Adaptive optics control using model-based reinforcement learning.

Reinforcement learning (RL) presents a new approach for controlling adaptive optics (AO) systems for Astronomy. It promises to effectively cope with some aspects often hampering AO performance such as temporal delay or calibration errors. We formulate the AO control loop as a model-based RL problem (MBRL) and apply it in numerical simulations to a simple Shack-Hartmann Sensor (SHS) based AO system with 24 resolution elements across the aperture. The simulations show that MBRL controlled AO predicts the temporal evolution of turbulence and adjusts to mis-registration between deformable mirror and SHS which is a typical calibration issue in AO. The method learns continuously on timescales of some seconds and is therefore capable of automatically adjusting to changing conditions.

RETRACTED: Direct imaging discovery of a Jovian exoplanet within a triple-star system.

Direct imaging allows for the detection and characterization of exoplanets via their thermal emission. We report the discovery via imaging of a young Jovian planet in a triple-star system and characterize its atmospheric properties through near-infrared spectroscopy. The semimajor axis of the planet is closer relative to that of its hierarchical triple-star system than for any known exoplanet within a stellar binary or triple, making HD 131399 dynamically unlike any other known system. The location of HD 131399Ab on a wide orbit in a triple system demonstrates that massive planets may be found on long and possibly unstable orbits in multistar systems. HD 131399Ab is one of the lowest mass (4 ± 1 Jupiter masses) and coldest (850 ± 50 kelvin) exoplanets to have been directly imaged.

Fast-moving features in the debris disk around AU Microscopii.

In the 1980s, excess infrared emission was discovered around main-sequence stars; subsequent direct-imaging observations revealed orbiting disks of cold dust to be the source. These 'debris disks' were thought to be by-products of planet formation because they often exhibited morphological and brightness asymmetries that may result from gravitational perturbation by planets. This was proved to be true for the ß Pictoris system, in which the known planet generates an observable warp in the disk. The nearby, young, unusually active late-type star AU Microscopii hosts a well-studied edge-on debris disk; earlier observations in the visible and near-infrared found asymmetric localized structures in the form of intensity variations along the midplane of the disk beyond a distance of 20 astronomical units. Here we report high-contrast imaging that reveals a series of five large-scale features in the southeast side of the disk, at projected separations of 10-60 astronomical units, persisting over intervals of 1-4 years. All these features appear to move away from the star at projected speeds of 4-10 kilometres per second, suggesting highly eccentric or unbound trajectories if they are associated with physical entities. The origin, localization, morphology and rapid evolution of these features are difficult to reconcile with current theories.

PCDD/F formation in an iron/potassium-catalyzed diesel particle filter.

Catalytic diesel particle filters (DPFs) have evolved to a powerful environmental technology. Several metal-based, fuel soluble catalysts, so-called fuel-borne catalysts (FBCs), were developed to catalyze soot combustion and support filter regeneration. Mainly iron- and cerium-based FBCs have been commercialized for passenger cars and heavy-duty vehicle applications. We investigated a new iron/potassium-based FBC used in combination with an uncoated silicon carbide filter and report effects on emissions of polychlorinated dibenzodioxins/furans (PCDD/Fs). The PCDD/F formation potential was assessed under best and worst case conditions, as required for filter approval under the VERT protocol. TEQ-weighted PCDD/F emissions remained low when using the Fe/K catalyst (37/7.5 µg/g) with the filter and commercial, low-sulfur fuel. The addition of chlorine (10 µg/g) immediately led to an intense PCDD/F formation in the Fe/K-DPF. TEQ-based emissions increased 51-fold from engine-out levels of 95 to 4800 pg I-TEQ/L after the DPF. Emissions of 2,3,7,8-TCDD, the most toxic congener (TEF = 1.0), increased 320-fold, those of 2,3,7,8-TCDF (TEF = 0.1) even 540-fold. Remarkable pattern changes were noticed, indicating a preferential formation of tetrachlorinated dibenzofurans. It has been shown that potassium acts as a structural promoter inducing the formation of magnetite (Fe3O4) rather than hematite (Fe2O3). This may alter the catalytic properties of iron. But the chemical nature of this new catalyst is yet unknown, and we are far from an established mechanism for this new pathway to PCDD/Fs. In conclusion, the iron/potassium-catalyzed DPF has a high PCDD/F formation potential, similar to the ones of copper-catalyzed filters, the latter are prohibited by Swiss legislation.

Effects of a combined Diesel particle filter-DeNOx system (DPN) on reactive nitrogen compounds emissions: a parameter study.

