The launch of the Defense Meteorological Satellite Program (DMSP) satellite F16 in 2003 provided the first opportunity to analyze extensive sets of high‐quality coincident auroral particle and FUV data obtained by the onboard sensors Special Sensor Ultraviolet Spectrographic Imager (SSUSI) and Special Sensor Auroral Particle Sensor (SSJ/5). Features of interest are Ly α (121.6 nm), Lyman‐Birge‐Hopfield short (LBHS, the SSUSI 140–150 nm channel), and Lyman‐Birge‐Hopfield long (LBHL, 165–180 nm). We report on comparisons of column emission rates (CERs) by deriving simulated SSUSI values using SSJ/5 electron and ion (treated as proton) spectra. Field‐line tracing is performed to determine the locations of coincidences. CERs are obtained by integrating the products of particle spectra and monoenergetic emission yields. A technique is reported for deriving these yields from nonmonoenergetic CERs obtained by our particle transport model. SSJ/5 ion spectra are extrapolated above 30 keV using a statistical representation based on Polar Orbiting Environmental Satellites particle data. Key quantities of interest are ratios of SSUSI to SSJ/5‐based CERs (S‐S ratios) and corresponding ratios of proton‐produced to total emission (unity for Ly α and from 0 to 1 for LBHS and LBHL). SSJ/5‐based CERs are used to derive the latter ratios. Median ratio values are determined in order to reduce the error budget to primarily calibration and model errors. The median LBH S‐S ratios increase by a factor of ∼2.5 from electron to proton aurora and support significantly higher proton LBH emission efficiencies (3 times the electron efficiencies) assuming reported calibration uncertainties. This calls for significant increases in proton and/or H‐atom LBH cross sections. In turn, FUV auroral remote‐sensing algorithms must explicitly address both electron and proton aurora.

A large number (~1000) of coincident auroral far ultraviolet (FUV) and ground-based ionosonde observations are compared. This is the largest study to date of coincident satellite-based FUV and ground-based observations of the auroral E region. FUV radiance values from the NASA Thermosphere, Ionosphere, Mesosphere Energetics and Dynamics (TIMED) Global Ultraviolet Imager (GUVI) and the Defense Meteorological Satellite Program (DMSP) F16 and F18 Special Sensor Ultraviolet Spectrographic Imager (SSUSI) are included in the study. A method is described for deriving auroral ionospheric E region maximum electron density (NmE) and height of maximum electron density (hmE) from N Lyman-Birge-Hopfield (LBH) radiances given in two channels using lookup tables generated with the Boltzmann 3-Constituent (B3C) auroral particle transport and optical emission model. Our rules for scaling (i.e., extracting ionospheric parameters from) ionograms to obtain auroral NmE and hmE are also described. Statistical and visual comparison methods establish statistical consistency and agreement between the two methods for observing auroral NmE, but not auroral hmE. It is expected that auroral non-uniformity will cause the two NmE methods to give inconsistent results, but we have not attempted to quantify this effect in terms of more basic principles, and our results show that the two types of NmE observations are well correlated and statistically symmetrical, meaning that there is no overall bias and no scale-dependent bias.

[1] A significantly higher N 2 Lyman-Birge-Hopfield (LBH) emission efficiency for auroral proton precipitation compared to model calculations was reported by Knight et al. (2008) based on a statistical study utilizing coincident far ultraviolet and particle data from the sensors Special Sensor Ultraviolet Spectrographic Imager (SSUSI) and Special Sensor J/5 (SSJ/5) on board the DMSP satellite F16. Here, the quantity of interest from that study is the median ratio of LBH column emission rates (CERs) from SSUSI and derived from SSJ/5 spectra using monoenergetic emission yields. The median ratio was found to be 2.83 for proton aurora, suggesting the need for significant increases in currently used LBH proton/H-atom impact cross sections. A key step in their analysis was extrapolation of SSJ/5 spectra above 30 keV. Limited testing of this algorithm using NOAA Polar Orbiting Environmental Satellites Total Energy Detector and Medium Energy Proton and Electron Detector (TM) data found no significant bias. This work reports on a more detailed investigation of the algorithm's performance, also using TM data, and has uncovered a bias that reduces the median column emission rates (CER) ratio to 1.75. Within expected uncertainties, including calibration, this still calls for cross section increases but to a lesser extent. The discovered bias becomes apparent with CER thresholding that was overlooked during testing by Knight et al. (2008). Thresholding at 400 Rayleighs (R) is necessary since Knight et al. excluded CERs < 400 R in deriving their median ratio. We show that the algorithm's performance degrades with increasing energy flux of the precipitation. A method is reported for eliminating most of the bias which utilizes auroral Ly a, whose emission strength is closely coupled to spectral hardness.

