Time-series analysis and forecasting of daily averaged surface pressure on Mars (VL1 and VL2 sites)
Stationarity of daily averaged surface pressure time series on Mars can be obtained (by differencing) for short time intervals only. Large pressure variance intervals are usually associated with the dust storm season. Medium and small pressure variance intervals can be found usually during clear sky periods. General auto-regressive integrated moving average (ARIMA) models were developed for daily average pressure forecasting at Viking Lander sites. A procedure to find the “appropriate” model for a given time series was developed. It is based on various degrees of differencing the original time series and a t-statistics-assisted estimation of the significance of the fitted coefficients. A method employing the cumulative spectrum of the residuals was used to check the models. ARIMA (3,2,3) seems to be the most “appropriate” model to forecasting the daily average surface pressure on Mars.
KeywordsDust Storm Lander Site Original Time Series Fitted Coefficient Storm Season
Unable to display preview. Download preview PDF.