Electron spectra were obtained during two rocket flights into pulsating aurora from Southend, Saskatchewan. The first rocket launched at 1143:24 UT, on February 15, 1980 flew into an aurora of background intensity 275 R of N2+ 4278 Å and showing regular pulsations with about a 17 s period. Electron spectra of Maxwellian energy distributions were observed with an average E0 = 1.5 keV, rising to 1.8 keV during the pulsations. There was one-to-one correspondence between the electron energy modulation and the observed optical pulsations. The second rocket, launched at 1009:10 UT on February 23, flew into a diffuse auroral surface of intensity 800 R of N2+ 4278 Å and with somewhat irregular pulsations. The electron spectra were again of Maxwellian energy distribution with an average E0 = 1.8 keV increasing to 2.1 keV during the pulsations. The results from these flights suggest that pulsating auroras occurring in the morning sector may be quite commonly excited by low energy electrons. The optical pulsations are due to periodic increases in the energy of the electrons with the source of modulation in the vicinity of the geomagnetic equatorial plane.

Rocket‐borne measurements are made of the particle pulsation characteristics in a pulsating aurora. A train of eleven 17‐s pulsations was observed in electron intensities at energies as low as 2 keV and as high as 28 keV. During the pulsations a factor of 2 increase in the electron intensities at the few‐keV range, and a factor of 3–5 increase in the intensities in the 20‐keV range, were measured. Time delays in pulsations of low‐energy electrons are observed; they indicate the source of particle injection to be near the equatorial plane. The electron energy spectra in pulsation maxima are characterized by higher Maxwellian temperatures than their pulsation minima counterparts. The few‐keV electrons have more trapped distributions at pulsation minima than at pulsation maxima. The energetic electrons are isotropic at all times. The altitude profiles of the electron fluxes at various energies are examined, in an attempt to identify the possible existence of thin optical pulsating auroral structure; they are found to be similar to those associated with discrete auroras. Current theoretical pulsation models are discussed in light of the observations.

Ground‐based auroral observations with a meridian‐scanning photometer and all‐sky camera were made at Rabbit Lake, Saskatchewan, Canada, simultaneously with nearby HILAT satellite passes during six winter periods near new moon from 1984 to 1986. Eight passes were selected in which HILAT overflew auroras visible from Rabbit Lake. Auroral intensity measurements made at Rabbit Lake were corrected for proton aurora and nightglow and compared with HILAT electron flux measurements at 800 km. The excitation efficiency for the 4278‐Å auroral emission by electron precipitation was found to be (0.29±0.08) kR/(erg/cm² s). Excitation efficiencies for 5577‐Å emission of (1.73±0.51) kR/(erg/cm² s) and (1.23±0.44) kR/(erg/cm² s) were measured on different days in auroras excited by 1.8‐keV and 3.1‐keV Maxwellian precipitation during moderate geomagnetic disturbance. After small night‐to‐night variations in atmospheric composition indicated by the MSIS‐86 model atmosphere were accounted for, the magnetic zenith column intensity ratio 6300/4278 was found to depend on the electron characteristic energy Eo approximately as 6300/4278 ∼ 3.3 Eo−2.1 for characteristic energies of a few keV. The 5577/4278 ratio was found to decrease by about 40% from 1 to 5 keV characteristic energy. These Results pertain mostly to weak auroras (I(5577 Å) = 2–6 kR). Most of these new measurements agree fairly well with previous results of other workers, but the two 5577‐Å excitation efficiencies and the decreasing 5577/4278 intensity ratio show that the green line excitation efficiency varies appreciably even for small changes in the precipitating electron spectrum and atmospheric composition. Published calculations of the auroral 6300/4278 and 5577/4278 ratios versus characteristic energy are compared with our measurements. The magnitude and variation of the 5577/4278 ratios are consistent with roughly 60% of green line emission resulting from O(¹S) excitation by N2(A) energy transfer.

An analysis of some 523 F layer patches observed over Eureka, Northwest Territories, Canada (89° corrected geomagnetic latitude) during the four winters from 1990–1991 to 1993–1994 has given some definitive results on their occurrence patterns and characteristics. They are observed on average about 25% of the time and are seen at all hours of the day but with more present in the local evening hours than the morning hours. For the 182 patches for which interplanetary magnetic field data were available, 143 of the patches occurred with Bz negative and 39 with Bz positive. In most of the latter cases, Bz was either near zero or had changed to positive only 30–120 min prior to the patch observation. The average optical emission intensity of the patches over the 4‐year period was 150 R of 630 nm and 55 R of 558 nm. The excitation is by dissociation recombination of the oxygen ion with a branching ratio O[1S]/O[1D] of 0.37. They were on average less bright during the midnight hours (by about 50 R) than through the midday, a fact perhaps due to the offset of the Eureka station from the magnetic pole. The average patch intensity has decreased from 190 R in 1990–1991 to 120 R in 1993–1994 as solar activity has declined. While patch sizes in the dawn‐dusk direction varied up to >2000 km, the average cross‐sectional dimension was 500–600 km.

No abstract