Objective Visibility Forecasting Techniques Based on Surface and Tower Observations

摘要

Three objective techniques are presented for forecasting the morning minimum visibility at Washington National Airport in the fall season. Excluded from forecasts are dates with precipitation or a frontal passage before 1200Z. The basic technique makes use of the 0400Z wind speed and dew-point depression at Washington; the wind speed, wind direction, and wind gustiness recorded on a 300-ft tower; and a simple yes-or-no forecast for any clouds below 7500 ft before 1200Z. Gustiness is measured by the wind direction variability on the tower for one hour, 0300-0400Z. A decision tree approach combined with graphical methods give the probability of visibility occurring in five classes. Three years of data provide the developmental sample. Verification statistics from an independent two-year sample show 66 percent of the forecasts in the correct class and a skill score of .40, slightly exceeding the skill of the official forecast. An alternate second technique uses only surface data from which tower data are objectively estimated; an alternate third technique employs only wind speed and dew-point depression to forecast visibility from isopleths. The techniques development procedures are described and illustrated. (Author)