Radiosoundings and AMDARS for HYMEX SOP TT01a A. Jansa, S Klink, P. Arbogast 26 may 2010

Radiosoundings and AMDARS

for HYMEX SOP TT01a

A. Jansa, S Klink, P. Arbogast 26 may 2010

2

TT01a

1. Routine Networks and observations dedicated to HYMEX2. Use of these networks following the “adaptive observation” strategy 3. Scientific goals addressed with these observing systems4. Recommendations for SOP, EOP5. To be done

3

1. Radiosoundings and AMDARS; Routine Networks and obs dedicated to HYMEX

Commercial aircraft data (AMDARs ACARs): Provide a significant amount of information Significant reduction of analysis uncertainty over Spain… (When

using KFS sensitivity, not shown)

Source: Météo-France

4

Radiosoundings (06 and 18UTC)

from 31 October to 13 December 2009 (DTS MEDEX)

. Round “dots” are radio-sounding targetable stations; their label indicates the WMO-identifier followed by the number of extra-sounding requested (deployed or not) during the sub-period. The coloured contours indicate the frequency of use of the targetable stations; this can be seen as a proxy of the mean sensitivity (mostly the sensitivity as computed from the ECMWF). The pink squares that located in the center of 2x2 lat x lon º squares (43ºN-03ºW, 43ºN-01ºW and so on) indicate the frequency of centres of target areas (the size of the square is proportional to this frequency in the sub-period). The “gravity centre” of target areas for the sub-period is indicated as a violet star.

Routine network at 06 and 18UTC

+ 2 ships and 2 ground stations (Lampedusa, Montpellier)

Source: AEMet

5

2. Adaptive observation strategy

1. Frequency of AMDAR observations: increase to hourly (instead of operational just 3-hourly) profiles at international airports close to the Mediterranean Sea

2. Time frequency of Radiosoundings from a pre-defined list

3. At 06 UTC and 18 UTC (since the potential for observation network is larger than 00 and 12UTC)

4. Activation following results from sensitivity computation using adjoint models and ensemble. (Metrics do ignore physics and Data Assimilation System failures.)

5. Potential for financial support by EUCOS (EUMETNET), exact amount to be negotiated (depends on EUCOS budget)

6. Previous experiments concerning Europe and the Mediterranean area: DTS PREVIEW (2008) and DTS MEDEX (Autumn 2009)

6

3. Adaptive observation strategy

Real time J-10 J-9 J-8 J-7 J-6 J-5 J-4 J-3 J-2 J-1 J

EARLY WARNING

FORECAST TO BE OPTIMISEDDELAI

INCOMPRESSIBLE POUR l’ACTIVATION DES OBS.

DTS ACTIF

(Target Time)18 UTC

Valid Time (HPE or synoptic scale factors leading to HPEs)

7

Radiosoundings launched during DTS/DEDEX 2009 (whole autumn 2009)

The image shows the frequency of extra-sounding during the whole DTS-MEDEX-2009. This is the result of the human interpretation of the objectively computed sensitivities, taking into account, in addition, the list of "targetable" radiosounding stations. Some Spanish and Algerian radiosoundings are the most critical for improving the forecasting of high impact weather in the Mediterranean

8

3. Scientific issues ?

1. To lead to an improvement of the global forecast (say DX=10 to 30 km) ?

2. Do we expect an impact within mesoscale DA systems ? (need an OSE of DTS-MEDEX to see whether additional data have an impact)

3. Description of Dynamics associated with HPEs ?

4. Cross-validation of different observing systems (contribution by O. Bock fir GPS network) ?

5. Is there any interest in an extension of the Network over North-Africa ?

9

4. Recommendations for SOP, EOP, LOP

• LOP: regular operational RS and AMDAR network. It is suggested to proceed to the storage of high- resolution RS data

• EOP: adaptive observation strategy is manpower consuming…It is suggested to implement an extra RS site in a poorly observed area but found to be “sensitive”.

• Winter SOP : adaptive observation strategy suitable for that purpose

• SOP dedicated to HPE : put effort on high resolution Sensitive Area Prediction

10

5. To be done

1. Observing system Experiment for some HPE (CNRM, other Met services contributing…) list of cases taken from DTS PREVIEW and DTS MEDEX (contributors) ?

2. Science behind Sensitivity Area Computation (works on metrics, use of ensembles, LAM… ? )

3. Funding issues (again EUCOS ?)

4. Involvement of North-African countries ?