WWRP/WCRP. S2S Project. Frédéric Vitart and Andrew Robertson

Size: px

Start display at page:

Download "WWRP/WCRP. S2S Project. Frédéric Vitart and Andrew Robertson"

Howard Preston
5 years ago
Views:

1 WWRP/WCRP S2S Project Frédéric Vitart and Andrew Robertson

2 Progress report

3 S2S Project 5-year project, started in Nov A 5-year extension (S2S Phase ) has been approved Project office: KMA/NIMR hosts the project office Trust Fund: Contributions from Australia and UK (Phase 1), Canada (Phase 2)

4 International Conferences on Subseasonal to Decadal Prediction September 2018 NCAR, Boulder, CO, USA

5 Second International Conference on Subseasonal to Seasonal Prediction (S2S) and Second International Conference on Seasonal to Decadal Prediction (S2D) Some highlights from the S2S conference: 347 Participants 224 Poster Presentations 144 Oral Presentations 92 Early Career Scientists 38 Countries Hundreds of remote users Good representation of the basic state is required for teleconnections between the MJO and North Atlantic to be correct. Increasing evidence of time scale interactions: MJO teleconnections modulated by ENSO, QBO/SSW, QBO/tropical convection. The stratosphere is a valuable source of predictability on S2S timescales. However models used for S2S forecast still have issues capturing stratosphere processes and stratosphere-troposphere interaction. Some operational centres are moving towards a unified, coupled forecast system that can work across timescales from days (or shorter) to seasons (or longer). Low order empirical models can provide valuable insight into S2S predictability and can be used as benchmarks for operational forecasts. Multimodel ensembling shown to have higher skill than individual models Weather regime analysis and teleconnection patterns provide a bridge between large-scale sources of predictability and impacts such as surface temperatures, heat/cold waves, and atmospheric rivers. S2S forecasts in real-time are needed for application research, but not provided by S2S database. Applicability of S2S forecasts depends on prediction skill, which strongly depends on the target-user variable. Further case studies and evaluation are needed from the application context.

by Andrew Robertson and Frédéric Vitart Contains 22 chapters on S2S sources of

6 S2S book "THE GAP BETWEEN WEATHER AND CLIMATE FORECASTING: SUB-SEASONAL TO SEASONAL PREDICTION to be published by Elsevier in October 2018 and co-edited by Andrew Robertson and Frédéric Vitart Contains 22 chapters on S2S sources of predictability, modelling and forecasting, applications. To be available in October 2018

8 Sub-seasonal to Seasonal (S2S) Prediction Project Phase 1 Teleconnections Sub-Projects Madden-Julian Oscillation Monsoons Africa Extremes Research Issues Predictability Teleconnection O-ACoupling Scaleinteractions Physicalprocesses Verification Modelling Issues Initialisation Ensemble generation Resolution O-A Coupling Systematic errors Multi-model combination S2S Database Needs & Applications Liaison with SERA (Working Group on Societal and Economic Research Applications)

and the role of the ocean and sea ice on the sub-seasonal forecasts; stratospheric processes; and understanding model systematic errors and error growth.

9 Gap Analysis To inform future plans, a questionnaire was circulated to the research, modelling and operational communities for feedback. Frequently mentioned gaps included: land-surface processes and initialization; ensemble generation, including initialization, perturbation methods and stochastic physics; coupled data assimilation and the role of the ocean and sea ice on the sub-seasonal forecasts; stratospheric processes; and understanding model systematic errors and error growth. Some of the database and operational gaps raised include: need for more convenient and faster access to popular suites of variables, including ensemble means, model climatologies, indices, and map displays; need for multi-model calibrated forecast product development; desire for more extensive re-forecast sets (number of years and ensemble members) for verification and forecast calibration, encouraging centres to harmonize re-forecasts; request for more ocean data including 3D fields, increased model horizontal and temporal resolution; and desire for real-time access.

