Discipline: Ecology Environmental and Earth Sciences
Subcategory: Geosciences and Earth Sciences
Session: 3
Room: Exhibit Hall
Michael Ward - California State University, Northridge
Co-Author(s): Ryan Lynch, University of Michigan - Ann Arbor, MI. Steven Lazarus, Florida Institute of Technology, FL. Mike Splitt, Florida Institute of Technology, FL. Nezamoddin N. Kachouie, Florida Institute of Technology, FL.
Reduced ceiling and visibility (C&V) is frequently the cause of weather delays in the United States and has contributed to many of the most devastating aviation disasters [1]. Poor C&V conditions and their corresponding FAA regulations can lead to lengthy and costly delays. According to the FAA, aviation delays cost as much as $33 billion in 2019 [2]. Fuel burned during flight delays released an additional 7.1 million metric tons of carbon dioxide into the atmosphere in 2007.A number of studies have used a variety of forecast parameters such as temperature, precipitation, relative humidity, and ENSO indices to model and predict Flight Rules Categories (FRCs). The previous work has demonstrated potential use of short term (up to 30 hours) ensemble re-forecasts and Climate Prediction Center (CPC) seasonal outlooks to model FRCs. However, there are also benefits in terms of FRC forecasting at the intermediate time scale (on the order of a week). There are four FRCs including VFR, MVFR, IFR, and LIFR corresponding to C&V. The main objective of this project is to study the potential use of climate and other local spatiotemporal predictors to predict each flight category. Beta regression is used to model FRC rates in this study. The full 20-year CPC 6-10 day gridded categorical and probabilistic precipitation/temperature outlooks and corresponding METARs at 15 airports were employed to identify the optimal model at each airport based on AIC and pseudo-R2 values. Using C&V, 5-min METAR observations are converted to FRCs and then averaged over the same 5-day window as the CPC outlooks. The METAR locations were selected based on a range of climate classifications and fog occurrence frequencies. Selected highlights from a 30-year monthly FRC climatology are presented and then, as a prelude to modeling, compared to the average monthly FRC rates as a function of the CPC categorical temperature and precipitation forecasts including below normal, normal, or above normal. Identified optimal models were verified using one month of observed FRCs for 5 airports and preliminary results yielded small mean squared error (MSE). The results indicate that CPC probabilities and seasonal months can be useful in predicting FRCs. Although the proposed models could explain some variation in the observed FRCs, additional predictors can be included to improve the proposed models. The future work will be focused on inclusion of more predictors such as soil moisture. References:[1] Andrew J. Fultz & Walker S. Ashley (2016) Fatal weather-related general aviation accidents in the United States, Physical Geography, 37:5, 291-312, DOI: 10.1080/02723646.2016.1211854[2] Committee, U., 2008. Your Flight Has Been Delayed Again – Your Flight Has Been Delayed Again – United States Joint Economic Committee. [online] Jec.senate.gov. Available at:
Funder Acknowledgement(s): Funding was provided by an NSF REU Grant to Dr. Nezamoddin N. Kachouie, Florida Institute of Technology.
Faculty Advisor: Dr. Nezamoddin N. Kachouie and Dr. Steven Lazarus, nezamoddin@fit.edu
Role: This project was one of the projects in the REU Summer 2021 Program of Statistical Modeling with Applications to Geoscience (SMAG ) at Florida Institute of Technology under direct supervision of Dr. Nezamoddin N. Kachouie and Dr. Lazarus. As an undergraduate student participant, I was involved in all aspects of this project working in a team of two undergrad students.