Transcript ｽﾗｲﾄﾞ ﾀｲﾄﾙなし

TYPHOON COMMITTEE Integrated Workshop on Urban Flood Risk
Management in a Changing Climate: Sustainable and Adaptation Challenges
Macao, China 06-10 September 2010
QPE/QPF of JMA
Application of Radar Data
Masashi KUNITSUGU
Head, National Typhoon Center
Japan Meteorological Agency
JMA
Japan
Meteorological
Agency
Purpose of QPE/QPF
JMA
Observation data
Products of QPE/QPF
High resolution data
(spatial, time, intensity)
local meteorological office,
local government ,
mass media,
weather companies
Weather
Information
JMA
Japan
Meteorological
Agency
Products of QPE/QPF of JMA
*Composite of echo intensity (every 30 min)
*Composite of echo top height (every 30 min)
* Radar/Raingauge-Analyzed Precipitation (every 30 min)
*Very-Short-Range-Forecast(VSRF) of precipitation
(every 30 min)
（Application）
*Analyzed 10-min precipitation (every 5 min)
*Precipitation Nowcasts (every 10 min)
* Soil Water Index
* Runoff Index
Technical base of QPE/QPF
Digitalized radar
(automatic rejection
of ground clutter)
Raingauge
network
Communication
network with rapid
transmission
High
performance
computer
NWP model with
high resolution
JMA
Japan
Meteorological
Agency
Precipitation observation equipment
Raingauges
Raingauges
Advantages
Disadvantages
Radar
Radar
Can measure actual amounts of
precipitation.
Can observe large areas with higher
spatial resolution than the raingauge
network.
Can observe precipitation at single May produce readings different from
precipitation observed on the ground, as
points only.
it measures the amount of rain overhead.
JMA
Japan
Meteorological
Agency
Radar/Raingauge-Analyzed Precipitation
○Calibrate ‘radar estimate’ with
raingauge data
precipitation
Radar
estimate
Raingauge data
JMA
Japan
Meteorological
Agency
Radar/Raingauge-Analyzed Precipitation
○Calibrated ‘radar estimate’ with
raingauge data are more accurate
precipitation
Analyzed
precipitation
Calibration
×
factor
Radar
estimate
Raingauge data
JMA
Japan
Meteorological
Agency
Radar/Raingauge-Analyzed Precipitation
process
Qualification of radar data
*remove ‘false echo’ like ground clutter
Calibration over the entire radar
detection range
Calibration over land
*modification of calibration factor
Composition
*using maximum value method
*replace with raingauge data
Radar/Raingauge-Analyzed precipitation
JMA
Japan
Meteorological
Agency
Calibration over the entire radar
detection range
○Conditions in calibration
（1）calibration factor is a function of beam
height and is calculated for each radar every
time
（2）analyzed precipitation estimates of each
radar in the area where multiple radars
overlap should be equal
（3） analyzed precipitation estimates should
be equal to the raingauge precipitation
JMA
Japan
Meteorological
Agency
Calibration over the entire radar detection range (1)
E1(x,y)=F1(x,y)E0(x,y)
F1(x,y)= Fa(x,y)｛1+Fx(x,y)・H(x,y)2｝
calculation of Fx
Calibration over the entire radar detection range (2)
E1(x,y)=F1(x,y)E0(x,y)
F1(x,y)= Fa(x,y)｛1+Fx(x,y)・H(x,y)2｝
calculation of Fa
Calibration over land (modification)
① modify calibrated estimates on raingauge
grids
Calibrated
estimates
Raingauge
precipitation
R(i) = C2(i)×E1(i)
raingauge grid
R(i) ：raingauge precip.
E1(i) ：calibrated estimates
C2(i) ：factor to modify
estimates
＊for all the raingauge grids
of the radar
JMA
Japan
Meteorological
Agency
Calibration over land (modification)
② Determine factors of all grids over land
Radar precipitation
Interpolate factors of
raingauge grids to a target
grid with weights.
