The TIGGE experience Richard Swinbank and Masayuki Kyouda

The TIGGE experience Richard Swinbank and Masayuki Kyouda with thanks to members of GIFS-TIGGE WG & THORPEX IPO S 2 S project planning group meeting, Exeter, February 2013

The TIGGE experience Ø Ø Ø Aims Database content Technical – how we run TIGGE-based products Looking ahead, post THORPEX

TIGGE THORPEX Interactive Grand Global Ensemble Ø A major component of THORPEX: a WMO World Weather Research Programme to accelerate the improvements in the accuracy of 1 -day to 2 -week highimpact weather forecasts Ø GEO task WE-01 -C 1 – “Global Multi-Model Prediction System for High-Impact Weather” Objectives: Ø Enhance collaboration on ensemble prediction, both internationally and between operational centres & universities. Ø Facilitate research on ensemble prediction methods, especially methods to combine ensembles and to correct systematic errors Ø Enable evolution towards a prototype operational system, the “Global Interactive Forecast System”

TIGGE data Ø Ten of the leading global forecast centres are providing regular ensemble predictions to support research on predictability, dynamical processes and the development of probabilistic forecasting methods. Ø TIGGE data is made available for research after a 48 hour delay. Ø The TIGGE dataset has proved invaluable as a resource for a wide range of research projects. Over 70 articles related to TIGGE have been published in the scientific literature

Summary of TIGGE database Centre Ensemble members Output data resolution Forecast length Forecasts per day Fields (out of 73) Start date Bo. M* 33 1. 50º x 1. 50º 10 day 2 55 3 Sep 07 CMA 15 0. 56º x 0. 56º 10 day 2 60 15 May 07 MSC 21 0. 9º x 0. 9º 16 day 2 56 3 Oct 07 CPTEC 15 0. 94º x 0. 94º 15 day 2 55 1 Feb 08 ECMWF 51 N 200 (Reduced Gaussian) N 128 after day 10 15 day 2 70 1 Oct 06 JMA 51 1. 25º x 1. 25º 9 day 1 61 1 Oct 06 KMA 24 0. 56º x 0. 38º 10 day 2 46 28 Dec 07 Météo-France 35 1. 50º x 1. 50º 4. 5 day 2 62 25 Oct 07 NCEP 21 1. 00º x 1. 00º 16 day 4 69 5 Mar 07 UKMO 24 0. 83º x 0. 55º 15 day 2 72 1 Oct 06 * Delivery of Bo. M data currently suspended

TIGGE data flows Ø The ensemble prediction data is transferred from the data providers to one of the data centres (using LDM, FTP or HTTP). Ø After checking, the data is then sent on to the other data centres. Ø The data is archived and made available to users 48 hours after initial forecast time.

TIGGE features Predictability Ø All data are archived at Applications science native resolution (on native grid when possible) other academic NHMS users Ø Data may be interpolated on any limited-area lat-lon grid defined by the user just before download Ø Field names, definitions, units, accumulation times, NCAR ECMWF CMA (etc. ) are fully standardized Ø Data gaps are continuously monitored and every effort is made to repair them quickly Ø All data provided in GRIB 2 EPS 1 EPS 2 EPS n (WMO standard data format) AMS - IIPS - 2012 Slide 7

TIGGE Archive Ø The TIGGE database now contains over five years of global forecast data. Ø Data volume now exceeds 520 terabytes (2. 6 billion fields). Ø Around 100 users access TIGGE data every month.

TIGGE – Research focus The TIGGE project (& GIFS-TIGGE working group) was originally focused on the technical work enabling the establishment of the TIGGE archive dataset. The main scientific aspects of THORPEX were covered by Predictability & Dynamical Processes WG, and also Data Assimilation & Observing Systems WG(s). Following the successful establishment of the TIGGE dataset, the focus of the GIFS-TIGGE working group has shifted towards research on ensemble forecasting. Particular topics of interest include: Ø a posteriori calibration of ensemble forecasts; Ø combination of ensembles produced by multiple models; Ø research on and development of probabilistic forecast products. More recently, we have broadened our range of interest to cover improvement of EPSs, including: Ø representing uncertainties in initial conditions Ø stochastic physics & other techniques to account for model error

Evolution of TIGGE & GIFS TIGGE development ts n, tio bina uc d o r p om c ent , m n e o rov rati p b m i i l Ca EPS Time Ø The GIFS-TIGGE should also be a forum to focus on R&D directed at improving our EPS systems, to help us develop a “virtuous circle”. Ø In addition to post-processing and applications of EPS and EPS developments, we will also maintain an interest in ensemble verification and links with convective-scale EPS and the new subseasonal to seasonal group. Ø … to be considered further in post-THORPEX plans.

