Vertical Thermospheric Density Profiles From EUV Solar Occultations Made by PROBA2 LYRA for Solar Cycle 24

Abstract

A new data set of summed neutral N2 and O number density profiles, spanning altitudes between 150 and 400 km, and observed during Northern Winter from 2010 to 2016 is presented. The neutral density profiles are derived from solar occultation measurements made by the 0.1–20 nm Zr channel on the Large Yield Radiometer (LYRA) instrument on board Project for Onboard Autonomy 2 (PROBA2). The climatology derived from the vertical profiles is consistent with that predicted by the NRLMSISE‐00 model, and the systematic error and random uncertainty of the measurements are less than 13% and 6%, respectively. The density profiles are used to characterize the response of thermospheric density to solar EUV irradiance variability. Peak correlation coefficients between neutral density and EUV irradiance occur near 300 km at the dusk terminator and 220 km at the dawn terminator. Density variability is higher at dawn than it is at dusk, and temperature variability increases with increasing altitude at both terminators.

Mars Thermospheric Variability Revealed by MAVEN EUVM Solar Occultations: Structure at Aphelion and Perihelion and Response to EUV Forcing

“…Nevertheless, the uncertainties for LYRA are comparable to those shown for EUVM in Figure . Further, Thiemann, Dominique, et al () found better than 5% agreement with the LYRA measurements and predictions made by the NRLMSISE‐00 empirical model for Earth's thermospheric density (Picone et al, ), which has been independently characterized and validated by available thermospheric density measurements at the Earth. The good agreement between LYRA solar occultation observations and NRLMSISE‐00 predictions also serve to validate the EUVM‐SO observations given the similarities in the instruments and near identical retrieval algorithms.…”

“…A brief description of the retrieval method is summarized here, but the reader is referred to those two papers for a detailed description of the retrieval method. In particular, the study of Thiemann, Dominique, et al () was motivated by an effort to validate methods used in this study by applying them at the Earth, where thermospheric densities are better understood. As such, the method of Thiemann, Dominique, et al () is nearly identical to the method used in this study.…”

“…The reference atmospheres used in the retrieval are isothermal with the form n ( z ) = n ( z 0 )exp(( z − z 0 )/ H ), where n , z , z 0 , and H are the number density, altitude, reference altitude, and scale height, respectively. Roble and Norton () and Thiemann, Dominique, et al () showed that isothermal reference atmospheres do not force the retrieved atmospheres to also be isothermal, and temperature gradients are accurately captured at altitudes above an abrupt change in scale height. However, error induced by the reference atmospheres increases markedly with decreasing altitude below an abrupt change in scale height.…”

“…Thiemann, Dominique, et al () performed a nearly identical MC uncertainty analysis for the LYRA solar occultation measurements at Earth. The two differences in their analysis are that they omitted the response function uncertainty because it was known to remain within 1% due to a more precise prelaunch calibration, and they had an additional systematic uncertainty associated with assuming equivalent cross sections for N 2 and O.…”

“…Additionally, if the observing latitude varies substantially during an occultation scan, as occurs when the spacecraft β angle approaches 90°, the retrieved densities no longer correspond with a vertical profile. This has no significant effect on the density retrieval (Thiemann, Dominique, et al, ), but because the temperature calculation assumes a vertical density profile, substantial error can be introduced to derived temperatures when the scan spans a large range of latitudes. We adopt the upper limit of a 4° change in latitude during an occultation scan for valid temperature calculations based on the fact that in the winter hemisphere, where latitudinal temperature gradients are steeper, temperature varies by less than 10 K over 4° latitude (e.g., González‐Galindo, Forget, López‐Valverde, Angelats i Coll, & Millour, ) or ~10% of the expected vertical temperature variation over the EUVM‐SO measurement range.…”