“In this research work we will quantify the value that the NMME Phase-2 system adds to the prediction skill of the subseasonal excessive heat outlook system with focus on human health (SEHOS-H) currently in development at NOAA’s Climate Prediction Center (CPC) and the Earth System Science Interdisciplinary Center (ESSIC) of the University of Maryland. The development of the SEHOS-H was motivated by the fact that adaptation to more frequent, more intense and longer lasting heat waves, as projected by the IPCC, necessitates excessive heat information at increasing lead time utilizing all sources of available predictability. Known sources of predictability at subseasonal lead times include the Madden-Julian Oscillation (MJO) and patterns of anomalous mid-latitude atmospheric planetary waves with a wavenumber of 5 that precede heat waves over the contiguous U.S. (CONUS) by 15–20 days. These sources of predictability have to be considered in conjunction with interfering, even slower modes of variability such as ENSO.
In ongoing work, the SEHOS-H uses ensemble forecasts of 2-meter apparent and dry-bulb temperature calculated by the NCEP Global Ensemble Prediction system (GEFS) for forecasts during Week-2 and the Climate Forecast System (CFS) for forecasts during Week 3-4. The SEHOS-H is based on effects of heat on human health and defines the intensity of heat events as the integral of the standard NOAA heat index over the number of consecutive days that the heat index exceeds a certain percentile threshold. Additional seasonal weighting of each day is performed in order to take into account acclimatization to heat. This weighting was optimized by using mortality data. Verification of this system is done by comparing the forecasts to observations considering: (1) the number of heat events successfully forecast by the system, (2) the number of false alerts, (3) the number of missed alerts and finally, (4) the forecast skill for the intensity of heat events.
Multi-model ensemble forecasting has been shown to be more skillful than each individual model forecast. This consideration was the motivation for the current proposal which seeks to compare the forecast skill of the baseline experimental SEHOS-H to the forecast skill of the SEHOS-H forced by every possible combination from the NMME Phase-2 models. We will first assess the forecast skill of apparent and dry-bulb temperature for all NMME models and their combinations and compare them to the baseline GEFS/CFS. We will then use these fields to forecast the duration and intensity of heat events thus considering the non-linear relations between atmospheric conditions and human health. Verification will be based on the Climate Test Bed (CTB) protocol and use the suggested categorical and probabilistic verification techniques.
This research will document the importance of the NMME system as part of a subseasonal excessive heat outlook system and the realism in the representation by the NMME models of subseasonal variability relevant to extreme heat patterns over the CONUS. As such this proposal is highly relevant to the targeted competition as well as highly visible due to recent initiatives announced at both the NOAA and White House level in recent months.”