＊weight＊
W(i)=W1(i)×W2(i)
W1 (i) ： for distance
・ raingauge
grids
■ target grid
■ factor of a target grid
＝ ∑W(i)C2(i)
W2 (i) ： for rain intensity
and beam attenuation
i : raingauge
number
Radar/Raingauge-Analyzed Precipitation
1-hour radar precipitation
raingauge precipitation
Radar/Raingauge-Analyzed Precipitation
Radar precipitation
Calibrated precipitation
Radar/Raingauge-Analyzed Precipitation
Composite
calibrated
radar data
Flowchart of VSRF
Extrapolation of precipitation with
orographic effect(EX6)
process
Merge EX6 and MSM(MRG)
VSRF of precipitation
up to 6 hours, spatial resolution 1km
JMA
Japan
Meteorological
Agency
Extrapolation method（EX6）
･Extrapolation method is effective up to ~ 3
forecast hours
「extrapolation」
after 1 hour
Move it with
the same speed
and the same direction
as it moved.
now
Non-linear extrapolation
was introduced
in 2006.
before 1 hour
JMA
Japan
Meteorological
Agency
Data used for making VSRF
Outputs of Numerical Weather Prediction
*MSM(mesoscale model) operation:8 times a day
*wind（700･900hPa） *temperature(900hPa)
*relative humidity(900hPa) *precipitation(surface)
Topography data
JMA
Japan
Meteorological
Agency
Precipitation enhanced by orographic
effect
○Estimate precipitation due to updraft along a
mountain
･MSM(900hPa）temperature, wind
･Analyzed precipitation
Pre-existing seeder cloud
Moist
lowlevel
flow
Feeder
cloud
seeder-feeder model
（Browning and Hill，1981）
Stationary part
of precipitation
over mountains
caused by
*cold air
outbreak across
the sea
*low pressure
JMA
Japan
Meteorological
Agency
Enhancement and dissipation of
precipitation by orographic effect
○enhancement
Low level wind
○dissipation
Mountain
(Pobability
of the)
Orographic
precipitation
Movement of a
echo
Orographic
enhancement
JMA
Japan
Meteorological
Agency
Enhancement and dissipation
of precipitation by orographic
effect
(1) Enhancement
Precipitable amount by
orographic effect
Difference caused by orographic effect
Without orographic effect
With orographic effect
Enhancement and dissipation of
precipitation by orographic effect
(2) Dissipation
Difference caused by orographic effect
Without orographic effect
With orographic effect
Merging method
Merging ratio considering space
Compare the accuracy of EX6/MSM
○ using pattern distance
MRG
Merging ratio considering time
Increase the ratio of MSM as
forecast time goes
Merge
○ every 10 minute up to 6 hours
JMA
Japan
Meteorological
Agency
Merging method
Compare the accuracy of EX6/MSM
○ using pattern distance
MRG
Calculate the reliability “r” of MSM
○ indicate the merging ratio
Calculate the merging ratio R(t)
○ using reliability r and weight function C(t)
Merge
○ every 10 minute up to 6 hours
JMA
Japan
Meteorological
Agency
Radar/raingauge precipitaion analysis
at initial time（upper left）
・EX6 FT=3 of the initial time 3 hours
before（lower right）
・latest MSM（lower left）
RA
MSM
calculate the reliability of MSM
comparing the similarity
EX6
ratio of merge considering time
reliability of MSM＝0.5
reliability of MSM＝１
FT
ratio of EX6 at forecast time
the larger the r of MSM, the smaller the ratio of EX6
red line : lower limit of the ratio of EX6 : C(t)
Ratio of MSM = r ･{1 － C(t)}
Examples of VSRF(Merging method)
Top left R-A, top right MSM, bottom left EX6, and
bottom right MRG at 11JST 31 May 2000 (6-hour
forecast). MRG could predict the extending western
rain systems.
Same as the left figure, at 15JST 31 May 2000(6-hour
forecast). MRG could predict isolated convective rain
systems.
JMA
Japan
Meteorological
Agency
Precipitation Nowcasts
Forecast 10 minute precipitation amount
up to 60 minutes every 10 minute
※ with extrapolation method
res.=1km, forecast every 10min
Disseminate within 3 minutes of
observation time
1) Previous calibration factors (10 min
before) are used.
2) Use the movement derived by VSRF
3) Only dissipation by orographic effect
is introduced.
10minutes rainfall
initial = 0730JST 06 June 2004
Thank you for your
attention!
JMA
Japan
Meteorological
Agency