Towards a Multi-Model Subseasonal Excessive Heat Outlook System

Abstract:

We developed an experimental realtime subseasonal excessive heat outlook and monitoring system (SEHOMS) based on the detection of heat events in dynamical forecasts and reanalyses. Our definition of a heat event takes into account both the challenges of subseasonal forecasting and the effects of heat stress on human physiology e.g., the dependence of heat impacts on duration, geographical location and timing of the heat event. The prototype outlook system focuses on forecast lead time week-2 and uses the Global Ensemble Forecast System (GEFS) reforecast conducted at ESRL and the NCEP-GEFS operational realtime ensemble forecasts. The prototype monitoring system, on which we base forecast verification, provides a dual output. The first product uses the NCAR/NCEP reanalysis; the second monitoring product is based on the day-1 forecast from the GEFS reforecast and from the operational GEFS realtime forecast. In this presentation we first show results from the prototype forecasting and monitoring system. We then compare these results with forecasts from the SEHOMS in which we gradually add reforecasts obtained from the S2S database (NCEP – Climate forecast System and ECMWF models). Finally we discuss the possibility of expanding the SEHOMS to week-3 and week-4 based on results from the CFS, ECMWF model, and the North American Multi-Model Ensemble system (NMME).