Ya. A. Virolainen

Ground-based measurements of total ozone content by the infrared method

To interpret the ground-based measurements of the spectra of direct solar infrared radiation with the help of a Brucker Fourier-spectrometer, a technique for determining the total ozone content (TOC) was developed and implemented. The TOC was determined using six spectral intervals of an ozone-absorption band of 9.6 μm and the shortwave panel of a carbon-dioxide-absorption band of 15 μm, where the impact of other atmospheric parameters on the measured solar radiation was reduced to a minimum. The potential errors of the infrared method for determining the TOC for the chosen spectral scheme with the influence of measurement errors and vertical profiles of temperature are less than 1% for different signal-to-noise ratios and zenith angles of the sun. We analyzed 269 high-resolution (0.005–0.008 cm−1) spectra of solar infrared radiation measured in Peterhof over 52 days from March to November, 2009. The resulting values of TOC were compared with the results of independent ground-based TOC measurements in Voeikovo (Main Geophysical Observatory) using a Dobson spectrophotometer and an M-124 ozonometer, as well as with the Ozone Monitoring Instrument (OMI) satellite data. The mean errors between the results of TOC measurements with the help of the three ground-based probes constitute no more than 0.4%. The rms errors between data obtained by the Brucker spectrometer and the given satellite and ground-based probes constitute 3–4%. A comparison between different series of measurements indicated that the upper estimate for the error of TOC measurements by the Brucker spectrometer was 2.5–3% (when the possible spatial and temporal errors in measurements are disregarded). An analysis of the diurnal variations in the TOC measurements for stable atmospheric conditions yields an upper estimate of ∼3 DU (around 1%) for the random component of error in TOC measurements by the Brucker spectrometer.

Intercomparison of satellite and ground-based ozone total column measurements

Ozone total column (OTC) measurements made in 2009–2012 near St. Petersburg by a Fourier Transform Infrared (FTIR) spectrometer (Peterhof, St. Petersburg State University (SPbSU)), an M-124 filter ozonometer, and a Dobson spectrophotometer (Voeikovo, MGO), as well as measurements made by a spectrometer ozone monitoring instrument (OMI) (onboard the AURA satellite) have been analyzed and compared. Comparisons have been performed both between ensembles of ground-based measurement data, as well as between ground-based and satellite data. It has been shown that the standard deviation for all devices is 2.5–4.5%; here, the FTIR and Dobson instruments measuring the direct sun are in better agreement with OMI than the M-124 ozonometer measuring the zenith-scattered solar radiation as well. A seasonal cycle in discrepancy with amplitude of 1.5% has been detected between two series of OTC measurements made by M-124 and OMI instruments for a total of 850 days. In fall and winter, the ground-based measurements underestimate the OTC values in comparison with satellite data; in spring and summer, the situation is reversed: ground-based data overestimate the OTC values. Also, it has been revealed that FTIR measurements systematically overestimate the OTC values in comparison with other instruments: from 1.4% (for Dobson) to 3.4% (for OMI). Taking into account the spatial and temporal discrepancy of independent ensembles of measurements and an analysis of standard deviations between ground-based and satellite measurement data, the FTIR spectrometer (SPbSU) can be recommended for OTC satellite data validation.

Comparison of ground-based FTIR and radio sounding measurements of water vapor total content

We compared two datasets of the total content of atmospheric water vapor received near St. Petersburg in 2009–2012 from ground-based Fourier transform spectroscopy measurements at the Peterhof station and from radio sounding at Voyeykovo station. Despite a good correlation of daily measurements in Peterhof and Voyeykovo, the standard mismatch is significant, 20% or more, for most subsets taken for the comparison. The high mismatch is mainly due to the natural spatial variability of the total content of water vapor, accounting for the 50-km distance between Peterhof and Voyeykovo. This variability needs to be considered when validating the satellite measurements of water vapor content by ground-based measurements.

