Statistical analysis of wave heights using Kalman Filtering methods
Cecilia Linroth · KTH Publication Database DiVA (KTH Royal Institute of Technology) · 2014
We present some of the basic theory of Kalman ltering on time series with missing data and then show how to implement this numerically for ARIMA-processes. Using these numerical tools we analyze the ocean wave data from US NODC Buoy 46005 obtained between January 1983 an December 2003 which is missing about 17 % of the measurements. We look some of the strengths and weaknesses of the Kalman lter that become apparent and discuss the problems that occurred in the implementation.