The impact of a combined diesel particle filter-deNO(x) system (DPN) on emissions of reactive nitrogen compounds (RNCs) was studied varying the urea feed factor (α), temperature, and residence time, which are key parameters of the deNO(x) process. The DPN consisted of a platinum-coated cordierite filter and a vanadia-based deNO(x) catalyst supporting selective catalytic reduction (SCR) chemistry. Ammonia (NH3) is produced in situ from thermolysis of urea and hydrolysis of isocyanic acid (HNCO). HNCO and NH3 are both toxic and highly reactive intermediates. The deNO(x) system was only part-time active in the ISO8178/4 C1cycle. Urea injection was stopped and restarted twice. Mean NO and NO2 conversion efficiencies were 80%, 95%, 97% and 43%, 87%, 99%, respectively, for α = 0.8, 1.0, and 1.2. HNCO emissions increased from 0.028 g/h engine-out to 0.18, 0.25, and 0.26 g/h at α = 0.8, 1.0, and 1.2, whereas NH3 emissions increased from <0.045 to 0.12, 1.82, and 12.8 g/h with maxima at highest temperatures and shortest residence times. Most HNCO is released at intermediate residence times (0.2-0.3 s) and temperatures (300-400 °C). Total RNC efficiencies are highest at α = 1.0, when comparable amounts of reduced and oxidized compounds are released. The DPN represents the most advanced system studied so far under the VERT protocol achieving high conversion efficiencies for particles, NO, NO2, CO, and hydrocarbons. However, we observed a trade-off between deNO(x) efficiency and secondary emissions. Therefore, it is important to adopt such DPN technology to specific application conditions to take advantage of reduced NO(x) and particle emissions while avoiding NH3 and HNCO slip.

New exposure system to evaluate the toxicity of (scooter) exhaust emissions in lung cells in vitro.

A constantly growing number of scooters produce an increasing amount of potentially harmful emissions. Due to their engine technology, two-stroke scooters emit huge amounts of adverse substances, which can induce adverse pulmonary and cardiovascular health effects. The aim of this study was to develop a system to expose a characterized triple cell coculture model of the human epithelial airway barrier, to freshly produced and characterized total scooter exhaust emissions. In exposure chambers, cell cultures were exposed for 1 and 2 h to 1:100 diluted exhaust emissions and in the reference chamber to filtered ambient air, both controlled at 5% CO(2), 85% relative humidity, and 37 degrees C. The postexposure time was 0-24 h. Cytotoxicity, used to validate the exposure system, was significantly increased in exposed cell cultures after 8 h postexposure time. (Pro-) inflammatory chemo- and cytokine concentrations in the medium of exposed cells were significantly higher at the 12 h postexposure time point. It was shown that the described exposure system (with 2 h exposure duration, 8 and 24 h postexposure time, dilution of 1:100, flow of 2 L/min as optimal exposure conditions) can be used to evaluate the toxic potential of total exhaust emissions.

Impact of low- and high-oxidation diesel particulate filters on genotoxic exhaust constituents.

Diesel exhaust contains several genotoxic compounds that may or may not penetrate diesel particulate filters (DPFs). Furthermore, the DPF-supported combustion of soot and adsorbed compounds may lead to the formation of additional pollutants. Herein, we compare the impact of 14 different DPFs on emissions of known genotoxic compounds. During a four year period, these DPFs were tested on a heavy duty diesel engine, operated in the ISO 8178/4 C1 cycle. Integral samples, including gas-phase and particle-bound matter were taken. All DPFs were efficient wall-flow filters with solid particulate number filtration efficiencies eta > 98%. On the basis of their CO, NO, and NO(2) emission characteristics, two different filter families were distinguished. DPFs with high oxidation potential (hox, n = 8) converted CO and NO besides hydrocarbons, whereas low oxidation potential DPFs (lox, n = 6) did not support CO and NO oxidation but still converted hydrocarbons. Lox-DPFs reduced NO(2) from 1.0 +/- 0.3 (engine-out) to 0.42 +/- 0.11 g/kWh (eta = 0.59), whereas hox-DPFs induced a NO(2) formation up to 3.3 +/- 0.7 g/kWh (eta = -2.16). Emissions of genotoxic PAHs decreased for both filter families. Conversion efficiencies varied for individual PAHs and were lower for lox- (eta = 0.31-0.87) than for hox-DPFs (eta = 0.75-0.98). Certain nitro-PAHs were formed indicating that nitration is an important step along PAH oxidation. For example, 1-nitronaphthalene emissions increased from 11 to 17 to 21 microg/L without, with lox-, and hox-DPFs respectively, whereas 2-nitronaphthalene emissions decreased from 25 to 19 to 4.7 microg/L. In contrast to our expectations, the nitration potential of lox-DPFs was higher than the one of hox-DPFs, despite the intense NO(2) formation of the latter. The filters converted most genotoxic PAHs and nitro-PAHs and most soot particles, acting as carriers for these compounds. Hox-DPF exhaust remains oxidizing and therefore is expected to support atmospheric oxidation reactions, whereas lox-DPF exhaust is reducing and consuming oxidants such as ozone, when mixed with ambient air.

