11/09/2015FINNISH METEOROLOGICAL INSTITUTE CARPE DIEM WP 7: FMI Progress Report Jarmo Koistinen, Heikki Pohjola Finnish Meteorological Institute.

04/19/23 FINNISH METEOROLOGICAL INSTITUTE

CARPE DIEMWP 7: FMI Progress Report

Jarmo Koistinen, Heikki Pohjola Finnish Meteorological Institute


· Area 1/WP3: Development of a variational assimilation scheme for doppler winds (FMI + SMHI, responsible persons at FMI Heikki Järvinen, Kirsti Salonen)

· Area 2/WP7: Advanced surface precipitation estimate from radars and a NWP model (HIRLAM) (Jarmo Koistinen, Heikki Pohjola)

· Objective: Improve the accuracy and quality of operational real time precipitation measurements


WP 7.3 Vertical reflectivity profile correction applying radars and

NWP

· New order:7.3 started first and continues during the whole 36 months. WPs 7.1-7.2 started later and feed 7.3.

· Operational method has been created and validated (preliminary presentations at ERAD and AMS Radar Conferences, Heikki’s MSc), a final paper should be written. Remaining problem: OP i.e. WP 7.2.

· Results applied in the GPM-project validation planning


1(2)

Measured vertical profile of reflectivity

• 7 C-band radars

• every 15 minutes• layer thickness 200 m• range 2 - 40 km• max bin count 5000 / layer


1(2)

Statistics of VPR from March 2001 to August 2003 (556 471 profiles)


Climatological profiles based on precipitation profiles March 2001 to

August 2003


1(2)

VPR correction ),(

),0(log10

rhZ

rZS

e

e


1(2)

Yearly average VPR correction


1(2)

Validation of the VPR correction with overlapping radars

Measurement area 5 – 15 km

1. Precipitation profile measured at Radar 2

2. > 50 measurement bins in the same location diagnosed as precipitation (max. 10020)

3. Difference in those bins (Rad 2 – Rad 1) tells under/over estimation at the distance Rad 2

4. Average bias and standard deviation calculated for distances 141 km, 180 km, 193 km, 198 km and 219 km.


Radar pair Anjalankoski-Vantaa (Vantaa-Anjalankoski) Feb – May 2003, distance

141 km


Radar pair Korppoo-Vantaa (Vantaa-Korppoo) Feb – May 2003, distance 180

km


Radar pair Ikaalinen-Vantaa (Vantaa-Ikaalinen) Feb – May 2003, distance 193

km


Radar pair Ikaalinen-Korppoo (Korppoo-Ikaalinen) Feb – May 2003, distance 198

km


Radar pair Kuopio-Utajärvi (Utajärvi-Kuopio) Feb – May 2003, distance 219 km


Radar pair Vantaa-Ikaalinen (Ikaalinen-Vantaa) Jun – Aug 2003, distance 193 km

•Calibration difference on Vantaa (+ 2-3 dB) and lowest elavation angle 0.8 instead of 0.4 on Ikaalinen because of technical reason (-2 dB)


Correction with different measurement angles

Snow profile Rain profile with bright band

2 dB


Nordrad Quality Assurance project

• Calibration difference for radar pair Vantaa Ikaalinen, period Jun – Aug 2003


WP 7.1: Attenuation correction based on 3D water phase diagnosis from NWP model quantities

Deliverables: Large attenuation statistics for rain-only (assumed in most existing radar systems) and variable water phase statisticsImprovement in surface precipitationOverestimation of attenuation in hail avoided by applying a hail algorithm


Calculation of attenuation

· Simple one: all precipitation is liquid, k(dB/km)=1.12*10-4 Ze

0.62 · Advanced one: freezing level height fixed at each

radar, obtained from NWP and VPR, attenuation calculated separately in each water phase layer

· We are testing Doppler data from elevation angle 90 degrees in the polar volume scans to improve the radar-based detection of freezing level

· Formula: rain (see above), dry snow (in Battan), bright band ?


WP 7.2 Elimination of overhanging precipitation (OP) from surface estimates

Altostratus: present in 15 % of all VPR in Finland in 2001


Diagnosis of OP in the network

· Apply precipitation base height field from the radar network

· Apply (an idea so far): nowcast movement of theVPR, measured above the radars with the analysed motion vector field (operationally available)

· Apply precipitating layers from NWP. First step: compare model profiles from HIRLAM with VPR from radars. Available quantities e.g. RH, and cloud condensate mixing ratio.


Example of OP case 9.3. 2002 02 UTC


9.3. 2002 03 UTC


9.3. 2002 04 UTC


9.3. 2003 05 UTC


9.3. 2003 06 UTC


9.3. 2003 07 UTC


1(2)

Overhanging precipitation 9.3. 2002 at Anjalankoski radar


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


1(2)


HIRLAM model relative humidity 09.03. 2003


HIRLAM model freezing level height 21.8. 2003 17 UTC


1(2)

24 h accumulated precipitation Nov 8, 2002, 23 UTC No VPR correction No VPR correction


1(2)

24 h accumulated precipitation Nov 7, 2002, 14 UTC No VPR correction With VPR correction


1(2)

24 h accumulated precipitation Dec 21, 2002, 21 UTC No VPR correction No VPR correction


1(2)

VPR correction in the network


1(2)

Time averaging of the single radar VPR correction

t

httt

cp

ccpp www

wCwCC

6

h

twt 6

1


Radar pair Anjalankoski-Kuopio (Kuopio-Anjalankoski) Feb – May 2003, distance

219 km


1(2)

11/09/2015FINNISH METEOROLOGICAL INSTITUTE CARPE DIEM WP 7: FMI Progress Report Jarmo Koistinen, Heikki Pohjola Finnish Meteorological Institute.

Documents

db slide

layer slide

average vpr correction

distance rad

precipitation profiles

ams radar conferences

statistics of vpr

measurement bins