-
During the 2003 SAGE (Stratospheric Aerosol and Gas Experiment) III Ozone Loss and Validation Experiment (SOLVE) II, the fourteen-channel NASA Ames Airborne Tracking Sunphotometer (AATS-14) was mounted on the NASA DC-8 aircraft and measured spectra of total and aerosol optical depth (TOD and AOD) during the sunlit portions of eight science flights. Values of ozone column content above the aircraft have been derived from the AATS-14 measurements by using a linear least squares method that exploits the differential ozone absorption in the seven AATS-14 channels located within the Chappuis band. We compare AATS-14 columnar ozone retrievals with temporally and spatially near-coincident measurements acquired by the SAGE III and the Polar Ozone and Aerosol Measurement (POAM) III satellite sensors during four solar occultation events observed by each satellite. RMS differences are 19 DU (7% of the AATS value) for AATS-SAGE and 10 DU (3% of the AATS value) for AATS-POAM. In these checks of consistency between AATS-14 and SAGE III or POAM III ozone results, the AATS-14 analyses use airmass factors derived from the relative vertical profiles of ozone and aerosol extinction obtained by SAGE III or POAM III. <P style="line-height: 20px;"> We also compare AATS-14 ozone retrievals for measurements obtained during three DC-8 flights that included extended horizontal transects with total column ozone data acquired by the Total Ozone Mapping Spectrometer (TOMS) and the Global Ozone Monitoring Experiment (GOME) satellite sensors. To enable these comparisons, the amount of ozone in the column below the aircraft is estimated either by assuming a climatological model or by combining SAGE and/or POAM data with high resolution in-situ ozone measurements acquired by the NASA Langley Research Center chemiluminescent ozone sensor, FASTOZ, during the aircraft vertical profile at the start or end of each flight. Resultant total column ozone values agree with corresponding TOMS and GOME measurements to within 10-15 DU (~3%) for AATS data acquired during two flights - a longitudinal transect from Sweden to Greenland on 21 January, and a latitudinal transect from 47° N to 35° N on 6 February. For the round trip DC-8 latitudinal transect between 34° N and 22° N on 19-20 December 2002, resultant AATS-14 ozone retrievals plus below-aircraft ozone estimates yield a latitudinal gradient that is similar in shape to that observed by TOMS and GOME, but resultant AATS values exceed the corresponding satellite values by up to 30 DU at certain latitudes. These differences are unexplained, but they are attributed to spatial and temporal variability that was associated with the dynamics near the subtropical jet but was unresolved by the satellite sensors. <P style="line-height: 20px;"> For selected cases, we also compare AATS-14 ozone retrievals with values derived from coincident measurements by the other two DC-8 based solar occultation instruments: the National Center for Atmospheric Research Direct beam Irradiance Airborne Spectrometer (DIAS) and the NASA Langley Research Center Gas and Aerosol Monitoring System (GAMS). AATS and DIAS retrievals agree to within RMS differences of 1% of the AATS values for the 21 January and 19-20 December flights, and 2.3% for the 6 February flight. Corresponding AATS-GAMS RMS differences are ~1.5% for the 21 January flight; GAMS data were not compared for the 6 February flight and were not available for the 19-20 December flight. Line of sight ozone retrievals from coincident measurements obtained by the three DC-8 solar occultation instruments during the SAGE III solar occultation event on 24 January yield RMS differences of 2.1% for AATS-DIAS and 0.5% for AATS-GAMS.
-
We report on the vertical distributions of Saharan dust aerosols over the N.E. Mediterranean region, which were obtained during a typical dust outbreak on August 2000, by two lidar systems located in Athens and Thessaloniki, Greece, in the frame of the European EARLINET project. MODIS and ground sun spectrophotometric data, as well as air-mass backward trajectories confirmed the existence of Saharan dust in the case examined, which was also successfully forecasted by the DREAM dust model. The lidar data analysis for the period 2000-2002 made possible, for the first time, an estimation of the vertical extent of free tropospheric dust layers [mean values of the aerosol backscatter and extinction coefficients and the extinction-to-backscatter ratio (lidar ratio, LR) at 355 nm], as well as a seasonal distribution of Saharan dust outbreaks over Greece, under cloud-free conditions. A mean value of the lidar ratio at 355 nm was obtained over Athens (53±1 sr) and over Thessaloniki (44±2 sr) during the Saharan dust outbreaks. The corresponding aerosol optical thickness (AOT) at 355 nm, in the altitude range 0-5 km, was 0.69±0.12 and 0.65±0.10 for Athens and Thessaloniki, respectively (within the dust layer the AOT was 0.23 and 0.21, respectively). Air-mass back-trajectory analysis performed in the period 2000-2002 for all Saharan dust outbreaks over the N.E. Mediterranean indicated the main pathways followed by the dust aerosols.