10 The 2 nd phase ( ) will focus on S2S Database enhancement Add more variables, including ocean variables Additional models, more variables available 6-hourly Include derived products (e.g. model climatology) Research activities (incl. process oriented studies): MJO and Teleconnections Ocean initialization and processes Land initialization and processes Aerosols Ensemble generation Stratosphere Enhancing operational infrastructure and user applications

11 Enhancing operational infrastructure and user Applications Research to Operations (R2O) and S2S Forecast and Verification Products Development Pursue research for testing and developing methodologies for calibration, combination, verification and generation of forecast products Coordination with the relevant WMO technical commissions to define the standards and protocols for operational implementation and exchange of S2S forecasts Real-time Pilot for S2S Applications research & demonstrations Goal is to demonstrate the value of S2S forecasts to different GFCS sectors Real time pilot: Make some derived variables available close to real-time for a limited period of time, such as 1 year. Promote interdisciplinary research the development of Ready-Set-Go type S2S applications

12 Barriers to S2S Forecast Uptake: Stakeholder Mini-Survey For the applications/service/donors/wider stakeholder audience, a set of 8 semistructured interviews was carried out by SERA. The interviewees were stakeholders in agriculture (Australia, Uruguay), energy (Uruguay), transport (Canada), water management (Canada, USA), bushfire management (Australia), and humanitarian aid (global and Peru). They generally agreed that while the potential benefits of skilful S2S forecasts are high, several barriers hinder their realization, namely: Lack of accuracy/poor skill -high level of accuracy is required for many types of decision-making; Lack of post-processing -need for statistical post-processing techniques to calibrate forecast for reliable probabilities; Lack of forecast verification Challenges in interpretation of probabilities -a large share of users struggle to interpret probabilities and can have low expertise in risk management.

13 Real-Time Pilot S2S Data available in real-time to a limited set of application users and for a limited time (2 years maximum) Next steps: 1. Select a list of users. Could include existing S2S projects: S2S4E Africa Swift Heat Waves US hydro California Water South Africa multi-sector Rimes ACtoday(Columbia U) 2. Send letter to data providers. 3. Give access to real-time data from the providers who accepted. Could start end of 2019 at the earliest.

MJO and teleconnections Representation of teleconnections (and their modulation) in S2S models MJO Pattern Metrics for assessment of model teleconnections and diagnosing sources of errors in

14 MJO and teleconnections Representation of teleconnections (and their modulation) in S2S models MJO Pattern Metrics for assessment of model teleconnections and diagnosing sources of errors in teleconnections Skill for MJO teleconnections in model based on MJO characteristics (slow vs. fast) MJO Metrics: OLR Pattern correlation ERAI Phase 3 MJO pattern vs Z500 pattern [40E-140W, 15S-15N] R= % confidence Better MJO pattern Better teleconnection pattern Courtesy Jiabao Wang and Hyemi KIm R ~ 0.72 for U850, U250 PART ONE

15 Land initialization and processes ILSTSS2S (Y. Xue, T. Yao, A. Boone) land temperature initialization, GEWEX/GASS proposed project, approached S2S SnowGLACE(Y. Orsolini, J.-H. Jeong) WGSIP project, boreal winter focus, snow IC impacts; initially focused on accumulation (1 Nov ICs), now realizing spring snowmelt focus (1 Mar ICs).

16 Ocean initialization and processes Science questions: What ocean feedbacks directly influence sub-seasonal variability and prediction skill? How is S2S predictability influenced by pre-existing ocean state? How does low-frequency variability (e.g., ENSO, climate change) affect S2S predictability? How does ocean mean state drift impact S2S predictability? Needs Ocean variables in S2S database! OSE coordinated by A. Subramanian (UCSD) to test impact of ocean observing system on S2S forecasts Link with TPOS 2020.

17 Evaluating Aerosols Impacts on Numerical Weather and Subseasonal Prediction and S2S Evaluating aerosols impacts on Numerical Weather and Subseasonal Prediction 1 2 Ariane Frassoni, 3François Engelbrecht Frederic Vitart and Angela Benedetti, on behalf of WGNE 3 on behalf of S2S 1-CPTEC, Brazil; 2-CSIR, S. Africa, 3-ECMWF, England

18 Evaluating Aerosols Impacts on Numerical Weather and Subseasonal Prediction Goals of the Project This project aims to improve our understanding about the following questions: What are the current capabilities of NWP models to predict aerosols on S2S time scales? What are the current capabilities of NWP models to simulate aerosol impacts on weather and sub-seasonal prediction? Are the S2S air quality forecasts useful for impacts purposes? 2 Coordinated experiments: - 72h forecasts with limited area models Dust in Egypt, Smoke in Brazil and S. Africa (field Campaign) - S2S re-forecasts for dust (Spring) and biomass burning Possible partners: ECMWF, NASA, NOAA, JMA, CMA, UKMO Links with WMO/GAW

Sub-Seasonal dust forecast (May 2015) DUST AEROSOL OPTICAL DEPTH @ 550nm

19 Sub-Seasonal dust forecast (May 2015) DUST AEROSOL OPTICAL 550nm CRPSS of dust aerosol optical depth N Africa and C. Asia Benedetti and Vitart, 2018

