Abstract
A high-resolution model of exospheric temperatures has been developed,
with the objective of predicting the global values of exospheric
temperatures with greater accuracy. From these temperatures, the neutral
densities in the thermosphere can be calculated. This model is derived
from measurements of the neutral densities on the CHAMP, GRACE, and
Swarm satellites. These data were sorted into 1620, triangular cells on
a spherical, polyhedral grid, using coordinates of geographic latitude
and local solar time (longitude). A least-error fit of the data is used
to obtain a separate set of regression coefficients for each grid cell.
Several versions of model functions have been tested, using parameters
such as the day-of-year, Universal Time, solar indices, and emissions
from nitric oxide in the thermosphere, as measured with the SABER
instrument on the TIMED satellite. Accuracy is improved with the
addition of parameters that use the total Poynting flux flowing into the
Northern and Southern hemispheres. This energy flux is obtained from the
solar wind velocity and interplanetary magnetic, using an empirical
model. Given a specific date, time, and other inputs, a global map of
the exospheric temperature is obtained. These maps show significant
variability in the polar regions, that are strongly modulated by the
time-of-day, due to the rotation of the magnetic poles around the
geographic pole. Values at specific locations are obtained using a
triangular interpolation of these results. Comparisons of the exospheric
temperatures from the model with neutral density measurements are shown
to produce very good results.