The photochemistry of the c1&Sgr;,C3Δ (now renamed the A'3Δu), and A3&Sgr; excited states of O2 is of relevance to the energetics of the lower thermosphere and upper mesosphere and to the atomic oxygen concentration in this region. A means of obtaining the atomic oxygen concentration profile by remote sensing is via the O(1S) emission at 5577 ¿ provided the production and loss mechanisms of the O(1S) are well understood. The O(1S) is now believed to be produced in a two-step process originating with one of the above three states of O2 (the Barth mechanism). Earlier, it was thought to be produced directly in three-body recombination of atomic oxygen (the Chapman mechanism). While the empirical evidence in recent years has tipped toward the first of these alternatives, the identity of the intermediate state is still in question. The O2 atmosphere bands arise from the b1&Sgr; state, which is populated via a two-step process similar to the Barth mechanism, again involving an intermediate state. In this paper we investigate these related issues using spectral measurements gathered on the nocturnal middle atmosphere with an array of imaging spectrometers flown on the Spacelab 1 shuttle mission. At this time the problem still contains too many uncertain parameters to be unambiguously resolved by the data. However, the data appear best supported by the O2(b1&Sgr;) arising from the O2(c1&Sgr;), although an additional source of O2(b1&Sgr;) is indicated.

The O(1S) is more closely represented by a process with the Chapmanlike characteristics of an [O>3 dependence, which seems to favor the O2(A'3Δ) as the intermediate state in this case if the previously accepted rate coefficients are used. (We do not have sufficient information in our data set to reject the (A3) state as the O(1S) precursor, although this has been rejected in other studies.) However, the state of knowledge of the rate coefficients and dominant quenching species for the potential intermediate states and their vibrationally excited forms is not yet stable. If the latest laboratory information on the O2(b1&Sgr;) state is correct in the relative roles of quenching by O and O2 of the intermediate state but incorrect in the rate of production of the intermediate state, another solution is possible. In this case the O2 atmospheric bands and O(1S) can arise from the same intermediate state but with O(1S) being generated only from vibrationally excited states with a yield of 0.01 and the O2 atmospheric generated also from the low vibrational states with a yield of at least .80.