Michael Bittner

An 18-year time series of OH rotational temperatures and middle atmosphere decadal variations

Abstract Rotational temperatures are derived from measurements of the (3-1) transitions of the OH Meinel bands. They are found to be a reasonable proxy for atmospheric kinetic temperatures at an altitude of 3×10 −3 hPa (87 km ). Measurements were taken in Wuppertal (51°N/7°E) from 1980 to 1998. A second (shorter) data set (1980–1991) is available for Northern Scandinavia (Andoya, 69°N/16°E, and Kiruna, 68°N/21°E). Data are analyzed for seasonal variations (annual, semi-annual, terannual) and for long-term changes of the temperature. Furthermore atmospheric wave activity is studied by means of measured short-term temperature fluctuations. Amplitudes and phases of the seasonal components do not change much during the measurement period. On the contrary annual mean temperatures and wave activity show strong long-term and even decadal variations. Similar decadal “episodes” are found in data from the Stratospheric Sounding Units (SSU) and from radiosondes at lower altitudes ( 1 hPa /47 km and 10 hPa /31 km , respectively). Temperature variations during these episodes are rather large ( 10 K ) and need to be taken into account when using reference or standard atmospheres. Episodic variations of temperature and wave activity are found to be correlated. They appear to be consistent with the “downward control principle” (Rev. Geophys. 33 (1995) 403).

Intercomparison of density and temperature profiles obtained by lidar, ionization gauges, falling spheres, datasondes and radiosondes during the DYANA campaign

Abstract During the course of the DYANA campaign in early 1990, various techniques to measure densities and temperatures from the ground up to the lower thermosphere were employed. Some of these measurements were performed near simultaneously (maximum allowed time difference: 1 h) and at the same location, and therefore offered the unique chance of intercomparison of different techniques. In this study, we will report on intercomparisons of data from ground-based instruments (Rayleigh- and sodium-lidar), balloon-borne methods (datasondes and radiosondes) and rocket-borne techniques (falling spheres and ionization gauges). The main result is that there is good agreement between the various measurements when considering the error bars. Only occasionally did we notice small but systematic differences (e.g. for the datasondes above 65 km). The most extensive intercomparison was possible between the Rayleigh lidar and the falling sphere technique, both employed in Biscarrosse (44°N,1°W). Concerning densities, excellent agreement was found below 63 km: the mean of the deviations is less than 1 % and the root mean square (RMS) is ~ 3%. Systematic differences of the order of 5% were noticed around 67 km and above 80 km. The former can be accounted for by an instrumental effect of the falling sphere (Ma = 1 transition; Ma = Mach number), whereas the latter is tentatively explained by the presence of Mie scatterers in the upper mesosphere. Concerning temperatures, the agreement is excellent between 35 and 65 km: the mean of the deviations is less than ± 3 K and the variability is ± 5 K. The two systematic density differences mentioned above also affect the temperatures: between 65 and 80 km, the Rayleigh lidar temperatures are systematically lower than the falling sphere values by ~ 5 K.

Long period/large scale oscillations of temperature during the DYANA campaign

Abstract During the DYANA campaign (January–March 1990) vertical temperature profiles were measured in the middle atmosphere above 11 rocket and four lidar stations in the northern hemisphere. Strong temperature variations were observed, especially at the medium to high latitude stations. Time series analysis was applied, and most oscillations were found to be quasi-periodic only, and restricted to certain altitude levels. Solely the longest periods (>4 weeks) and the shortest periods (around 5 days) were consisterrtly observed in the whole altitude regime, and were, therefore, further analysed. These temperature variations were found to be compatible with the assumption that the Quasi-Stationary Planetary Wave No. 1 (QSW 1) of the CIRA 1990 Model Atmosphere was modulated with the respective periods. Especially close similarity of the vertical phase structures was obtained. The amplitudes measured were, however, much larger than those of the model, and hence some amendment to the model may be appropriate. The importance of QSW 1 modulation appears to be considerable, as more than 50% of the temperature variance could be ascribed to it. The DYANA results were checked by an analysis of respective data from two other campaigns (Winter Anomaly campaign, 1976, and MAP/WINE campaign, 1984). Essentially the same results were obtained. Considering the strength of QSW 1 control, the midwinter middle atmosphere shows itself to be quite strongly and persisterrtly structured in both the vertical and horizontal directions. This applies to all parts of the stratosphere and mesosphere.

Large-scale structures in hydroxyl rotational temperatures during DYANA

Abstract Comparison of nightly means of OH∗ rotational temperatures measured at Yakutsk, Kiruna, Andoya, Biscarrosse, and El Arenosillo during the DYANA campaign reveals information about dynamics in the 86 km altitude region that includes relatively short-range temperature gradients at high latitudes associated with a stratospheric warming, and long-range oscillations with periods from about 2 to more than 30 days that can be attributed to travelling or quasi-stationary planetary waves. The DYANA campaign averages terrd to exceed zonal mean temperatures of the CIRA 86 model by up to 15 K.

Gravity wave characteristics in the middle atmosphere derived from the Empirical Mode Decomposition method.

The recently developed Empirical Mode Decomposition (EMD) method is applied to analyzing gravity wave characteristics in the middle atmosphere. By establishing a close connection between the fundamental Intrinsic Mode Functions (IMFs) derived from the EMD method and WKB solutions of a dispersive-dissipative wave equation, we show that the EMD method can provide useful insights into physical processes in the middle atmosphere where dispersive-dissipative wave phenomena are dominant. A local power spectrum function P is introduced which provides a quantitative description of the spectrum at any particular location within a data series. The sharp localization of P in space and wavenumber leads to an identification of unphysical small scale oscillations by falling spheres embedded in the wind profiles above 60 km. Further analyses of the horizontal wind profiles derived from the Dynamics Adapted Network for the Atmosphere (DYANA) campaign suggest that for horizontal wind fluctuations with vertical wavenumber m≤3 km−1 (or vertical wavelength Lz≥2 km) the previously observed m−3 Fourier spectra could be produced by a linear wave packet whose characteristic vertical wavenumber decreases with altitude. For small vertical scale disturbances with m>3 km−1 (Lz<2 km) a near −3 slope in the marginal distribution exists locally in the middle atmosphere with a great degree of universality, suggesting that nonlinear energy cascade processes may dominate the spectral formation in this wavenumber range.

