I. S. Boyd

Trends and variability in vertical ozone and temperature profiles measured by ozonesondes at Lauder, New Zealand: 1986–1996

A first analysis of trends in vertical ozone and temperature profiles from ozonesonde flights made at Lauder (45.045°S, 169.684°E) between August 1986 and July 1996, is presented. To calculate the trends and determine the magnitude of the forcing mechanisms underlying the variability in ozone and temperature, a linear least squares regression model was applied to ozone mixing ratios, ozone number densities, and temperatures, interpolated onto 100 pressure levels from the surface (969 hPa/370 m) to 12.1 hPa (∼30.1 km), ∼300 geopotential meters apart. Ozone trends indicate wintertime upper tropospheric decreases of more than -30±24% per decade (2σ), post vortex breakup trends in a narrow altitude region above the tropopause of -20±20% per decade (2σ), and positive trends of up to 30±14% per decade (2σ) in the lower stratosphere during late winter, spring, and early summer. The predominant temperature trend is +1.5% per decade and greater above the ∼50 hPa level during winter. Derived trends were sensitive to inclusion of tropopause height forcing which was found to influence ozone and temperature at a high level of statistical significance. Ozone at Lauder shows significant QBO dependence throughout the lower stratosphere during winter, spring, and early summer, but little or no dependence on the solar cycle. Temperatures, however, show little dependence on QBO but were influenced by the solar cycle. The Mt. Pinatubo eruption had little influence on Lauder ozone but significantly cooled the troposphere. The ENSO cycle in ozone and temperature was weak except at the uppermost analysis levels.

Stratospheric NO2 variations from a long time series at Lauder, New Zealand

Eighteen years of NO2 measurements using zenith-scattered sunlight are analyzed for seasonal, cyclic, and episodic variability and secular trends. The analysis shows a marked increase in stratospheric NO2 over the period, corresponding to a trend of 5% per decade, and the influences of both the El Chichon and Pinatubo eruptions are clearly evident. Smaller effects of the El Nino-Southern Oscillation and quasi-biennial oscillation are apparent, but correlation with the solar cycle is poor after correction for autocorrelation. All of these effects are similar for sunrise and sunset NO2.

An assessment of ECC ozonesondes operated using 1% and 0.5% KI cathode solutions at Lauder, New Zealand

The effects of reducing the concentration of the potassium iodide (KI) cathode electrolyte from 1% to 0.5% in electrochemical concentration cell (ECC) ozonesondes flown at Lauder, New Zealand (45° S, 170° E) have been investigated. Four studies were made to assess these effects: 1% and 0.5% KI ozonesonde performance was compared directly in three dual flights, one year of 1% KI and one year of 0.5% KI ozonesonde profiles were compared with near-simultaneous lidar profiles, integrated ozonesonde profiles were compared with Dobson spectrophotometer measurements over the same period, and ascent and descent profiles were compared for both KI concentrations. Ozonesondes flown with a 1% KI solution showed positive differences of 3% to 8% in the ozone profile and ∼5% in the ozone column compared with the 0.5% KI ozonesondes, which also showed better agreement in profiles and ozone column compared with the lidar and Dobson spectrophotometer measurements respectively.

Analysis of Record-Low Ozone Values During the 1997 Winter over Lauder, New Zealand.

Record-low ozone (O3) column densities (with a minimum of 222 DU) were observed over the Lauder NDSC (Network for the Detection of Stratospheric Change) sta- tion (45S, 170E) in August 1997. Possible causes are ex- amined using height-resolved O3 measurements over Lauder, and high-resolution reverse trajectory maps ofO3 (initialised with Microwave Limb Sounder measurements) and of poten- tial vorticity. The analysis shows thatO3 poor air originated from two regions: Below the 550 K isentrope (22 km) sub- tropical air was observed, while between 600 and 1000 K (25 { 33.5 km) the polar vortex tilted over Lauder for sev- eral days. A rapid recovery of the O3 column density was observed later, due to an O3 rich polar vortex lament mov- ing over Lauder between 18 and 24 km, while simultaneously the O3 poor higher vortex moved away.

OPAL: Network for the Detection of Stratospheric Change ozone profiler assessment at Lauder, New Zealand 2. Intercomparison of revised results

Following a blind intercomparison of ozone profiling instruments in the Network for the Detection of Stratospheric Change at Lauder, New Zealand, revisions to the analyses were made resulting in a new data set. This paper compares the revised results from two differential absorption lidars (RIVM and GSFC), a microwave radiometer (Millitech/LaRC), and electrochemical concentration cell (ECC) balloon sondes (NIWA). In general, the results are substantially improved compared to the earlier blind intercomparison. The level of agreement was similar both for single profiles and for the campaign average profile and was approximately 5% for the lidars and the sondes over the altitude range from 15 to 42 km (32 km for sondes). The revised microwave data show a bias of 5–10% high in the region from 22 to 42 km. Starting at 42 km, the lidar errors increase significantly, and comparisons of the microwave results were not possible above this altitude.