[1] Model-derived electron and proton auroral FUV emission efficiencies of relevance to auroral FUV remote sensing methods are evaluated using coincident observations by Special Sensor Ultraviolet Spectrographic Imager (SSUSI) and Special Sensor J/5 (SSJ/5), both on board the Defense Meteorological Satellite Program (DMSP) satellite F16. This follows earlier work by Knight et al. (2008), which reported higher than expected proton Lyman-Birge-Hopfield (LBH) emission efficiencies based on F16 SSUSI and SSJ/5 comparisons, and Correira et al. (2011), which suggested a downward revision in the proton LBH efficiencies from Knight et al. (2008). These proton efficiency results rely on proton extrapolation methods to supply the unmeasured proton flux above 30 keV (the upper limit of SSJ/5). Correira et al. (2011) determined that there was a bias in the proton extrapolation method used by Knight et al. (2008) that was caused by column emission rate (CER) thresholding in the coincident SSUSI and SSJ/5 sets. In the latest work, a more robust proton flux extrapolation method is introduced which does not have the problem of CER threshold dependence. The new extrapolation method uses coincident SSUSI Lyman alpha observations to constrain the extrapolated proton flux above 30 keV in such a way that unknown Lyman alpha model yield errors and SSUSI and SSJ/5 calibration errors drop out without biasing the extrapolation. With the latest extrapolation method, SSUSI-SSJ/5 comparisons indicate that proton aurora is typically a factor of $4.5 more efficient per unit of energy flux in producing LBH than electron aurora.Citation: Knight, H. K., D. J. Strickland, J. Correira, J. H. Hecht, and P. R. Straus (2012), An empirical determination of proton auroral far ultraviolet emission efficiencies using a new nonclimatological proton flux extrapolation method,

[1] Near peak activity of two X-class solar flares, on 28 October and 4 November 2003, the Thermosphere Ionosphere Mesosphere Energetics and Dynamics (TIMED)/Solar EUV Experiment (SEE) instrument recorded order of magnitude increases in solar EUV irradiance, the TIMED/Global Ultraviolet Imager (GUVI) observed simultaneous increases in upper atmosphere far ultraviolet (FUV) dayglow, and the European Incoherent Scatter Scientific Association (EISCAT) radar and the Ionospheric Occultation Experiment onboard the PICOSat spacecraft recorded corresponding changes in E-region electron densities. Calculations of the FUV dayglow and electron density profiles using Version 8 SEE flare spectra overestimate the actual observed increases by more than a factor of 2.0. This prompted the development of an alternative approach that uses the FUV dayglow and associated E-layer electron density profiles to derive and validate, respectively, the increases in the solar EUV irradiance spectrum. The solar EUV spectrum required to produce the FUV dayglow is specified between 45 and 27 nm by SEE's EGS measurements, between 27 and 5 nm by GUVI dayglow measurements, and between 5 and 1 nm using a combination of the GOES X-ray data and the NRLEUV model. The energy fluxes in the 5-to 27-nm bands (at 5-10, 10-15, 15-20, and 20-27 nm) are randomly varied in search of combinations such that the full spectrum (l < 45 nm) replicates the GUVI dayglow observations. In contrast to the Version 8 SEE XPS observations, solar EUV spectra derived using the multiband yield approach produce electron densities that are consistent with those observed independently. The new multiband yield algorithm thus provides a unique tool for independent validation of solar EUV spectral irradiance measurements using FUV dayglow observations.