC V2
vˆ  2 1 1 2  Wˆ 2 1 1 2  v   vWˆ R1 R2
D D
z-z1
where the determinant of coefficients is
Radar2 (x2,y2,z2)
D  A1 B2  A2 B1
z1-z2
y -y 1 x -x
Ai  ( x  xi ), Bi  ( y  yi ), Ci  ( z  zi ), Di  Vi Ri 1
x -x y -y 2
Horizontal wind O Radar1 (x1,y1,z1) 2
uˆ  u 2   u w
ˆ u2  u    uVt
form x
vˆ  v 2   v w
ˆ v2  v   vVt
22

The vertical air velocity (w) was determined as a difference between the measured Doppler vertical
velocity (W) and a bulk-estimate of the precipitation particle fall speed, V t (Rogers, 1964).
Wˆ  w  Vt From Marks (1987)
wtop = 0 ms-1
The mean terminal fall speed of the precipitation particles was assumed to be a
function of reflectivity and height; V t (ms-1) was estimated from Z (mm6m-3) as
follows: snow
0.45
  2.600Z 0.107 for altitudes  ML  0.5 km 
Vt   0    Rain+snow Melting
    
0.063
 0 .817 Z for altitudes ML 0.5 km  Layer (ML)
z

 ( z )  0e H Rain
The empirical relationship for rain : Joss and Waldvogel (1970)
The empirical relationship for snow : Atlas et al. (1973)
15
12
Mass continuity equation
Altitude (km)
9
6
  ( w)  ( u2 )  ( v2 )  (  u w)  (  u w)
  ( V )      0 3
z x y x y 0
0.0 0.3 0.6 0.9 1.2 1.5
Air Density (kgm-3) 23

24

Multiple Doppler Radar analysis
Wind field Overlapped observation area
25

Limitation of traditional anlaysis
• Sensitive to radar position and observation strategy
1. Along/near radar baseline 2. Vertical velocity boundary condition
 As the angle between baseline and analysis point  Setting of vertical velocity boundary conditions and
increase, accuracy of wind field decrease rapidly. vertical integration of continuity equation
 Wind vector can be calculated only over the over-  Error of vertical integration of continuity equation due
lapping area of two or more radar observations. to radar observation gap at lower layer
 ( w ) ( u ) ( v ) (  w ) (  w )
[Effective area]   ( V )   2
 2
 u
 u
0
z x y x y
26

Variational Doppler radar analysis
• Overcome the limitation of traditional method through minimization
of cost function composed of various constrains
- However, it takes a long computational time due to iterative calculation to minimize cost function
𝑱 = 𝑱 𝒐 + 𝑱 𝒃+ 𝑱 𝒅 + 𝑱 𝒍 +
(Gao et al., 1999) (Liou and Chang, 2009)
o: radial velocity observation, b: background wind, d: continuity equation, v: vorticity equation, l: laplacian filter
bot: boundary condition of bottom, top: boundary condition of top 27

Multiple wind field Variational wind field
28

Multiple wind area Variational wind area
63% 100%
29

Operational status of
Doppler velocity analysis
in KMA
30

WISSDOM
WISSDOM: WInd Synthesis System using DOppler Measurements
31

WISSDOM
 Definition
• Variational-based three-dimensional wind synthesis method using multiple-Doppler

radars
 Technical exchange
• WISSDOM was originally developed by Prof. Yu-Chieng Liou of National Central Univer-
sity (NCU) in 2009 and 2012.
• Weather Radar Center of KMA (WRC) visited NCU for technical exchange in Oct. 2017.
• WRC has conducted the localization, modularization, parallelization tasks for real-time
implementation of WISSDOM.
32

WISSDOM
 Characteristics
33

WISSDOM
 Characteristics
WISSDOM configuration
Horizontal grid (x,y) 151, 151
Vertical levels 21
Horizontal & vertical spacing 1 km & 0.5 km
The number of iteration 500
71 min 27 min 7 min 13 sec
Localization Use of supercom- Parallelization

puter of KMA (MPI mode,
Original source of (serial mode) 72 processor)
Modularization
NCU
Debugging ICE-POP support Long term validation
WRC server WRC server Supercomputer Supercomputer
34

Operational applications
35

Dynamic vertical grid size
Altitude # of Grid size
No.
(meter) layer (meter)
1 0 ∼ 1,000 21 50
2 1,000 ∼ 2,000 10 100
3 2,000 ∼ 5,000 15 200
4 5,000 ∼ 10,000 10 500
Regular grid size

