Visibility forecast verification at Jomo Kenyatta International Airport
View/ Open
Date
2013Author
Oundo, Kizito Amua
Type
ThesisLanguage
enMetadata
Show full item recordAbstract
The economic growth of any airline is largely determined by the optimum operationalization of
the flight schedules. Bad weather and in particular low visibility at the airport may lead to
interruption of the schedule. Low visibility may also lead to delays for several hours when taking
off and holding for some time or diversion. Diversions to nearby airport lead to large economic
losses to the airline. Visibility forecasts are very critical in flight planning hence considerable
interest in assessing its accuracy, skill and value. The visibility forecasts form part of the
Terminal Aerodrome Forecast (TAF) provided by aerodrome meteorological office.TAF
comprises of forecast for wind speed and direction, visibility, weather and ceiling conditions. In
this research the forecasts for visibility was verified against visibility observation for 0hr and 6hr
lead times. The lowest observed value was used to score against the lowest forecasted value.
Visibility was categorized into ranges of poor covering range between 0-1000m, fair between
2000-4000m and clear between 5000-9999. The entries were then made into a 3-category
contingency table for six hour lead time. Accuracy at 0hr and 6hr lead time was evaluated using
the scatter plot and the root mean square error for days with determinable visibility both in the
forecast and in the observations, visibility of 9999 were dropped since specific visibility value
could not be determined. Various skill score were then evaluated from the contingency table. The
accuracy and skill scores for visibility forecasts are high for 0hr lead time as compared to 6hr
lead time hence the visibility forecasts should be used immediately after production and limited
to not more than six hours since its accuracy and skill drops drastically. For flight planning the
new and amended forecasts should be treated with urgency since convey more accurate forecasts.
The verification results should be able to show the strengths and weaknesses attached on the
visibility forecasts thereby acting as a tool to foster further improvement in low visibility
forecasting on the part of the forecasters. The skill scores generated from the contingency table
will assist the management in advancing ways of improving low visibility forecasting by
improving the equipments for observing and forecasting low visibility and also initiating further
training in low visibility forecasting for forecasters.
Publisher
School of Physical Sciences