Heat index products are issued from the medium-range forecast function of the HPC for days 3 through 7 during the period from May 1 through September 30. With the exception of day 3, these products are based on an ensemble of model and human forecasts. The day 3 forecast is based on the gridded Model Output Statistics (MOS) temperature and dew point temperature forecasts; while, forecasts for days 4 through 7 are based on the HPC medium-range forecast desk products. The graphic plots display the cities to which the heat index forecast is interpolated, as well as shaded contours indicating the probability of the daily maximum, daily minimum, and daily mean heat index reaching or exceeding a specific threshold. The three thresholds for the maximum are 95, 100, and 105 degrees F; for the minimum, they are 75, 80, and 85 degrees F; and for the mean, they are 85, 90 and 95 degrees F. Contours are drawn for 10%, 40% and 70% probabilities, with shading between the lines of all areas greater than 10%. Alphanumeric tables are also available displaying each forecast heat index parameter (maximum, minimum, and mean) value and the probabilities of that heat index parameter exceeding the 70, 75, 80, 85, 90, 95 and 100 degree thresholds, except 80, 85, 90, 95, 100, 105, 110 degrees for the maximum.

        Studies indicate the duration of high heat indices over an extended period of time presents a greater danger to the general public than a high index over a shorter period of time; therefore, the daily mean heat index value over a 24-hour period is presented along with the daily extrema of heat index values.

        The forecast heat index values are calculated using grids of 24-hour maximum and minimum temperature forecasts and grids of 12 UTC and 00 UTC dew point temperatures for each forecast day. These forecast grids for days 4 through 7 are derived from the HPC point forecasts in the manner described in the documentation of the HPC 5-km resolution grid data products. The day 3 forecast information is obtained from gridded MOS data. Relative humidity values are computed using the temperature and dew point temperature forecasts. These relative humidity values and forecast temperatures are substituted into an equation derived using multiple regression to calculate the maximum and minimum heat indices, which are averaged to get the daily mean heat index.

        The uncertainty information for the heat index probability forecasts is obtained from an ensemble composed of the NCEP high-resolution Global Forecast System (GFS) model and the members of the NCEP Global Ensemble Forecast System (GEFS). The mean and spread of the heat index is available from this ensemble at six-hour intervals through each forecast day, beginning at valid time 06 UTC, at each grid point.  From this data, means and spreads of the minimum, mean, and maximum heat index are obtained for each forecast day. The probability of the maximum, minimum, or mean heat index exceeding a threshold is computed on the assumption of normal distribution of probability centered on the HPC or MOS heat index value using a standard deviation equal to the spread of the ensemble about the HPC or MOS heat index value. These calculations are done separately at each point on a grid having 20-km resolution.

        A contouring program is used to make color filled displays of the heat index values and the probabilities. The tabulated forecast values of the probability of a heat index parameter exceeding a specific threshold are interpolated linearly from the gridded data and made available at the display locations.

For more information on the generation of these products please click here.