Level: Intermediate
JMP time series analysis, ARIMA models and State-Space Smoothing Models (SSM) were used to investigate Antarctic glacier mass loss for the projection period 2021-2041. Ice sheet mass data from the GRACE-FO satellite was also considered in a geoscience and COVID-19 context. Non-seasonal/seasonal ARIMA models show 12-month seasonality and long-term year-over-year trend patterns (average 2% increase in melt rate). SSMs improved forecast accuracy by reducing random noise and extended the forecast horizon to at least 5 years in the future.
*Please note that due to the presenter's circumstances, only the JMP data used in the presentation and analysis will be made public.