Towards the Global Interactive Forecast System (GIFS) Ø The THORPEX implementation plan had a grand vision for the development of a Global Interactive Forecast System. Although our plans are now more modest, one key element of GIFS that we are developing is the use of probabilistic forecast products. Ø Many weather forecast situations are low probability but high risk – unlikely but potentially catastrophic. Probabilistic forecasting is a powerful tool to improve early warning of high-impact events. Ø Using the TIGGE dataset, we are developing and evaluating probabilistic products to deliver improved forecasts of high-impact weather. Ø The products are focused on Ø Tropical cyclones Ø Heavy precipitation Ø Strong winds

Links with Severe Weather FDP Ø We are collaborating with WMO Severe Weather Forecast Demonstration Project (SWFDP) and other projects Ø to ensure that products address needs of operational forecasters and end users; Ø to provide an environment for the evaluation of prototype products. Ø GIFS will use global-regionalnational cascade pioneered by the SWFDP. Ø Web-enabled technology will be used for generation and distribution of products.

Tropical cyclones Ø The GIFS-TIGGE working group first set up a pilot project for the exchange of real-time tropical cyclone track predictions using “Cyclone XML” format, initially to support the THORPEX Pacific. Asia Regional Campaign (T-PARC) Ø Several types of products to support TC forecasting have been developed using those data. Example of combined TC track forecasts (Met Office + ECMWF)

Warnings of other severe weather events Ø More recently, prototype products have been developed based on gridded forecast data, such as this product which highlights the risk of heavy rainfall. Similar products also available for • hot & cold temperatures, • strong winds Prototype product courtesy Mio Matsueda

Steps to progress use of GIFS products in SWFDP Ø Progress so far Ø TC products based on CXML data; prototype products based on gridded TIGGE forecast data Ø Provided documentation of prototype products Ø GIFS-TIGGE WG co-chair attended recent SWFDP SG meeting Ø Seek feedback from RSMCs coordinating SWFDP regional subprojects Ø Current Ø Develop real-time products for SWFDP based on preferred prototypes, e. g. , Multi-model versions of TC products; near real-time versions of highest priority rainfall products. Ø Supply products to SWFDP regional websites Ø Future Ø Provide training on GIFS products via SWFDP

TIGGE-LAM: TIGGE for limited-area models To complement the global TIGGE project, a TIGGE-LAM panel was set up to coordinate the development of Limited Area Model Ensemble Prediction Systems. The panel facilitates the interoperability of the different regional ensemble systems and coordinates the archiving of limited-area ensemble forecasts. The TIGGE-LAM panel is chaired by Tiziana Paccagnella and works closely with the GIFS-TIGGE working group. Ø The three TIGGE archive centres have agreed to host a sub-set of high priority data, and archiving of LAM EPS data has now started. Ø With the increasing emphasis on convective-scale ensembles, TIGGE-LAM works closely with the WWRP Mesoscale Weather Forecasting Research group, as well as GIFS-TIGGE. Ø The TIGGE LAM Panel now structured in regional sub-groups to facilitate the focus on regional activities, with links to THORPEX regional committees. Ø More information is available from the TIGGE-LAM website http: //www. smr. arpa. emr. it/tiggelam/.

Ø GEOWOW (GEOSS interoperability for Weather, Ocean and Water) is a 3 -year EU-funded FP 7 project started September 2011. Ø The Weather component (WP 4) includes TIGGE archive enhancements: Ø TIGGE-LAM archive Ø Time-series data – for specified set of locations Ø Provision of data in netcdf format as alternative to GRIB 2 GIFS-TIGGE 31 August - 2 September 2011

Other developments Ø Clear need for coordination with S 2 S archive, as well as TIGGE-LAM – a trio of complementary datasets Ø Supplementary data Ø Observations – NCAR model validation portal Ø Possible future enhancements (to be discussed) Ø Additional fields (e. g. ice-related parameters for polar project) Ø Reforecast data Ø Resolution Ø Currently TIGGE archives data at native model resolution, but need to consider standard or maximum resolution in future.

Post-THORPEX landscape Still fluid, but a plausible scenario is: Ø WWRP “Predictability and Ensemble Forecasting” working group, focusing on science of dynamics & predictability and ensemble forecasting. Ø Small technical “liaison group” to manage TIGGE database (could include S 2 S & TIGGE-LAM too? ) Ø New “high-impact weather” project to continue some of THORPEX R&D, with more of a focus on short-range convective-scale resolution. To be discussed at Karlsruhe workshop in March.

WWRP-THORPEX GIFS-TIGGE working group PDP working group TIGGE-LAM panel TIGGE-LAM dataset TIGGE dataset Users Predictability, dynamics, probabilistic forecasting

WWRP HIW project team P&EF expert team WCRP S 2 S project team Dataset liaison group TIGGE-LAM dataset TIGGE dataset S 2 S dataset Users Sub-seasonal to seasonal and polar predictability, high-impact weather, probabilistic forecasting, RDPs, FDPs

Any Questions?