On multivariate imputation and forecasting of decadal wind speed missing data

Area of Science:

• Meteorology and Climatology
• Statistical Modeling

Background:

• Missing historical wind speed data is a significant challenge in meteorological studies.
• Accurate wind speed data is crucial for renewable energy assessments and climate research.

Purpose of the Study:

• To apply multiple imputations by chained equations (MICE) for handling missing wind speed data.
• To develop and validate a time series forecasting model for wind speed prediction.
• To estimate maximum decadal wind speed for Entebbe International Airport.

Main Methods:

• Utilized multiple imputations by chained equations (MICE) with a fully conditional specification.
• Employed time series forecasting techniques to analyze wind speed patterns.
• Statistical significance testing and error bound estimation were performed.

Main Results:

• MICE provided reliable imputations for missing wind speed data.
• The forecasting model indicated a smoothing parameter (alpha = 0.014) close to zero, favoring recent observations.
• Estimated maximum decadal wind speed at Entebbe International Airport: 17.6 m/s (95% CI: 6.8-28.4 m/s).

Conclusions:

• MICE is a robust method for addressing missing wind speed data.
• Time series forecasting effectively captures wind speed dynamics.
• The high bound on error estimation highlights the inherent variability of wind speed at the study location.