29 May 2019 The Added Value of UserDriven

29 May 2019 The Added Value of User-Driven Climate Predictions F. J. Doblas-Reyes, M. Donat, D. Verfaillie, B. Solaraju, R. Bilbao, D. Bojovic, M. Terrado, I. Christel, S. Wild

Near term: Sector readiness In all sectors there are potential applications of climate predictions, but in some sectors the decision making processes that would benefit from decadal predictions, understood as a mature climate prediction tool, are better defined. Agriculture Energy Water High management Insurance Transport Forestry Fire management Health Fisheries Tourism Low This is valid provided the added value of predictions-projections is illustrated to the users. From project deliverables of EUCP (D 6. 4), PRIMAVERA (D 11. 6) and EUPORIAS (D 12. 3). Additional sectoral comments in user engagement by S 2 S 4 E, APPLICATE, MED-GOLD, HIATUS and VISCA. 3

Near-term projections Seasonal-mean air temperature change for the RCP 4. 5 scenario over 2016 -2035 (wrt Stippling for significant changes, hatching C 1986 -2005). a for non-significant. n Ca the us nt er he sg us e ta er n s yth gr ge an t a ing ula dd be rit itio t ter y? na ? l ti me IPCC AR 5 WGI (2013) 4

Climate time scales Progression from initial-value problems with weather forecasting at Thmulti-decadal to century projections as a forced one end and e b problem at the other, with climate prediction (sub inf condition boundary o rm igand -seasonal, seasonal decadal) in the middle. Prediction involves c h initialization andatsystematic io al comparison with a simultaneous reference. len n be fo ge r s Subseasonal to seasonal t Climate-change cli the. Decadalisforecasts (18 forecasts (2 weeks-18 t projections Weather months-30 oyears) n months) m ate ext pro forecasts da 30 vid Time ta ye e c Initial-value driven so ars lim ur ce from ate s Boundary-condition driven the Adapted from Meehl et al. (2009) 5

Applications: Energy 7

Applications: Agriculture Multi-model correlation between the predicted ensemble mean and reference (from GHCN and GPCC) standardised precipitation evapotranspiration index of the previous six months (SPEI 6) for the boreal summer averaged over forecast years 2 to 5. INIT - No. INIT: Initialized decadal prediction No. INIT: Non initialized climate projection Sep Aug Jul INIT • Solaraju Murali (BSC) C 3 S Climate Projections Workshop: Near-term predictions and projections, 21 April 2015 8

Real-time decadal climate prediction The multi-model real-time decadal prediction exchange is a research exercise that guarantees equal ownership to the contributors. BSC is one of the four centres recognised as global producers of decadal climate predictions by WMO-CCl. 13

Lots yet to explore A non-exhaustive list of aspects to be explored in climate prediction: • Definition of benchmarks from the user perspective (not just climatology, persistence or projections). • Standards and quality assurance. • Entry-level documentation. • Integration of the observational uncertainty in the production chain. • Model weighting, model selection and prediction-projection merging. • Use of new paradigms like storylines and use of emergent constrains. 14