Abstract. Direct measurements of OH and HO 2 over a tropical rainforest were made for the first time during the GABRIEL campaign in October 2005, deploying the custom-built HORUS instrument (HydrOxyl Radical measurement Unit based on fluorescence Spectroscopy), adapted to fly in a Learjet wingpod. Biogenic hydrocarbon emissions were expected to strongly reduce the OH and HO 2 mixing ratios as the air is transported from the ocean over the forest. However, surprisingly high mixing ratios of both OH and HO 2 were encountered in the boundary layer over the rainforest.The HORUS instrumentation and calibration methods are described in detail and the measurement results obtained are discussed. The extensive dataset collected during GABRIEL, including measurements of many other trace gases and photolysis frequencies, has been used to quantify the main sources and sinks of OH. Comparison of these measurementderived formation and loss rates of OH indicates strong previously overlooked recycling of OH in the boundary layer over the tropical rainforest, occurring in chorus with isoprene emission.

Abstract. We present measurements of NO, NO y , O 3 , and N 2 O within the lowermost stratosphere (LMS) over Europe obtained during the SPURT project. The measurements cover all seasons between November 2001 and July 2003. They span a broad band of latitudes from 30 • N to 75 • N and a potential temperature range from 290 to 380 K. The measurements represent a comprehensive data set of these tracers and reveal atmospheric transport processes that influence tracer distributions in the LMS. Median mixing ratios of stratospheric tracers in equivalent latitude-potential temperature coordinates show a clear seasonal cycle related to the Brewer-Dobson circulation, with highest values in spring and lowest values in autumn. Vertical tracer profiles show strong gradients at the extratropical tropopause, suggesting that vertical (cross-isentropic) mixing is reduced above the tropopause. Pronounced meridional gradients in the tracer mixing ratios are found on potential temperature surfaces in the LMS. This suggests strongly reduced mixing along isentropes. Concurrent large gradients in static stability in the vertical direction, and of PV in the meridional direction, suggest the presence of a mixing barrier. Seasonal cycles were found in the correlation slopes O 3 / N 2 O and NO y / N 2 O well above the tropopause. Absolute slope values are smallest in spring indicating chemically aged stratospheric air originating from high altitudes and latitudes. Larger values were measured in summer and autumn suggesting that a substantial fraction of air takes a "short-cut" from the tropical tropopause region into the extratropical LMS. The seasonal change in the composition of the LMS has direct implications for the ozone chemistry in this reCorrespondence to: M. I. Hegglin (michaela@atmosp.physics.utoronto.ca) gion. Comparisons of measured NO with the critical NO value at which net ozone production changes from negative to positive, imply ozone production up to 20 K above the local tropopause in spring, up to 30 K in summer, and up to 40 K in autumn. Above these heights, and in winter, net ozone production is negative.

As a major source region of the hydroxyl radical OH, the Tropics largely control the oxidation capacity of the atmosphere on a global scale. However, emissions of hydrocarbons from the tropical rainforest that react rapidly with OH can potentially deplete the amount of OH and thereby reduce the oxidation capacity. The airborne GABRIEL field campaign in equatorial South America (Suriname) in October 2005 investigated the influence of the tropical rainforest on the HO_x budget (HO_x = OH + HO₂). The first observations of OH and HO₂ over a tropical rainforest are compared to steady state concentrations calculated with the atmospheric chemistry box model MECCA. The important precursors and sinks for HO_x chemistry, measured during the campaign, are used as constraining parameters for the simulation of OH and HO₂. Significant underestimations of HO_x are found by the model over land during the afternoon, with mean ratios of observation to model of 12.2 &plusmn; 3.5 and 4.1 &plusmn; 1.4 for OH and HO₂, respectively. The discrepancy between measurements and simulation results is correlated to the abundance of isoprene. While for low isoprene mixing ratios (above ocean or at altitudes >3 km), observation and simulation agree fairly well, for mixing ratios >200 pptV (<3 km over the rainforest) the model tends to underestimate the HO_x observations as a function of isoprene. <br></br> Box model simulations have been performed with the condensed chemical mechanism of MECCA and with the detailed isoprene reaction scheme of MCM, resulting in similar results for HO_x concentrations. Simulations with constrained HO₂ concentrations show that the conversion from HO₂ to OH in the model is too low. However, by neglecting the isoprene chemistry in the model, observations and simulations agree much better. An OH source similar to the strength of the OH sink via isoprene chemistry is needed in the model to resolve the discrepancy. A possible explanation is that the oxidation of isoprene by OH not only dominates the removal of OH but also produces it in a similar amount. Several additional reactions which directly produce OH have been implemented into the box model, suggesting that upper limits in producing OH are still not able to reproduce the observations (improvement by factors of ≈2.4 and ≈2 for OH and HO₂, respectively). We determine that OH has to be recycled to 94% instead of the simulated 38% to match the observations, which is most likely to happen in the isoprene degradation process, otherwise additional sources are required

Abstract. Strongly enhanced tropospheric ozone (O3) mixing ratios have been reported in the Arabian Basin, a region with intense solar radiation and high concentrations of O3 precursors such as nitrogen oxides (NOx) and volatile organic compounds (VOCs). To analyze photochemical O3 production in the marine boundary layer (MBL) around the Arabian Peninsula, we use shipborne observations of NO, NO2, O3, OH, HO2, HCHO, the actinic flux, water vapor, pressure and temperature obtained during the summer 2017 Air Quality and Climate in the Arabian Basin (AQABA) campaign, and we compare them to simulation results from the ECHAM-MESSy Atmospheric Chemistry (EMAC) general circulation model. Net O3 production rates (NOPRs) were greatest over both the Gulf of Oman and the northern Red Sea (16 ppbv d−1) and over the Arabian Gulf (32 ppbv d−1). The NOPR over the Mediterranean, the southern Red Sea and the Arabian Sea did not significantly deviate from zero; however, the results for the Arabian Sea indicated weak net O3 production of 5 ppbv d−1 as well as net O3 destruction over the Mediterranean and the southern Red Sea with values of −1 and −4 ppbv d−1, respectively. Constrained by HCHO∕NO2 ratios, our photochemistry calculations show that net O3 production in the MBL around the Arabian Peninsula mostly occurs in NOx-limited regimes with a significant share of O3 production occurring in the transition regime between NOx limitation and VOC limitation over the Mediterranean and more significantly over the northern Red Sea and Oman Gulf.

Abstract. A new, two-channel instrument for simultaneous NO3 and N2O5 monitoring was used to make the first comprehensive set of nocturnal NOx measurements (NO, NO2, NO3 and N2O5) at the Taunus Observatory, a rural mountain site (Kleiner Feldberg) in South-western Germany. In May 2008, NO3 and N2O5 mixing ratios were well above the instrumental detection limit (a few ppt) on all nights of the campaign and were characterised by large variability. The concentrations of NO3, N2O5 and NO2 were consistent with the equilibrium constant, K2, defining the rates of formation and thermal dissociation of N2O5. A steady-state lifetime analysis is consistent with the loss of nocturnal NOx being dominated by the reaction of NO3 with volatile organic compounds in this forested region, with N2O5 uptake to aerosols of secondary importance. Analysis of a limited dataset obtained at high relative humidity indicated that the loss of N2O5 by reaction with water vapour is less efficient (>factor 3) than derived using laboratory kinetic data. The fraction of NOx present as NO3 and N2O5 reached ~20% on some nights, with night-time losses of NOx competing with daytime losses.