|Download Postscript File||Download Compressed Postscript File|
Jhoon Kim and Gwang Rae Cho
Space Div., Korea Aerospace Research Institute
Taejon, 305-600, Korea
|Editor's note: This article is being presented at the Fall 1998 Meeting of the Korean Meteorological Society.|
To date, extensive efforts have been put to measure ozone concentration
profiles using ground-based, balloons, rockets and satellites. Sounding
rockets have played a significant role in measuring ozone concentration
profiles of the intermediate altitudes where the balloons and satellites
cannot reach for the precise in situ measurements. Remote sensing from the
satellite has been most popular for its global and continuous coverage.
Korea Aerospace Research Institute (KARI) has flown four sounding rockets
since 1993. KSR-I series which was first launched in June 4, 1993 obtained
two sounding data on stratospheric ozone concentration profiles. KSR-II, a
two-stage sounding rocket of KARI was launched successfully at the west
coast of the Korean Peninsula at 1000 LST, June 11th, 1998. Its scientific
mission were to measure ozone density profile, ionospheric electron density
and temperature profile, and celestial X-ray events. For this mission,
8-channel UV and visible radiometers, Langmuir probe, electron temperature
probe, and proportional counter for X-ray measurements were onboard the
rocket. The apogee of the rocket was 137 km and the total flight time was
365 seconds. The rocket measured a stratospheric and mesospheric ozone
density profile during its ascending phase using the 8-channel radiometer
and transmitted the data to ground station in real time (Fig. 1).
Ozone absorbs the solar UV radiation at Huggins and Hartley bands. This property is the important basis for the KSR ozone detector. The current KSR ozone detecting system is a modified version of KSR-I's (Kim et al. 1997) which measures the attenuation of sunlight in ultraviolet absorption bands of ozone as a function of height with eight radiometers (see Fig. 2). These radiometers were fixed on the rocket skin looking at an angle of 40 deg from the rocket body axis, to capture the Sun during the rocket flight within the FOV (field of view) of the detector considering the rocket trajectory, the predicted rocket attitude, and the position of the sun at the time and location of the launch. The ozone sensor consists of phototubes and interference filters to detect direct solar radiation at threeUV and one visible wavelength (255, 290, 310, and 450 nm) during the ascending period of the rocket. This simple design provides FOV of about 50. The incoming diffusive component of sunlight was assumed to be negligible compared to the total intensity measured by the instrument. One visible radiometer is for reference.
These radiometers use spin of the rocket to sweep the FOV past the solar disc. This captured solar radiation induces current from the phototube, which is then converted to analog voltage (0 - 5 V) in the circuit box. Analog output voltages and six-step gain levels for each radiometer channel were sampled at every 4 ms and telemetered in S band to the KARI's ground station. Ozone data are transmitted in a 16-channel, 10-bit pulse code modulation (PCM) system.
As a redundant way to see the effect of solar aspect angle change, the angle
was reduced from the onboard inertial navigation system (INS). Fig. 3 shows
the comparison between the calculated and measured values of the aspect
angle during the flight. In general, the difference between the calculated
and measured angles were within 2 deg up to 20 km due to active attitude
control system by canard fins. These differences become larger as altitude
increases due to nutational motion of the rocket, but still within the FOV
of the detector up to 70 km.
3. Data Reduction
Figure 4 shows analog signals (in volts) as a function of flight time (in seconds) of the selected radiometer channels. The upper envelope of signals corresponds to ascending motion of the rocket, as shown in Fig. 1 (KSR-II reached its apogee at t = 186 s). Thus these upper envelopes are expected to increase as time(thus altitude) increases until the rocket reached the top of the ozone layer, above which the solar UV radiation at these wavelengths are no longer absorbed by ozone. Thus the oscillation of the upper envelope corresponds to the change of the rocket attitude, and the individual peak signals at 2.5-sec intervals are the result of changing solar radiation intensities during spinning motion of the rocket. Peak signals at 2.5-sec intervals were used to retrieve O3 profiles with an altitude resolution of approximately 2.5 km.
Assuming horizontally uniform atmosphere, which is reasonable for the solar
zenith angle less than 60 (Watanabe 1986), the relationship between the
measured signal and ozone slant column density can be written as
4. Results and Discussion
After the attitude correction of signals as in Fig. 4, the measured solar radiation intensities at selected wavelengths are obtained as a function of altitudes (see Fig. 6). The increasing UV intensities at altitudes greater than 73 km is due to the very weak signals resulted from the attitude change of the rocket. Thus the ozone densities up to 70 km were obtained as shown in Fig. 7. Comparisons with Dobson spectrophotometer, ozonesonde, and HALOE onboard the UARS are shown together and are in reasonable agreements. The bulge of rocket measurement near 50 km is considered to be the effect of upward motion which can bring the ozone-rich atmosphere from the lower altitudes near the peak. Detail analysis using the comprehensive model needs to be carried out for this.
The ozone density profiles of different channels suggest that systematic error is very small in the measurements. The sources of errors in this rocket measurement include the error in characterizing filter response function, error in cross section data, error in attitude correction, telemetry and instrument noise, and error in estimating rocket altitude by radar. The largest uncertainties are due to the characterization of filter response function where the out-of-band filter response may cause nonnegligible errors. The errors in ozone density due to uncertainties in spectroscopic data are estimated to be less than 3and differences in the measurement of absorption cross section (e.g., Molina and Molina 1986). The error in attitude correction was estimated to be about 2% for our case.
Random errors due to telemetry and instrument noise are estimated to be less
than 2% for the current measurement. The rockets were tracked by radar at
the ground. The error in rocket altitude is about 15 m based on the
manufacturers specification. This error in determining rocket altitude can
lead to less than 0.5% error in the O3 density. Total random errors (1),
estimated considering all sources discussed above, are 15% for the altitudes
below 20 km, 7% for the altitude range above 20 km.