Statistical Characterization of Vertical Wind Drivers from GITM Simulations

Characterization of the thermospheric vertical wind is a vital challenge in atmospheric modeling, due to the importance of the vertical wind in driving composition change throughout the atmosphere. Several mechanisms for what drives vertical winds have been proposed, such asa redistribution of pressure levels in a hydrostatic atmosphere, as the result of convergence and divergence of horizontal winds, and as a result of non-hydrostatic processes emerging from conditions created by the pressure gradient differing significantly from gravity. Recent analyses of the vertical wind derived from linear acceleration measurements of the Gravity field and steady-state Ocean Circulation Explorer (GOCE) satellite have indicated a strong response of thevertical wind to geomagnetic activity at high latitudes, a dependence on magnetic local time that varies with altitude, and little to no response to solar activity. The capacity of the Global Ionosphere-Thermosphere Model (GITM) to capture this behavior of the vertical wind was explored. A statistical analysis of the standard deviation of the vertical wind (σ(Vz)) was performed at polar, auroral, and mid- latitudes. An increase in dispersion as a function of both Auroral Electrojet index and magnetic latitude in the GITM vertical winds was observed.Probability distributions were computed, and suggest that GITM performs most well in the mid-latitudes for AE below 250. We also constructed truth tables to aid in describing possible drivers of the vertical winds at different velocity levels, and show that GITM preforms fairly well, but isoften overly generous in how often it predicts wind speeds in excess of 50 and 75 m/s compared to the GOCE data. Both analysis methods indicate that GITM normally has lower vertical winds than GOCE, but has stronger vertical winds more often (i.e., a higher tail in the distribution function).