USE OF REFLECTIVITY TO VALIDATE HIRLAM

1
USE OF REFLECTIVITY TO VALIDATE HIRLAM

• Irene Sanz Zoydo
• Jose A. García-Moya
• INM

2
OUTLINE

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.
• Case of study.
• 4.1. General description
• 4.2. Experiments
• 4.2.1 HIR (0.2)
• 4.2.2 RAD (0.16)
• 4.2.3 RAH (0.05)
• Results obtained.
• Conclusions.
• Future lines of work.

3
1. Hirlam Model

• HIRLAM FEATURES AT INM.
• - Hirlam version 6.1.2. Outputs between 00 and
  72 hours every hour.
• Area 0.2.
• Boundary conditions Experiment of 0.5 degs.
• Area 0.16.
• Boundary conditions ECMWF (0.5 degs).
• Area 0.05.
• Boundary conditions Experiment with 0.16 degs.

Resol Lon x lat Vert.Levels Area Coordinates Area Coordinates Area Coordinates Area Coordinates
0.2 194x100 31 50.0N 66.5W 15.5S 30.0E
0.16 582x424 40 32.18N 46.5W 35.5S 46.46E
0.05 366x272 40 10.75N 15.0W 10.7S 15.25E
4

1. Hirlam model.
2. Radars at INM.

5
2. Radars at INM

• 14 radars cover all Iberian Peninsula and Canary
  Island.
• Radius of action 240 Km normal mode ? 2x2 Km2.
• Frequency 5620 MHz
• 20 elevation angles ( 0.5 - 25 degs )
• 360 degs at elevation angle
• Time rate 10 minutes

6

• Products obtained from Regional Radar Centre are
  cartesian coordinates (CAPPI or vertical
  sections) (12 Normal mode).
• CAPPI heights over sea surface level are
• Cappi 1 ? (1000-2100) m Cappi 7 ? 7500 m
• Cappi 2 ? 2500 m Cappi 8 ? 8500 m
• Cappi 3 ? 3500 m Cappi 9 ? 10000 m
• Cappi 4 ? 4500 m Cappi 10 ? 12000 m
• Cappi 5 ? 5500 m Cappi 11 ? 14000 m
• Cappi 6 ? 6500 m Cappi 12 ? 16000 m

7

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.

8
3. Radar Simulation Model

• RSM was done at the University of Bonn, adapted
  to Hirlam by FMI and was implemented at INM by
  Jana Sánchez in 2002 (thanks to Carl Fortelius
  from FMI)
• To use RSM some variables related with large
  scale and convective precipitation either in
  solid and liquid form are needed.
• RSM is able to simulate radar reflectivities from
  data of prediction models and to calculate
  statistic about simulated and actual them.

9

• Inputs and outputs

• HIRLAM FILES
• fchh
• fchhmd
• (GRIB)

• RSM FILES
• fchhrsm (ASIMOF)
• (Simulated and actual
• reflectivity files)

RSM
Metview

• RADAR FILES
• Actual Reflectivity
• Latitude and longitude
• (ASCII)

Radar Images
McIdas
Statistics
Radar Images
10

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.
• Case of study.
• 4.1. General description
• 4.2. Experiments
• 4.2.1 HIR (0.2)
• 4.2.2 RAD (0.16)
• 4.2.3 RAH (0.05)

11
4.Case of study (20011010)

• 4.1. General description
• Development of the severe convective situation
  over SE Iberian Peninsula, it was intensified and
  moved towards to northeast, caused important
  economic damages.
• SYNOPTIC DESCRIPTION

Middle-Upper levels Surface
CUT-OFF low over Cadiz Gulf area. High pressure on the Southern French coast.
Cloudy bands by S and SE. Convective cells in Mediterranean area Moist and warm Eastern flow over the Spanish Mediterranean coast.
12
Radar Data

• Murcia Radar (1270m over sea surface level)
• Horizontal Structure
• Great cell located at 35-50 Km moving as far as
  120 Km.
• Vertical Structure

• Actual images

2100
2130
2200
2230
13
Comparison of reflectivities

• 10-Oct-2001 at 00 UTC (from HH21 to HH26)
• 3 resolutions (3 experiments)
• 1) HIR (0.2 degs)
• 2) RAD (0.16 degs)
• 3) RAH (0.05 degs)

14
4.2.1 HIR (0.2 degs ? 22.2 Km)

• Simulated reflectivity Actual reflectivity
• H 21
• Cappi1

Radar image (2x2Km2)
15
4.2.2. RAD (0.16 degs ? 16.6Km)

• Simulated reflectivity Actual reflectivity
• H 23
• Cappi 3

2100
2200
2300
Radar image (2x2Km2)
Cell shifts Northwards
16
4.2.3. RAH (0.05 degs ? 5.5 Km)

• Simulated Reflectivity Actual
  Reflectivity

H 22 Cappi4
H 23 Cappi4
17

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.
• Case of study.
• 4.1. General description
• 4.2. Experiments
• 4.2.1 HIR (0.2)
• 4.2.2 RAD (0.16)
• 4.2.3 RAH (0.05)
• Results obtained.

18
5. Results obtained

• Z simulated lt Z observed ( the difference is
  about 0.30 dBZ. in the whole grid).
• Best adjustment of t-Student is for a value
  between (0.2 and 0.3).
• Occupation rate
• Its bigger for observed reflectivity.
• In the simulated reflectivity images the cell is
  located further South than in the actual image.

19

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.
• Case of study.
• 4.1. General description
• 4.2. Experiments
• 4.2.1 HIR (0.2)
• 4.2.2 RAD (0.16)
• 4.2.3 RAH (0.05)
• Results obtained.
• Conclusions.

20
6. Conclusions

• The more similar the horizontal resolution of
  Hirlam and the resolution of radar is, closer the
  simulated reflectivity can be, compared to the
  actual one.
• This model can be useful for nowcasting when RSM
  will be operational (real time).
• RSM will be useful in validation and verification
  of Hirlam operational model at INM.

21

• Hirlam model.
• Radars at INM.
• Radar Simulation Model.
• Case of study.
• 4.1. General description
• 4.2. Experiments
• 4.2.1 HIR (0.2)
• 4.2.2 RAD (0.16)
• 4.2.3 RAH (0.05)
• Results obtained.
• Conclusions.
• Future lines of work.

22
7. Future lines of work

• Parallelization of the code in the Cray-X1, to
  reduce the computer time.
• Use of RSM to assess hydrostatic Hirlam model and
  non hydrostatic MM5.
• Implementation of objectives scores of observed
  reflectivities in RSM to obtain them directly in
  the output.
• Use of new Hirlam convection scheme
  (Kain-Fristch/Rasch-Kristjansson) in case studies
  (vertical structure of reflectivities).

23
0.2 Area (? 22.2 Km)
Back.
24
0.16 Area (16.6 Km)
South Pole Coordinates LAT - 35 degs.
South Pole Coordinates LON - 15 degs.
Back
25
0.05 Area (5.5 Km)
South Pole Coordinates LAT -49.5 degs
South Pole Coordinates LON -6.0 degs
Back