NCEP global forecasts, Reynolds weekly sea surface temperature, and GOES-East, GOES-West, MTSAT-1R and Meteosat-7 channel-3 (water vapor) imagery are used as input to an algorithm to estimate the probability of tropical cyclone formation within 500km of each grid point within the next 48 hours 45 S to 45 N and 0 to 360 E. The product domain is divided into 7 main basins (below) based on satellite coverage and warning agency boundaries.
Figure: TCFP basins
**Note: all parameters except DNST computed over a 500km radius area centered on each grid point
Climatological input parameters are computed over the following time periods:
The formation probability is calculated by a three-step algorithm. First, all cases where input parameter values exceed predetermined thresholds are screened out. Then, a statistically-derived linear function of the input parameters is used to compute the linear discriminant function value at each grid point that passed screening. Lastly, a statistically-derived relationship between the linear discriminant function value and the corresponding past TC formation occurence frequecy at used to determine the probability of TC formation.
TC formation probability (TCFP) is computed for each 24-hr forecast interval (0-24hr and 24-48hr) based on the 24-hr averages of the input parameters. The cumulative probability of TC formation (CUMP) is computed using the formula:
CUMP(0-48h) = TCFP(0-24h) + TCFP(24-48h) – [TCFP(0-24h)*TCFP(24-48h)]
To provide some time continuity of the product, the formation probability and the primary parameters used in the screening and discrimination analysis are summed (probabilities) or averaged (all other parameters) over 6, 3, 6, and 3 sub-basins for the Atlantic, Eastern Pacific, Western Pacific, and Indian Ocean basins, respectively. Time series products comparing the current and climatological parameters over the sub-basins are provided on the web site.
|Figure: TCFP sub-basins|
Schumacher, A. B., M. DeMaria, J. A. Knaff, 2009: Objective Estimation of 24-h Probability of Tropical Cyclone Formation. Wea. Forecasting, 24, 456-471.