-
Four halons (H-1301, H-1211, H-2402 and H-1202) have been measured in air samples collected from polar firn from Dome Concordia (Dome C), Antarctica, from Devon Island, Canada and the North Greenland Ice Core Project (NGRIP) site, Greenland. H-2402 and H-1202 are reported for the first time in firn air. The depth profiles show the concentrations of all four halons to be close to zero (i.e. below the detection limit of 0.001 ppt) at the base of the firn thus demonstrating their entirely anthropogenic origin. This is the first evidence of this for H-2402 and H-1202. A 2-D atmospheric model was run with emissions previously derived using archive air measurements from the southern hemisphere mid-latitudes to produce historical trends in atmospheric concentrations at the firn sites, which were then input into a firn diffusion model to produce concentration depth profiles for comparison with the firn measurements. This comparison provides an evaluation of the model-derived atmospheric concentration histories in both hemispheres and thus an indirect evaluation of the emission rates and distributions used in the atmospheric model. Atmospheric concentration trends produced using global emissions previously determined from measurements at Cape Grim are found to be consistent with the firn data from the southern hemisphere. Further emissions of H-1202 in recent years (late 1980s onwards) are likely to have come from latitudes mostly south of 40° N, but emissions prior to that may have come from further north. Emissions of H-1211 may also have shifted to latitudes south of 40° N during the late 1980s. Following comparison of the atmospheric model output with the firn data, modelled atmospheric trends of total organic bromine in the form of halons were derived for both polar regions.
-
Ozone profiles retrieved from limb scattering measurements of the SCIAMACHY instrument based on the satellite ENVISAT are compared to ground-based low altitude resolution remote sensors. All profiles are retrieved using optimal estimation. Following the work of Rodgers and Connor (2003) the retrievals of the ground-based instruments are simulated using the SCIAMACHY retrieval. The SCIAMACHY results and the results of the ground-based microwave radiometer in Bremen and Ny Ålesund agree within the expected covariance of the intercomparison.
-
The extent of springtime Arctic ozone loss does not reach Antarctic ``ozone hole' dimensions because of the generally higher temperatures in the northern hemisphere vortex and consequent less polar stratospheric cloud (PSC) particle surface for heterogeneous chlorine activation. Yet, with increasing greenhouse gases stratospheric temperatures are expected to further decrease. To infer if present Antarctic PSC occurrence can be applied to predict future Arctic PSC occurrence, lidar observations from McMurdo station (78° S, 167° E) and NyÅlesund (79° N, 12° E) have been analysed for the 9 winters between 1995 (1995/1996) and 2003 (2003/2004). Although the statistics may not completely cover the overall hemispheric PSC occurrence, the observations are considered to represent the main synoptic cloud features as both stations are mostly situated in the centre or at the inner edge of the vortex. Since the focus is set on the occurrence frequency of solid and liquid particles, the analysis has been restricted to volcanic aerosol free conditions. In McMurdo, by far the largest part of PSC observations is associated with NAT PSCs. The observed persistent background of NAT particles and their potential ability to cause denoxification and irreversible denitrification is presumably more important to Antarctic ozone chemistry than the scarcely observed ice PSCs. Meanwhile in Ny-Ålesund, ice PSCs have never been observed, while solid NAT and liquid STS clouds both occur in large fraction. Although they are also found solely, the majority of observations reveals solid and liquid particle layers in the same profile. For the Ny-Ålesund measurements, the frequent occurrence of liquid PSC particles yields major significance in terms of ozone chemistry, as their chlorine activation rates are more efficient. <P style="line-height: 20px;"> The relationship between temperature, PSC formation, and denitrification is nonlinear and the McMurdo and Ny-Ålesund PSC observations imply that for predicted stratospheric cooling it is not possible to directly apply current Antarctic PSC occurrence to the Arctic stratosphere. Future Arctic PSC occurrence, and thus ozone loss, is likely to depend on the shape and barotropy of the vortex rather than on minimum temperature alone.