Evaluation of ozone content in different atmospheric layers using ground-based Fourier transform spectrometry

For the first time in Russia, using ground-based measurements of direct solar infrared radiation, we derived data on ozone content in different layers of the atmosphere. The measurements were conducted with the help of a Bruker IFS-125HR Fourier spectrometer in 2009–2012 in Petergof, which is 30 km west of the center of St. Petersburg. The errors in determining the ozone content by this method in the troposphere (0–12 km), in the stratosphere (12–50 km), in the layers of 10–20 and 20–50 km, and in the layers of 12–18, 18–25, and 25–50 km were ~4, 3, 3–5, and 4–7% (taking into account the instrumental and methodological errors, as well as the errors in specifying the temperature profile), respectively. The seasonal variation of tropospheric ozone content in the layer of 12–18 km is characterized by a clearly expressed maximum in March and a minimum in November, with amplitudes of 30 and 40%, respectively. For the layer of 18–25 km, the maximum and minimum are reached in the winter-spring period and late summer, respectively; the amplitude of the seasonal variation is ~20%. The amplitude of the annual variation in ozone content in the layer of 25–50 km is around 30%, with a maximum close to the summer solstice and a minimum close to the winter solstice. Over the three years of observations, the growth in the ozone content in this layer was ~10% per year of its value averaged over the time period. Comparisons of ground-based measurements with satellite measurements (by the IASI instrument) of tropospheric ozone revealed a discrepancy of (3.4 ± 17)% for both ensembles. The correlation between the two ensembles is 0.76–0.84 (depending on the season). Comparisons between ground-based and satellite measurements (by the MLS instrument) of stratospheric ozone revealed no systematic discrepancies of the two ensembles. The rms errors were 13, 6, and 5% for the layers of 10–20, 20–50, and 10–50 km, respectively; the coefficients of correlations between the two types of measurements were 0.82–0.94.

Polar stratospheric clouds from satellite observational data

The spectral aerosol-extinction coefficients (SAECs) obtained from SAGE III measurements are used to study the physical and integral microphysical characteristics of polar stratospheric clouds (PSCs). Different criteria for PSC identification from SAEC measurements are considered and analyzed based on model and field measurements. An intercomparison of them is performed, and the agreement and difference of the results obtained with the use of different criteria are shown. A new criterion is proposed for PSC identification, which is based on the estimate of how close the measured vector of the spectral attenuation coefficient is to a model distribution of the PSC ensemble. On the basis of different criteria, cases of PSCs are isolated from all SAGE III observations (over 30000). All selection criteria lead to a qualitatively and quantitatively similar space-time distribution of the regions of PSC localization. The PSCs observed in the region accessible to SAGE III measurements are localized in the latitudinal zones 65°–80° in the Northern Hemisphere and 45°–60° in the Southern Hemisphere during the winter-spring period. In the Northern Hemisphere, PSCs are observed within the longitudinal zone 120° W–100° E with the maximum frequency of PSC observation in the vicinity of the Greenwich meridian. In the Southern Hemisphere, the region of PSC observation is almost the same in longitude but with a certain shift in the maximum frequency of PSC observation to the west. This maximum is observed in the vicinity of 40°W, and the region of usual PSC observation is the neighborhood of 60° of the maximum’s longitude. The physical parameters of PSCs are estimated: the mean heights of the lower and upper boundaries of PSCs are 19.5 and 21.9 km, respectively, and the mean cloud temperature is 191.8 K. The integral microphysical parameters of PSCs are estimated: the total surface of NAT particles SNAT = 0.41 μm2/cm3; the total volume of NAT particles VNAT = 1.1 μm3/cm3; and, for all aerosol and cloud particles together, S is 2.9 ± 1.5 at a standard deviation of 2.7 μm2/cm3 and V is 2.8 ± 1.5 at a standard deviation of 4.2 μm3/cm3. A high frequency of PSC occurrence and high values of S and V in PSCs both for all particles and for NAT particles have been noted in January–February 2005 as compared to the rest of the period of SAGE III measurements for 2002–2005.

Analysis of Solutions to the Inverse Problem on the Retrieval of the Microstructure of Stratospheric Aerosol from Satellite Measurements

A statistical ensemble of microphysical parameters of the background stratospheric aerosol at altitudes of 15 to 30 km is modeled on the basis of experimental data. The aerosol attenuation coefficients (AACs) in the wavelength range 0.38–16.3 μm are calculated for all realizations of the ensemble by algorithms of the Mie theory. Analysis of correlations between the AACs and the microphysical parameters indicate that the AAC correlates most strongly with the total volume V and area S of all particles. The errors of determining the microphysical parameters from AAC measurements are analyzed via the method of linear regression. It is shown that, if the AAC is measured with an error of 5%, the errors of determining both the particle size distribution (PSD) for particles with sizes of 0.4 to 4 μm and the parameter S are an order of magnitude smaller than the prior uncertainty, whereas the error of determining V is two orders of magnitude smaller than the prior uncertainty. Schemes of AAC measurements with the SAGE III, ISAMS, CLAES, HALOE instruments and an IR interferometer in the visible and IR regions are discussed. It is shown that combining the schemes makes it possible to extend the range of particle sizes for which the PSD is retrieved with a satisfactory accuracy and to increase the accuracy of determining S and V substantially and the accuracy of determining the total number of particles Nopt to a lesser extent. Examples of interpreting AAC measurements carried out simultaneously with the SAGE III and HALOE instruments within the same spatial region are presented. A systematic discrepancy between vertical profiles of S and V obtained from SAGE III and HALOE measurements is revealed.

