Aviation Verify Program: Statistics
Aviation Verify generates a multitude of statistics, with output in different formats depending on what is requested. When the monthly report is run, a file called "Raw Stats", if desired, may be generated that contains all of the output statistics. Users may then use this file to create their own specialized output to meet local needs. In all statistical output files, decimal points are not used. For statistics that normally range between zero and one, i.e., POD, FAR, the number is multiplied by 100. Thus, a POD of 0.76 would be listed as 76. All other statistics are rounded to the nearest whole number. The following terms are used in the statistical output: Hrs Hours (Hrs) are the number of hours a phenomenon (e.g., light winds or wind gusts) was observed during TAF verification. Whenever statistics for all TAF initiation times (00, 06, 12, and 18 UTC) are parsed together, each hourly observation is covered by four TAFS; therefore, the numbers in the "hours" row or column are four times higher than actually observed. This definition does not apply to TEMPO and PROB groups (see sections 4. 4 and 4.5). % C The percentage (% correct) of total forecasts issued that were correct. It is similar to POD, but % correct is usually used to evaluate all categories of a given element (e.g., all eight ceiling categories), whereas POD is usually used for evaluating a single category or threshold (e.g., ceilings below 200 feet). m%C The percent correct for MOS. POD Probability of detection (POD) is equal to the number of times the TAF matched the corresponding observation divided by the total number of times the element was observed. It is similar to % correct, but POD is usually used for evaluating a single category or threshold (e.g., ceilings below 200 feet), whereas % correct is usually used for evaluating all categories of a given element. POD is usually expressed with a false alarm ratio because it is possible to achieve a very high POD by issuing a lot of "false alarms". mPOD The same as POD except it uses the MOS forecast. FAR False alarm ratio (FAR) is equal to the number of times element was forecast but not observed divided by the total number of times it was forecast. mFAR The same as FAR except it uses the MOS forecast. FA Hr False Alarm Hours (FA Hr) is the actual number of hours an element was forecast but not observed. This provides clearer information on events that rarely occur. CSI Critical Success Index (CSI) is equal to the number of times a forecast matched the observation divided by the sum of the total number of times the element was observed plus total number of times the element was forecast but did not occur. mCSI The Critical Success Index for MOS. %T>M TAF better than MOS (%T>M) is the percentage of time the TAF was more accurate than the MOS forecast. %T<M TAF worse than MOS (%T<M) is the percentage of time the MOS forecast was more accurate than the TAF. T busts The percentage of time the hours the TAF was two or more ceiling or visibility categories different from the observed category. Note: The cases of VFR ceiling busts (ceiling categories 6 versus 8) are not counted. M busts Same as T busts, except M busts are computed for MOS. T = M = Ob The percentage of time the MOS and TAF agreed and were correct. T = M <> Ob The percentage of time the MOS and TAF agreed but both were incorrect.
Overview | Downloads | Multiple Forecasts At One Time | Statistics | Output | Quick Set Up Printing problems | Transferring data | Aviation Home | SR Home