20 Priority AAs

21 AA15 : Enhance access to services (observations, model output, data collection and pre-processing and global models) that require exceptional HPC and data handling, as an enabler for WWRP research S2S Database: 11 S2S models are available and real-time forecasts have been synchronized. S2S to make recommendation for synchronising re-forecasts (Phase 2). Possible reduction of the realtime delay from 3 weeks to 1 week. Model Updates: Several models have been updated in 2017/2018 (JMA/CNRM/ISAC). BoM new model (ACCESS-2) will be included in New S2S Models: Indian Meteorological Department to join the S2S database. NASA (already contributes to SubX) asked to join the S2S database. NCAR and NRL could also be interested to join. Add new ocean/sea ice variables in netcdf(ongoing) Tropical cyclone tracks MJO Indices computed from all the S2S models are now publicly available from ECMWF ftp server in addition to MJO indices. Sudden stratospheric warming and weather regime indices to be added in Development of a Github repository joint with NOAA MAPP (Included in some of the subproject plans)

22 Statistics > 1000 registered users Archive size: 84 Tbs ~300 Tbsof data delivered > 45 publications using S2S database October 29, 2014

Data download: Data download amount CMA S2S data service year amount of data 2015 0.3TB 2016 1.9TB 2017 3.1TB 2018 (Jan~Aug) 2.4TB Total:7.

23 Data download: Data download amount CMA S2S data service year amount of data TB TB TB 2018 (Jan~Aug) 2.4TB Total:7.7 TB User preference download top five centers NCEP, ECMWF, CMA, UKMO,JMA download top five parameters t, 2t, tp, gh, wind (V-velocity,U-velocity)

24 69 TB 174 registered users (Aug 2018) October 29, 2014

TIGGE-S2S Workshop April 2018 Workshop on Predictability, dynamics and applications research using the TIGGE and S2S ensembles ECMWF, Reading, 2-5

Predictability and Dynamics Prediction and Verification Multi-Model approaches to prediction Application Studies Registration and abstract

25 TIGGE-S2S Workshop April 2018 Workshop on Predictability, dynamics and applications research using the TIGGE and S2S ensembles ECMWF, Reading, 2-5 April 2019 To bring together the users and data providers of the TIGGE and S2S databases around the following themes: Database technical development Predictability and Dynamics Prediction and Verification Multi-Model approaches to prediction Application Studies Registration and abstract submission close on 30 Nov.

S2S database near-real time products ECMWF: http://www.

int/en/research/projects/s2s/charts/s2s/ S2S Museum at

26 S2S database near-real time products ECMWF: S2S Museum at University of Tsukuba, Japan (Mio Matsueda) NAO Index 8 Feb 2018 start date

27 S2S Linkage with WMO CBS NMHSs GPC GPC Institutes S2S Data Base (Near RT data + refcsts ) Data Portals (3weeks behind RT ) WMO LC- LRFMME (Subset in near real-time) Research & Application Communities RCCs RCOFs RCC GPC GPC NMHS S2S Producing Centres WMO Users

28 S2S Linkage with WMO CBS Prototype for real-time provision of MME S2S forecasts by WMO LC-LRFMME Variables: SST, T2m, precipitation, u200, v200, u850, and OLR Participating Models: ECMWF, UKMO, JMA, NCEP/CPC, KMA, ECCC and BoM Needsto beformalizedby CBS

30 AA1: Use diagnostic and verification information to identify capabilities and limitations in predictions of high-impact weather on time scales out to seasonal Case studies. Eastern Canada Flooding 2017 and its Sub-seasonal Predictions Accepted in Atmosphere-Ocean Chapter on Extreme events in S2S Book. Special issue on S2S extremes in Nature Climate and Atmospheric Science Tropical cyclone indices made available in ftp site Extreme Forecast Index charts available in near real-time at ECMWF website NOAA/MAPP S2S-TF projects (e.g. projects on tropical cyclones/atmospheric rivers in the S2S database). Special issue in GRL/JGR S2S session will take place at the IWTC-9 (Hawaii, Dec 2018) to encourage use of S2S database for Tropical cyclone research and discuss possible use of S2S real-time forecasts for tropical cyclone prediction. A main goal of the real-time pilot in S2S Phase 2 will be to demonstrate the value of S2S forecasts to different GFCS sectors and Ready-Set-Go type S2S applications.