Comparing data obtained from ground-based measurements of the total contents of O3, HNO3,HCl, and NO2 and from their numerical simulation

Chemistry climate models of the gas composition of the atmosphere make it possible to simulate both space and time variations in atmospheric trace-gas components (TGCs) and predict their changes. Both verification and improvement of such models on the basis of a comparison with experimental data are of great importance. Data obtained from the 2009–2012 ground-based spectrometric measurements of the total contents (TCs) of a number of TGCs (ozone, HNO3, HCl, and NO2) in the atmosphere over the St. Petersburg region (Petergof station, St. Petersburg State University) have been compared to analogous EMAC model data. Both daily and monthly means of their TCs for this period have been analyzed in detail. The seasonal dependences of the TCs of the gases under study are shown to be adequately reproduced by the EMAC model. At the same time, a number of disagreements (including systematic ones) have been revealed between model and measurement data. Thus, for example, the EMAC model underestimates the TCs of NO2, HCl, and HNO3, when compared to measurement data, on average, by 14, 22, and 35%, respectively. However, the TC of ozone is overestimated by the EMAC model (on average, by 12%) when compared to measurement data. In order to reveal the reasons for such disagreements between simulated and measured data on the TCs of TGCs, it is necessary to continue studies on comparisons of the contents of TGCs in different atmospheric layers.

Interannual and seasonal variations in ozone in different atmospheric layers over St. Petersburg based on observational data and numerical modeling

This paper analyzes atmospheric ozone variability at different altitudes over St. Petersburg for the period 2009–2014 on the basis of surface observations at the Peterhof station, satellite measurements with an SBUV instrument, and numerical simulations. Simulation data on temperature, wind velocity, humidity, and surface pressure are taken from the MERRA reanalysis database. Based on ozone measurements, numerical modeling, and reanalysis data, characteristics of ozone seasonal and interannual changes are identified; the role of photochemical and dynamic factors in ozone variations is estimated.

Using artificial neural networks in the temperature and humidity sounding of the atmosphere

The application of the radiative data inversion technique based on artificial neural networks (ANN) for the meteorological satellite sounding of the atmosphere is described. To increase the efficiency of solving inverse problems, the principal component method is used for the temperature and humidity profiles, as well as for IR radiation spectra, which allows the problem dimensionalities to be reduced substantially. Based on numerical experiments, errors of the temperature and humidity sounding are analyzed from the spectra of outgoing IR radiation (that were measured by the IKFS-2 instrument onboard the Meteor Russian satellite) using the iterative physical-mathematical (IPM) algorithm, multiple linear regression (MLR), and ANN-based methods. Appreciable advantages of the ANN-based method are revealed as compared to the MLR method. Therefore, in temperature sounding, the MLR method has a markedly large error at heights of 1–12 km (a difference of up to 1 K), while the IPM algorithm has almost the same error as the ANN method. The humidity determination error is about 10% when the ANN method is used at heights of 0–12 km. The IPM approach yields approximately the same error in the lower troposphere, but as the height increases the advantages of the ANN method grow.

Optimal parameterization of the spectra of outgoing thermal radiation with the data of the IKFS-2 spaceborne IR sensing device taken as an example

From the ensemble of the calculated spectra of the outgoing thermal radiation in the 660- to 2010-cm−1 range (2311 realizations) simulating global measurements by the IKFS-2 spaceborne device, the informativeness of the measurements of the outgoing thermal radiation has been analyzed in terms of Kozlov’s data volume, the degrees of freedom, and the Shannon information gain. In the entire spectral range (660–2010 cm−1), there are 106 independent parameters in the measurements of the outgoing radiation. The accuracy of the optimal parameterization of the spectral behavior of the radiation based on the radiation resolution into eigenvectors of the spectral covariance matrix has been analyzed. It is shown that to reach an rms error of the parameterization comparable with the random measurement noise for different spectral regions, it is sufficient to use from 20 to 50 first eigenvectors.