1. Report from the Sixth WMO Workshop on the Impact of Various Observing System on NWP
Lars Peter Riishojgaard, WMO
Scientific Organizing Committee:
Erik Andersson, Co-Chair, ECMWF; Yoshiaki Sato, Co-Chair, JMA
Carla Cardinali, ECMWF; John Eyre, Met Office; Ron Gelaro, NASA
Jianjie Wang, CMA; Jochen Dibbern (DWD)
Thibaut Montmerle; Météo-France
Lars Peter Riishojgaard, WMO; Wenjian Zhang, WMO

2. WMO Workshops on the Impact of Various Observing Systems on NWP
1st – Geneva, 1997
2nd – Toulouse, 2000
3rd – Alpbach, 2004
4th – Geneva, 2008
5th – Sedona 2012
6th – Shanghai, May
Impact-workshop_Shanghai-May2016.html
Extremely important for WIGOS due to their role in the WMO Rolling Review of Requirements, WMO Impact Workshops bring together major NWP centers, research community and other stakeholders to discuss the contribution to forecast skill of various WIGOS/GOS elements; guidance to participants provided well in advance of Workshop

3. The 6th WMO Impact Workshop in brief
Held at the Shanghai Meteorological Center HQ May at the kind invitation of the government of China
Largest WMO Impact Workshop so far:
4 days; 80 participants;14 countries
43 oral presentations, 41 poster presentations distributed in three sessions
Ample discussion time during and after the sessions
Very broad attendance from NWP community; space agencies, observing system managers also represented, as well as private sector
Growing awareness of the power of NWP output to support decision making, not only for weather but also for observing system design
Importance to NWP of free and open data sharing was emphasized by many participants, in response to recent shift by private sector toward selling observational data rather than observing systems

4. Science Questions for Shanghai (distributed to participants 18 months ahead of the Workshop)
S1Marine: Surface pressure over ocean
S2AMDAR: Coverage of AMDAR
S3Radar: Radar observations (impact)
S4Strat: In situ observations of the stratosphere (requirements)
S5PBL: Observations of the PBL for regional and high resolution NWP (requirements)
S6SatLand: Satellite sounding over land and ice (methodology and impact)
S7Sounders: Impact of multiple satellite sounders (orbital spacing, question of redundancy)
S8AMVs: Atmospheric Motion Vectors (impact and product generation)
S9UA: Regional upper-air network design studies (including radiosonde launch schedules)
S10AdjEns: Application of adjoint and ensemble methods (impact assessment methodology)
S11Ocean: Impact in ocean coupled assimilation (no such studies presented)
S12Land: Impact in land coupled assimilation
S13 Time frequency What is the required time frequency of observations?
S14 Atmospheric composition (impact and observational data requirements)
S15 OSSEs Observing system simulation experiments (future satellite systems, orbit configuration scenarios, etc)

5. Key messages from Shanghai
Less evidence of resilience/robustness/redundance of the overall Global Observing System than reported at the 5th Workshop in Sedona in 2012
Addition of observations almost always leads to improvements in skill
Addition of observations from secondary (operational back-up) satellites often leads to significant improvement in skill beyond the baseline achieved using the primary satellites
Positive impact now also of some sensors for which this was difficult to demonstrate in the past (e.g. scatterometers, wind profilers, MISR winds, …)
Regional/mesoscale/convective scale data assimilation and NWP has made significant progress in the direct use of observational data

6. A FEW SAMPLE RESULTS FROM SHANGHAI

7. Daily data volumes (assimilated)
13 M/day
Hunger of NWP systems for observational data continues; volume of observations assimilated into models keeps increasing; now ≈107 observations per day
(slide from Environment Canada)
AMSU-A: NOAA , METOP A/B, AQUA; AIRS & IASI A/B: 142 channels each, CrIS: 103 channels.