31 AA4: Applications Humanitarian Sector: News articles on Applications of S2S Forecasts: From Disaster Early Warning to Early Action published in the S2S and BRACED (Building Resilience and Adaptation to Climate Extremes and Disasters) web sites. S2S training workshops and meetings with applications components Subset of S2S database at IRI to facilitate access for application development. S2S stakeholder survey Real-time pilot sub-project is proposed for S2S Phase 2 (S2S-SERA). This subproject will build upon the stakeholder survey carried out by SERA, to co-develop a set of demonstration projects in partnership with users spanning the GFCS priority areas, representing both developing and developed countries. S2S is collaborating with several application projects: W2-SIP (WISER), ForPAc (SHEAR), SWIFT (GCRF), S2S4E (Energy, Horizon 2020)

in West and East Africa S2S WP has three objectives identify sources of predictability for Africa on sub-seasonal timescales evaluate operational S2S prediction systems build capability in the UK and

32 GCRF African SWIFT - Science for Weather Information and Forecasting Techniques Project overview 7.8M programme funded by UK Global Challenges Research Fund to built capability for weather forecasting in Africa for hours to months 5 UK and 10 African partners including NHMS, RCCs and Universities in West and East Africa S2S WP has three objectives identify sources of predictability for Africa on sub-seasonal timescales evaluate operational S2S prediction systems build capability in the UK and Africa to carry out research to inform the development of operational forecast products on the sub-seasonal timescale for decision making across a range of sectors. As part of this WP we propose to run a year long S2S testbed This work was supported by UK Research and Innovation as part of the Global Challenges Research Fund, grant number NE/P021077/1.

33 General Matters

34 Conferences/Workshops AGU sessions on S2S (Dec 2018) -More than 70 abstracts -3 sessions EGU session on S2S (April 2019) TIGGE-S2S workshop at ECMWF (April 2019) S2S session at IUGG/IAMAS meeting in Montreal (July 2019)

35 Trainings July 2017, Paraguay: First South American School on Sub-Seasonal Predictability and Prediction(WMO, WCRP, WWRP, IRI, ) July 2018, Ecuador: 10 th International Training Workshop on Climate Variability and Prediction(NOAA) August 2018, Singapore: Capability Building Programme in S2S Prediction for Southeast Asia-ASEAN2(ASMC, WCRP, IRI, RIMES, UN-ESCAP). 3-7 December 2018, Guatemala: Central American and Caribbean Workshop on Sub-Seasonal Predictability of the Mid-Summer Drought (IRI, ICTP, UNESCO, CIAT/CCAFS)?? 2019, somewhere in the Caribbean: 11 th International Training Workshop on Climate Variability and Prediction(NOAA).?? Summer School. Looking at possibilities at NCAR or Canada.

months Operate the S2S mailing list : 421 subscribers in Sept 2018 (361 in Nov

37 S2S ICO WMO and KMA signed a MoU for a second five year term ICO Activities include: Design and maintenance of the S2S website Issue a Newsletter every 4 months Operate the S2S mailing list : 421 subscribers in Sept 2018 (361 in Nov 2017) Organize regular training courses on S2S products in collaboration with WMO LC-LRFMME and APCC.

38 Membership

39 Some Links with other groups CBS : An important link has been established between S2S and thelead Center for Long-range forecast Multi-model Ensemble (LC-LRFMME) at KMA for the provisionof multi-model real-time SS2S forecasts. WGNE: Common research activities with MJO-TF and aerosols SPARC: 10 projects have been launched by members of the SPARC community to analyse the S2S database for stratospheric studies. S2S-SNAP Stratosphere subproject in Phase 2. WGSIP: Collaboration in the organization of the S2D workshops (Sept 2018) PDEF: Organization of TIGGE-S2S workshop, ensemble generation sub-project SERA: Organization of a S2S stakeholder survey. Real-time pilot project.

40 Issues Trust fund for Phase 2 (Only Canada has committed so far). Financial support for the maintenance and further development of the S2S (as well as TIGGE) database at ECMWF. Same for S2S at the IRI DL. Commitment from NOAA (~50k a year) but money transfer protocol needs to be sorted (through S2S trust fund?)

41 Thank you Merci

MJO prediction Intercomparison using the S2S Database Frédéric Vitart (ECMWF)

MJO prediction Intercomparison using the S2S Database Frédéric Vitart (ECMWF) Slide 1 WGNE Meeting 29 April 2016 1 INDEX The S2S project and S2S Database MJO prediction in S2S models MJO teleconnections