SEAS5 became operational in November 2017, following almost two years of effort dedicated to the upgrade. To date, the seasonal forecasting configuration at ECMWF has been upgraded only infrequently, so each new version represents a major change, and SEAS5 is no exception. As well as many scientific developments, SEAS5 highlights include major increases in resolution for both atmospheric and oceanic models, the introduction of an active sea-ice model, and a shift in strategy to align the seasonal forecast configuration more closely to that used for medium-range forecasts. ENSO SST forecasts, which were already good, show further substantial improvement. Many other aspects of forecast performance are also improved, although score differences are often within sampling error and a few specific areas of deterioration have also been identified.
This paper gives some background to the development of SEAS5 (Section 1) and a description of the SEAS5 configuration (Section 2). Section 3 gives an assessment of performance, including various studies to help understand the factors affecting forecast performance. Finally, Section 4 is forward looking, with an overview of scientific priorities and a roadmap for future developments, including SEAS6.
BT - ECMWF Technical Memoranda DA - 11/2018 DO - 10.21957/z3e92di7y LA - eng M1 - 835 N2 -SEAS5 became operational in November 2017, following almost two years of effort dedicated to the upgrade. To date, the seasonal forecasting configuration at ECMWF has been upgraded only infrequently, so each new version represents a major change, and SEAS5 is no exception. As well as many scientific developments, SEAS5 highlights include major increases in resolution for both atmospheric and oceanic models, the introduction of an active sea-ice model, and a shift in strategy to align the seasonal forecast configuration more closely to that used for medium-range forecasts. ENSO SST forecasts, which were already good, show further substantial improvement. Many other aspects of forecast performance are also improved, although score differences are often within sampling error and a few specific areas of deterioration have also been identified.
This paper gives some background to the development of SEAS5 (Section 1) and a description of the SEAS5 configuration (Section 2). Section 3 gives an assessment of performance, including various studies to help understand the factors affecting forecast performance. Finally, Section 4 is forward looking, with an overview of scientific priorities and a roadmap for future developments, including SEAS6.
PB - ECMWF PY - 2018 T2 - ECMWF Technical Memoranda TI - SEAS5 and the future evolution of the long-range forecast system UR - https://www.ecmwf.int/node/18750 ER -