8. Impact on 24h forecast error (FSOI)
Increasing FSOI as more AMV data assimilated
Jan-Mar 2012
Apr-July 2013
May 2014
6.7%
8.6%
4.3%
Most impact results reported in the form of FSOI diagnostics; contribution of individual observing systems to the reduction of 24-h forecast error
(slide from Met Office, UK)
2 hr temporal thinning of AMVs – 2-3x volume
+ low level Met-10
+ GOES hourly
The figure on the left shows the impacts from around the time of the last Winds Workshop in 2012.
The Geo AMVs sit in 8th place on 4.3%, behind the scatterometers.
By mid 2013, the Geo AMVs have jumped up to 5th (6.7%) ahead of Scat, AIRS and SYNOP
AVHRR slightly ahead of MODIS winds due to addition of Metop-B (note: have also since lost Terra WV channel winds)
Contributions to the total observation impact on a moist 24-hour forecast-error energy-norm, surface-150 hPa (Richard Marriott and James Cotton)
© Crown copyright Met Office

9. 500 hPa Northern Hemisphere AC scores for 20140101 – 20140131 00Z
Also classical Observing System Impact (OSE) experiments; data addition/denial studies; valuable complement although increasingly difficult to diagnose due to volume of observations and observing systems
(slide from NCEP)
Infrared
Microwave
WMO 6th Workshop

10. 500 hPa Southern Hemisphere AC scores for 20140101 – 20140131 00Z
Southern hemisphere shows greater sensitivity to satellite data due to lack of conventional data
(slide from NCEP)
Infrared
Microwave
WMO 6th Workshop

11. Observation fits – global Normalised change in std. dev. of FG dep.
All-sky microwave WV
Clear-sky microwave WV
100% = control (no microwave WV)
AMSU-A
GPSRO
SATOB (AMVs)
Encouraging progress in the assimilation of microwave satellite data over cloudy and precipitating areas
(slide from ECMWF)
Radiosonde T
Radiosonde q
Conventional wind
Improvement
Improvement
Improvement
6th WMO Workshop on the Impact of Various Observing Systems

12. Assimilating Radar Reflectivity Data
3D Reflectivity:
volume reflectivity measurements (10 PPI scans) of 17 C-Band radar stations
Radar forward operator:
Simulate synthetic 3D radar scan based on COSMO-DE model fields (EMVORADO)
No-reflectivity: threshold data at 5 dBZ
Superobbing: achieve homogeneous horizontal data distribution:
Assimilation experiment (20 – 29 May 2015)
Conv. Obs. Only
Conv obs + LHN
Conv. Obs + radar reflectivietes

13. ensemble mean and std dev.
high-resolution obs: radar reflectivity
 Theresa Bick et al., QJRMS 2016
7 days / 29 forecasts
(22 – 29 May 2014)
FSS precipitation
ensemble mean and std dev.
precip
precip
CONV
CONV + RAD
CONV + LHN
CONV + RAD
Beginning to see impact of the assimilation radar data; long–lasting impact on quality of precipitation forecasts
(slide from DWD)
rather large, long-lived positive impact from use of radar reflectivity in LETKF
use of radar reflectivity slighltly more impact than Latent heat nudging in first 4 hours

14. Mean daily precipitation pattern from June to August 2013~2014
OBS
ECMWF
JMA
WRF_NJU
NCEP
CMA
Coastal
rainband
Meiyu rainband
XiNan
Very encouraging progress in the local data impact of (conventional) observations assimilated into limited area models
(slide from CMA)
Here is the mean daily precipitation from June to August 2013 to The left upper panel is the rain guage observations and the right upper panel is our real time forecasts. We could see the general pattern are well simuluated. The dry and wet area are well distingished. And also the three main heavy rain centeral are well predicted. The other four panels are the global forecast. Although the global model also predicted heavy rain central. But the pattern and intensity are not as good as the convective scale model. The other disadvantage is the global model failed to distinguishe dry and wer region. For an example, ECWMF overestimated the intensity of dry region

