AIRBORNE MEASUREMENTS OF THE PHOTOLYSIS FREQUENCY OF NO2

A set of photoelectric detectors for airborne measurements of the phot olysis frequency of NO2, i.e., J(NO2), was developed and integrated ab oard the research aircraft Hercules C-130 operated by the U.K. Meteoro logical Office. The instrument consists of two separate sensors, each of which provides an isotropic response over a solid angle of 2 pi ste radian (sr). The sensors are mounted on top and below the aircraft, re spectively, to obtain a field of view of 4 pi sr, and permit the discr imination of the upwelling and downwelling components of the actinic f lux. From experimental tests and model calculations it is demonstrated that small differences between the spectral sensitivity of the sensor s and the spectral response of J(NO2) can lead to significant errors i n the determination of J(NO2), especially under cloudy conditions. We present correction factors for clear sky conditions and suggest the us e of a new filter combination in the sensors which requires only small corrections and provides acceptable accuracy, even under cloudy condi tions. A climatology of J(NO2) values is presented from a series of fl ights made in 1993 at latitudes of 36 degrees-59 degrees N. For clear sky conditions and solar zenith angles of 33 degrees-35 degrees, J(NO2 ) was 8.3 x 10(-3) s(-1) were observed for very short periods in the u ppermost layers of clouds. Enhancement of the actinic flux due to ligh t scattered from clouds was also observed at altitudes below 0.5 km. C omparison of the clear sky data with predictions from different radiat ive transfer models reveals the best agreement for models of higher an gular resolution. The Delta Eddington method underpredicts the measure ments significantly, whereas the J(NO2) values by the discrete ordinat e method and multidirectional model are only about 5% smaller than our measurements, a difference that is within the experimental uncertaint ies.