No.
1 0 ∼ 1,0000 51 200
Detailed terrain effect
vertical grid size : 200m vertical grid size : 50m
36

X-Z cross-section (shade region: vertical velocity)
Dynamic grid
Regular grid
37

X-Y cross-section at 1.5km altitude

(shade region: vertical velocity)
Regular grid Dynamic grid
38

WISSDOM-AiR (Airport wind information)
 WISSDOM-AiR Domain
Domain of WISSDOM-AiR Spatio-temporal resolution
Horizontal size 200 m
Vertical size Dynamic
# of
Horizontal grid 251 × 251
# of
vertical grid 101
Time resolution 5min

No.
1 0 ~ 1,000 51 20
2 1,000 ~ 2,000 20 50
3 2,000 ~ 3,000 10 100
4 3,000 ~ 5,000 10 200
5 5,000 ∼ 10,000 10 500
39

 WISSDOM VS WISSDOM-AiR
WISSDOM(regular grid) WISSDOM-AiR
Altitude Grid size # of Altitude Grid size # of
No. No.
(meter) (meter) layer (meter) (meter) layer
1 0 ~ 1,000 20 51
2 1,000 ~ 2,000 50 71
1 0 ~ 10,000 200 51 3 2,000 ~ 3,000 100 81
4 3,000 ~ 5,000 200 91
5 5,000 ~ 10,000 500 101
 Case study (2019.03.15 15:30~15:55)
[Low pressure] [Cold front] [Suface observation]
40

 X-Y cross-section at 500m altitude

Time
rainfall
WISSDOM
wind direction
WISSDOM-AiR
41

 X-Y cross-section at 500m altitude

Time
rainfall
WISSDOM
wind speed
WISSDOM-AiR
42

 X-Z cross-section Runway
Time
rainfall
WISSDOM-AiR WISSDOM WISSDOM-AiR WISSDOM
wind direction
wind speed
43

Microburst detection using WISSDOM
 OSSE (Observation System Simulation Experiment)
Numerical simulation Convert to Doppler velocity
Validation
WISSDOM wind retrieval
(NWP VS WISSDOM)
44

 Numerical simulation of Microburst
45

 Numerical wind fields

Alt.
46

 Convert to Doppler velocity

Alt.
P1
P2
P3
P4
47

 WISSDOM wind fields

Alt.
48

 Validation (NWP vs WISSDOM)
Wind vector U V W
NWP
(reference)
WISSDOM
49

Case studies
50

Warm front(2019.03.20. 16:00 ~ 17:00LST)
51
Typhoon: DANAS (2019.07.20. 08:00KST)
NWP
WISSDOM
52
Heavy rainfall (2019.05.27) 00KST
HSR 수상체
레이더바람장 53

54

55

56

57

58

59

60

61

62

Q&A
63

Doppler Radar Wind Analysis

Uploaded by

Copyright:

Available Formats

Doppler Radar Wind Analysis

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Doppler Radar Wind Analysis

Uploaded by

Copyright:

Available Formats

Doppler Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

Anemometer Radio sonde

21 June 2022 Radar wind analysis

Doppler weather radar

Wind profiler radar

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

 Consideration of radial velocity

21 June 2022 Radar wind analysis

True Velocity Equals

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

Doppler velocity Wind fields

21 June 2022 Radar wind analysis

 Doppler velocity equation

21 June 2022 Radar wind analysis

Multiple Doppler radar

(Lhemitte & Atlas, 1961)

Dual Doppler radar (Amijo, 1969)

21 June 2022 Radar wind analysis

 Vertical wind profile

21 June 2022 Radar wind analysis

 Time series of vertical wind profile

21 June 2022 Radar wind analysis

uˆa m  vˆbm  Wˆ c m  d m ; m  1, 2, 3, ...... m Where, dm is the term involving radial velocity.

For 2 radars, the linear system of equations to be solved at time t.

The solutions are: y

21 June 2022 Radar wind analysis

Wˆ  w  Vt From Marks (1987)

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

Wind field Overlapped observation area

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

Multiple wind field Variational wind field

21 June 2022 Radar wind analysis

Multiple wind area Variational wind area

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

• Variational-based three-dimensional wind synthesis method using multiple-Doppler

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis

71 min 27 min 7 min 13 sec

Localization Use of supercom- Parallelization

Debugging ICE-POP support Long term validation

WRC server WRC server Supercomputer Supercomputer

21 June 2022 Radar wind analysis

21 June 2022 Radar wind analysis