15. Forecast Sensitivity - Observation Impact (FSOI)
Inter-comparison Experiment
Tom Auligné, Joint Center for Satellite Data Assimilation (JCSDA)
Ron Gelaro, NASA, Global Modeling and Assimilation Office (GMAO)
Rahul Mahajan, David Groff, NOAA, National Weather Service (NWS)
Rolf Langland, Naval Research Laboratory (NRL)
Jianjun Liu, NOAA’s Satellite and Information Service ( NESDIS)
James Cotton, Larry Morgan, UK Met Office
Yoichiro Ota, Japan Meterological Agency (JMA)
Coordinated data impact studies by multiple NWP centers using same observing system classification and same diagnostics; direct result of OPAG-IOS guidance!
(slide from JCSDA)

16. Fractional Impact at 00UTC: Satellite Radiances
Polar microwave and hyperspectral IR most important satellite radiance data
(slide from JCSDA)

17. Fractional Impact at 00UTC: Other Observations
RAOBS, Aircraft observations, AMVs are most valuable non-radiance observations
(slide from JCSDA)

18. Formal output from the 6th WMO Impact Workshop
Final Report from the Workshop, published as a WIGOS Report, includes
Brief (10-page) Summary Report, including 23 formal recommendations, mostly addressed to
WMO (Commission for Basic Systems)
Space agencies
Numerical Weather Prediction community
All oral presentations from the meeting
(In principle) all poster presentations – however, many authors did not deliver these yet

19. A few highlights from the report
“In terms of the most important observing systems contributing to forecast skill of global NWP models, the top five system were, in no particular order, microwave sounders (AMSU-A, ATMS), hyperspectral infrared sounders (AIRS, IASI, CrIS), radiosondes, aircraft data and atmospheric motion vectors (AMVs).”
“For some data assimilation systems using very short observation cut-off times, aircraft observations taken during ascent/descent are the most important data, together with radiosondes and geostationary satellite data products. Over the Arctic regions, buoys and LEO satellite observations are essential, but additional in situ vertically resolved data (soundings) are needed.”
“The positive impact of aircraft observations, both flight-level and profile data, on skill continues to be very substantial. […]”
”Currently the radiosondes are launched simultaneously (primarily at 00 UTC and 12 UTC) […] However, both for climate applications and for certain local forecaster applications, some flexibility in the launch schedule may be desirable.”

20. Key recommendations addressed to WMO (in most cases this means to CBS in particular)
Recommendation 7; WMO is strongly encouraged to investigate and publicize the benefits of aircraft observations in general and of humidity observations in particular, in order to help sustain and further expand the AMDAR program.
Recommendation 8; WMO to develop specific alternative scenarios for radiosonde launch schedules; NWP centres were encouraged to perform data impact experiments for such scenarios
 Recommendation 10; All data providers are encouraged to continue to share all observations internationally, especially those observations that are essential for numerical weather prediction, e.g. all GNSS-RO soundings.

21. Key recommendations addressed to WMO (II)
Recommendation 12; WMO to articulate the requirement for international sharing of all observations used in NWP systems, e.g. via the new RBON (Regional Basic Observing Network) development; data providers (including NMHSs and space agencies) are encouraged to make these data available to all NWP centers.
Recommendation 13; NWP community to carry out impact studies regarding proposed observing systems for high-resolution sensing of the atmospheric boundary layer; WMO to document and record the requirements for such observations through its Rolling Review of Requirements.
Recommendation 16; WMO is encouraged to continue and strengthen its efforts on the development of protocols and formats for both national and international exchange of weather radar data.

22. Key recommendations addressed to WMO (III)
Recommendation 17; WMO and the Global Cryosphere Watch (GCW) to investigate possibilities of obtaining additional soundings over the Arctic.
Recommendation 18; Further investigation into the use of impact studies for the design of climate observing networks is encouraged.
Recommendation 19; Longer range (seasonal to decadal range prediction) observation impact studies are encouraged.
Recommendation 21; WMO is strongly encouraged to foster further coordination between impact studies undertaken by different NWP centers; this should extend also to issues such as common methodologies and diagnostics and common quantities used when presenting the results.

23. Next steps
Incorporate other relevant Workshop recommendations in the work programmes of Technical Commissions, ICG-WIGOS and its task teams and other relevant elements of WMO working structure
Decision on 7th WMO Impact Workshop in 2020 to be made by EC-71 in 2018; CBS Recommendation would be appropriate

24. Thank you
Merci