Computer animation of remote sensing-based time series data sets

Visualization has always played a major role in the exploitation of remote sensing data sets as well as in the confirmation of scientific hypotheses. With the new techniques available for two-dimensional (2D) and three-dimensional (3D) computer animation, the synthesis of different information layers as well as high quality visualizations for the presentation of research tasks and results are becoming increasingly important and popular. The German Remote Sensing Data Center (DFD) of the German Aerospace Center (DLR) is operationally generating remote-sensing-based time series data sets. This data can be used to generate long term, high quality computer animations for analyzing and presenting the information contained in Earth observation data. Due to clouds or to system specification, data gaps occur in satellite derived time series, which preclude the generation of highest quality computer animations. For this reason different interpolation techniques have been developed primarily for atmospheric research, and they now prove to be a valuable tool for interpolation of a wide range of remote sensing data sets to be used for computer animations.

Vertical velocities measured at Biscarrosse (44°N) and by EISCAT at Tromsø (69.6°N) during the DYANA campaign

Abstract The vertical velocities of the atmosphere were measured at Biscarrosse (southwest France, 44°N) by the foil cloud method and at Tromso (northern Norway, 69.6°N) using incoherent scatter (EISCAT) during the DYANA campaign from January 1990 to March 1990. A new type of foil which had a fairly low velocity of descent alleviated the investigation of small-scale vertical structures. The vertical velocities measured at Biscarrosse by the foil cloud method turned out to be rather moderate (±2–3 m/s) but were found to be superimposed by short-wavelength vertical waves (1–1.4 km) with about half that amplitude (1–1.5 m/s). This was considered to be a major result and a maximum entropy, a bispectral and a triad analysis of the vertical velocities measured in DYANA salvo 2 (on 20 February 1990) suggest a non-linear wave-wave interaction (‘parametric amplification’). A detailed analysis of this aspect which includes all flights is considered necessary. Vertical velocities measured by EISCAT at Tromso turned out to be quite moderate and one sample (on 5–6 March 1990) is discussed. The large distance between Biscarrosse and Tromso (3064 km) renders a direct comparison quite difficult

Wavelet analysis of upper mesosphere temperature variations

Abstract More than 2000 mean night temperatures (1987–1995) were derived from OH* near infrared emissions in the upper mesosphere above Wuppertal (51 °N, 7 °E). Variations of 4–50 days period were analysed using maximum entropy and wavelet methods. Seasonal dependence of occurrence and long term evolution of the strength of these variations is investigated. Dependencies on solar activity and possibly to the quasi biennal oscillation (QBO) are believed to show up in these parameters.

The STREAMER Project: An overview

In recent years it has become evident that the distribution of ozone is strongly characterized by small scale structures both in time and space. This extraordinary variability is mostly due to the formation of finger-like structures (streamers) reaching out from equatorial regions, towards higher latitudes and obviously carrying ozone-low air towards densely populated areas such as Europe. As a consequence, strong modulations in the intensity of ground-level UV-B radiation is observed. Today, there is broad agreement that high UV dosages can cause considerable biological damage, to man, crops and animals. The objective of STREAMER (Small Scale Structure Early Warning and Monitoring in Atmospheric Ozone and Related Exposure to UV-B Radiation) is the development, installation and validation of an operational information and early warning system for providing daily nowcast and forecast of ozone vertical profiles, total column ozone and UV ground level intensity. Focus is on smaller scale structures such as streamers and mini-holes approaching Europe. STREAMER uses the end-to-end philosophy: it starts with the processing of the satellites bits and bytes, continues with their interpretation into meaningful information and finally ends up with the provision of a comfortable on-line information system that allows customers easy access to exactly that kind of information they require. For the realization of the projects objectives a consortium was established within the 4th Framework Programme of the European Commission: German Aerospace Center (coordination), German Weather Service, Federal Environment Agency Austria, German Federal Environment Agency, Aristotle University Thessaloniki (Greece), Instituto Nacional de Technica Aerospatial (Spain), Federal Office for Radiation Protection (Germany).

Global Gap-Free MERIS LAI Time Series (2002–2012)

This article describes the principles used to generate global gap-free Leaf Area Index (LAI) time series from 2002–2012, based on MERIS (MEdium Resolution Imaging Spectrometer) full-resolution Level1B data. It is produced as a series of 10-day composites in geographic projection at 300-m spatial resolution. The processing chain comprises geometric correction, radiometric correction, pixel identification, LAI calculation with the BEAM (Basic ERS & Envisat (A)ATSR and MERIS Toolbox) MERIS vegetation processor, re-projection to a global grid and temporal aggregation selecting the measurement closest to the mean value. After the LAI pre-processing, we applied time series analysis to fill data gaps and to filter outliers using the technique of harmonic analysis (HA) in combination with mean annual and multiannual phenological data. Data gaps are caused by clouds, sensor limitations due to the solar zenith angle (<10°), topography and intermittent data reception. We applied our technique for the whole period of observation (July 2002–March 2012). Validation, carried out with VALERI (Validation of Land European Remote Sensing Instruments) and BigFoot data, revealed a high degree (R2 : 0.88) of agreement on a global scale.