Secondary effects of catalytic diesel particulate filters: conversion of PAHs versus formation of nitro-PAHs.

Diesel particulate filters (DPFs) are a promising technology to detoxify diesel exhaust. However, the secondary combustion of diesel soot and associated compounds may also induce the formation of new pollutants. Diesel soot is rated as carcinogenic to humans and also acts as a carrier for a variety of genotoxic compounds such as certain polycyclic aromatic hydrocarbons (PAHs) or nitrated PAHs (nitro-PAHs). Furthermore, diesel exhaust contains considerable amounts of nitric oxide (NO), which can be converted to more powerful nitrating species like nitrogen dioxide (NO2), nitric acid (HNO3), and others. This mix of compounds may support nitration reactions in DPFs. Herein we report effects of two cordierite-based, monolithic, wall-flow DPFs on emissions of genotoxic PAHs and nitro-PAHs and compare these findings with those of a reporter gene bioassay sensitive to aryl hydrocarbons (AHs). Soot combustion was either catalyzed with an iron- or a copper/iron-based fuel additive (fuel-borne catalysts). A heavy duty diesel engine, operated according to the 8-stage ISO 8178/4 C1 cycle, was used as test platform. Emissions of all investigated 4- to 6-ring PAHs were reduced by about 40-90%, including those rated as carcinogenic. Emissions of 1- and 2-nitronaphthalene increased by about 20-100%. Among the 3-ring nitro-PAHs, emissions of 3-nitrophenanthrene decreased by about 30%, whereas 9-nitrophenanthrene and 9-nitroanthracene were found only after DPFs. In case of 4-ring nitro-PAHs, emissions of 3-nitrofluoranthene, 1-nitropyrene, and 4-nitropyrene decreased by about 40-60% with DPFs. Total AH-receptor (AHR) agonist concentrations of diesel exhaust were lowered by 80-90%, when using the iron- and copper-based DPFs. The tested PAHs accounted for < 1% of the total AHR-mediated response, indicating that considerable amounts of other aryl hydrocarbons must be present in filtered and unfiltered exhaust. We conclude that both DPFs detoxified diesel exhaust with respect to total aryl hydrocarbons, including the investigated carcinogenic PAHs, but we also noticed a secondary formation of selected nitro-PAHs. Nitration reactions were found to be stereoselective with a preferential substitution of hydrogen atoms at peri-positions. The stereoisomers obtained are related to combustion chemistry, but differ from those formed upon atmospheric nitration of PAHs.

Secondary effects of catalytic diesel particulate filters: copper-induced formation of PCDD/Fs.

Potential risks of a secondary formation of polychlorinated dibenzodioxins/furans (PCDD/Fs) were assessed for two cordierite-based, wall-through diesel particulate filters (DPFs) for which soot combustion was either catalyzed with an iron- or a copper-based fuel additive. A heavy duty diesel engine was used as test platform, applying the eight-stage ISO 8178/4 C1 cycle. DPF applications neither affected the engine performance, nor did they increase NO, NO2, CO, and CO2 emissions. The latter is a metric for fuel consumption. THC emissions decreased by about 40% when deploying DPFs. PCDD/F emissions, with a focus on tetra- to octachlorinated congeners, were compared under standard and worst case conditions (enhanced chlorine uptake). The iron-catalyzed DPF neither increased PCDD/F emissions, nor did it change the congener pattern, even when traces of chlorine became available. In case of copper, PCDD/F emissions increased by up to 3 orders of magnitude from 22 to 200 to 12 700 pg I-TEQ/L with fuels of < 2, 14, and 110 microg/g chlorine, respectively. Mainly lower chlorinated DD/Fs were formed. Based on these substantial effects on PCDD/F emissions, the copper-catalyzed DPF system was not approved for workplace applications, whereas the iron system fulfilled all the specifications of the Swiss procedures for DPF approval (VERT).

Fast calibration of high-order adaptive optics systems.

We present a new method of calibrating adaptive optics systems that greatly reduces the required calibration time or, equivalently, improves the signal-to-noise ratio. The method uses an optimized actuation scheme with Hadamard patterns and does not scale with the number of actuators for a given noise level in the wavefront sensor channels. It is therefore highly desirable for high-order systems and/or adaptive secondary systems on a telescope without a Gregorian focal plane. In the latter case, the measurement noise is increased by the effects of the turbulent atmosphere when one is calibrating on a natural guide star.