OPAL: Network for the Detection of Stratospheric Change Ozone Profiler Assessment at Lauder, New Zealand. 1. Blind Intercomparison.

An intercomparison of ozone-profiling instruments, two differential absorption lidars, a microwave radiometer, electrochemical concentration sondes, and the SAGE II satellite instrument is presented. The ground-based instruments were located at the Network for the Detection of Stratospheric Change (NDSC) primary station at Lauder, New Zealand. The campaign, which took place between April 15 and 29, 1995, strictly followed the NDSC guidelines for a blind intercomparison. Agreement between the measurements was within 15% for single profiles and within 10% for the campaign average, in the region from 20 to 40 km altitude. Outside of this region the differences were greater but can generally be ascribed to the limits of a particular instrument.

Comparison of infrared and Dobson total ozone columns measured from Lauder, New Zealand

Ozone total columns have been derived from 13 spectral intervals in 5 infrared bands and compared with values deduced from correlative measurements with a Dobson spectrophotometer. The observations were recorded on 10 days in 1994 at the Network for the Detection of Stratospheric Change station in Lauder, New Zealand. The infrared total columns were derived from spectral fittings of unblended, temperature-insensitive ozone lines in high resolution solar absorption spectra. The line parameters on the 1992 HITRAN compilation were assumed with the O3 H2O relative volume mixing ratio and temperature profiles specified from correlative balloon ozonesonde, microwave O3, and radiosonde measurements. The retrieved IR/Dobson total column ratios ranged from 0.96 to 1.02 with the lower wavenumber bands yielding lower ratios. The results do not support the revised O3 intensity scale currently used to process O3 infrared measurements from 2 instruments on the Upper Atmosphere Research Satellite.

NDSC millimeter wave ozone observations at Lauder, New Zealand, 1992–1998: Improved methodology, validation, and variation study

A ground-based millimeter wave radiometer for the Network for the Detection of Stratospheric Change (NDSC) was installed at Lauder, New Zealand (45°S, 169.7°E) in November 1992. It has been monitoring the middle atmospheric ozone with nearly continuous operation since then. Owing to special complications in the observing conditions at this southern midlatitude site, three refinements to the data analysis and calibration techniques were proposed: (1) the use of a radiative model of local tropospheric climate adopted to the low surface elevation of the observing site, (2) the correction of observing angle measurements due to the settling of the foundation of the site, and (3) the improved method of radiometric temperature determination of calibration sources. All data from 1992 to 1998 were reprocessed with these modifications implemented. The retrieved ozone profiles are compared to sonde, two lidars, and satellite (Halogen Occultation Experiment (HALOE), Stratospheric Aerosol and Gas Experiment (SAGE II)) overpass measurements. The agreement is very good, with mean differences from 56 to 1 mbar of generally 2–3% for the comparisons with sonde, HALOE, and SAGE II, and generally <5% for the comparisons with lidars when large samples are considered. The root-mean-square scatter about the mean differences is mostly consistent with the expected (combined) precision. In comparisons with these correlative measurements, the millimeter wave ozone observations are found to have no seasonal bias, no comparison bias due to the a priori profiles used in millimeter wave data retrievals, and no observable instrument drift from 1992 to 1998. Better agreement is found in the comparison with sonde data if suspected vertical shifts in the sonde profiles are considered. The variations seen in the 6 year millimeter wave ozone data are shown to be mostly in line with photochemistry and dynamic transport processes in the mid austral latitudes. These processes, however, are apparently modulated somewhat by the Antarctic polar vortex circulation in winter seasons in the upper stratosphere.

The total ozone anomaly at Lauder, NZ in 1997

A record low annual mean value of total ozone was observed at Lauder, New Zealand in 1997, namely 294.5 DU, compared to a 1985–96 mean of 307.9±3.7. Ozone values were low through most of the year, especially in winter and spring. These observations are similar to TOMS results for Southern mid-latitudes, although the long-term comparison is clouded by the change in TOMS instruments. A regression analysis of the Lauder record through 1996 allows a generally accurate prediction of the 1997 values, shows that the QBO dominates the 1996–97 difference, and that the 1997 values are consistent with a linear long-term trend in ozone. The analysis does not completely explain low monthly values in austral spring. Enhancements in biologically active UV radiation were also observed, with monthly mean values increased by as much as 20–40% from recent years. The unusual ozone values were the dominant cause of the UV enhancement.