-
A transient simulation with the interactively coupled chemistry-climate model (CCM) E39/C has been carried out which covers the 40-year period between 1960 and 1999. Forcing of natural and anthropogenic origin is prescribed where the characteristics are sufficiently well known and the typical timescales are slow compared to synoptic timescale so that the simulated atmospheric chemistry and climate evolve under a "slowly" varying external forcing. Based on observations, sea surface temperature (SST) and ice cover are prescribed. The increase of greenhouse gas and chlorofluorocarbon concentrations, as well as nitrogen oxide emissions are taken into account. The 11-year solar cycle is considered in the calculation of heating rates and photolysis of chemical species. The three major volcanic eruptions during that time (Agung, 1963; El Chichon, 1982; Pinatubo, 1991) are considered. The quasi-biennial oscillation (QBO) is forced by linear relaxation, also known as nudging, of the equatorial zonal wind in the lower stratosphere towards observed zonal wind profiles. Beyond a reasonable reproduction of mean parameters and long-term variability characteristics there are many apparent features of episodic similarities between simulation and observation: In the years 1986 and 1988 the Antarctic ozone holes are smaller than in the other years of that decade. In mid-latitudes of the Southern Hemisphere ozone anomalies resemble the corresponding observations, especially in 1985, 1989, 1991/1992, and 1996. In the Northern Hemisphere, the episode between the late 1980s and the first half of the 1990s is dynamically quiet, in particular, no stratospheric warming is found between 1988 and 1993. As observed, volcanic eruptions strongly influence dynamics and chemistry, though only for few years. Obviously, planetary wave activity is strongly driven by the prescribed SST and modulated by the QBO. Preliminary evidence of realistic cause and effect relationships strongly suggests that detailed process-oriented studies will be a worthwhile endeavour.
-
Trends in cirrus cloud cover have been estimated based on 16 years of data from ISCCP (International Satellite Cloud Climatology Project). The results have been spatially correlated with aircraft density data to determine the changes in cirrus cloud cover due to aircraft traffic. The correlations are only moderate, as many other factors have also contributed to changes in cirrus. Still we regard the results to be indicative of an impact of aircraft on cirrus amount. The main emphasis of our study is on the area covered by the METEOSAT satellite to avoid trends in the ISCCP data resulting from changing satellite viewing geometry. In Europe, which is within the METEOSAT region, we find indications of a trend of about 1-2% cloud cover per decade due to aircraft, in reasonable agreement with previous studies. The positive trend in cirrus in areas of high aircraft traffic contrasts with a general negative trend in cirrus. Extrapolation in time to cover the entire period of aircraft operations and in space to cover the global scale yields a mean estimate of 0.03 Wm<sup>-2</sup> (lower limit 0.01, upper limit 0.08 Wm<sup>-2</sup>) for the radiative forcing due to aircraft induced cirrus. The mean is close to the value given by IPCC (1999) as an upper limit.
-
The chemical nature of a large mass fraction of ambient organic aerosol particles is not known. High molecular weight compounds (often named humic-like substances) have recently been detected by several authors and these compounds seem to account for a significant fraction of the total organic aerosol mass. Due to the unknown chemical structure of these compounds quantification as well as a determination of their molecular weight is difficult. In this paper we investigate water soluble humic-like substances in ambient urban aerosol using size exclusion chromatography-UV spectroscopy and laser desorption/ionization mass spectrometry (LDI-MS). LDI-MS was used for the first time to investigate HULIS from atmospheric aerosols. A careful evaluation of the two method shows that both methods complement each other and that both are needed to learn more about the molecular weight distribution and the concentration of humic-like substances. An upper molecular weight limit of humic-like substances of about 700 Da and a concentration of 0.3-1.6 µg/m<sup>3</sup> air can be estimated, corresponding to 9-30% of the total organic carbon for an urban background site.
-
We perform surface pressure retrievals from cloud-free Oxygen A band measurements of SCIAMACHY. These retrievals can be well validated because surface pressure is a quantity that is, in general, accurately known from meteorological models. Therefore, surface pressure retrievals and their validation provide important insight into the quality of the instrument calibration. Furthermore, they can provide insight into retrievals which are affected by similar radiation transport processes, for example the retrieval of total columns of H<sub>2</sub>O, CO, CO<sub>2</sub> and CH<sub>4</sub>. In our retrieval aerosols are neglected. Using synthetic measurements, it is shown that for low to moderate surface albedos this leads to an underestimation of the retrieved surface pressures. For high surface albedos this generally leads to an overestimation of the retrieved surface pressures. The surface pressures retrieved from the SCIAMACHY measurements indeed show this dependence on surface albedo, when compared to the corresponding pressures from a meteorological database. However, an offset of about 20 hPa was found, which can not be caused by neglecting aerosols in the retrieval. The same offset was found when comparing the retrieved surface pressures to those retrieved from co-located GOME Oxygen A band measurements. This implies a calibration error in the SCIAMACHY measurements. By adding an offset of 0.86% of the continuum reflectance at 756 nm to the SCIAMACHY reflectance measurements, this systematic bias vanishes.
-
Verification of modeled rainfall with precipitation observed by a rain gauge network has been performed in a case study over the Cyprus Island. Cyprus has a relatively dense rain gauge network. The applied verification method is the Contiguous Rain Area (CRA) analysis. Some drawbacks of the CRA method are pointed out when it is applied to such a case study. Impact on the CRA results, when considering different dimensions of the verification sub-domain and different types of indicators (correlation and mean square error) used in the comparison, are discussed. Results indicate that care should be taken when verification of modeled rainfall is performed over a domain smaller than the model one.
