Development of a GPU-based high-performance radiative transfer model for the Infrared Atmospheric Sounding Interferometer (IASI)

Satellite-observed radiance is a nonlinear functional of surface properties and atmospheric temperature and absorbing gas profiles as described by the radiative transfer equation (RTE). In the era of hyperspectral sounders with thousands of high-resolution channels, the computation of the radiative transfer model becomes more time-consuming. The radiative transfer model performance in operational numerical weather prediction systems still limits the number of channels we can use in hyperspectral sounders to only a few hundreds. To take the full advantage of such high-resolution infrared observations, a computationally efficient radiative transfer model is needed to facilitate satellite data assimilation. In recent years the programmable commodity graphics processing unit (GPU) has evolved into a highly parallel, multi-threaded, many-core processor with tremendous computational speed and very high memory bandwidth. The radiative transfer model is very suitable for the GPU implementation to take advantage of the hardware’s efficiency and parallelism where radiances of many channels can be calculated in parallel in GPUs.In this paper, we develop a GPU-based high-performance radiative transfer model for the Infrared Atmospheric Sounding Interferometer (IASI) launched in 2006 onboard the first European meteorological polar-orbiting satellites, METOP-A. Each IASI spectrum has 8461 spectral channels. The IASI radiative transfer model consists of three modules. The first module for computing the regression predictors takes less than 0.004% of CPU time, while the second module for transmittance computation and the third module for radiance computation take approximately 92.5% and 7.5%, respectively. Our GPU-based IASI radiative transfer model is developed to run on a low-cost personal supercomputer with four GPUs with total 960 compute cores, delivering near 4 TFlops theoretical peak performance. By massively parallelizing the second and third modules, we reached 364× speedup for 1 GPU and 1455× speedup for all 4 GPUs, both with respect to the original CPU-based single-threaded Fortran code with the –O₂ compiling optimization. The significant 1455× speedup using a computer with four GPUs means that the proposed GPU-based high-performance forward model is able to compute one day’s amount of 1,296,000 IASI spectra within nearly 10 min, whereas the original single CPU-based version will impractically take more than 10 days. This model runs over 80% of the theoretical memory bandwidth with asynchronous data transfer. A novel CPU–GPU pipeline implementation of the IASI radiative transfer model is proposed. The GPU-based high-performance IASI radiative transfer model is suitable for the assimilation of the IASI radiance observations into the operational numerical weather forecast model.     

GPU-based single-cluster algorithm for the simulation of the Ising model

We present the GPU calculation with the common unified device architecture (CUDA) for the Wolff single-cluster algorithm of the Ising model. Proposing an algorithm for a quasi-block synchronization, we realize the Wolff single-cluster Monte Carlo simulation with CUDA. We perform parallel computations for the newly added spins in the growing cluster. As a result, the GPU calculation speed for the two-dimensional Ising model at the critical temperature with the linear size L = 4096 is 5.60 times as fast as the calculation speed on a current CPU core. For the three-dimensional Ising model with the linear size L = 256, the GPU calculation speed is 7.90 times as fast as the CPU calculation speed. The idea of quasi-block synchronization can be used not only in the cluster algorithm but also in many fields where the synchronization of all threads is required.     

Efficient high voltage pulser for piezoelectric air coupled transducer

The design of high voltage pulser for air coupled ultrasound imaging is presented. It is dedicated for air-coupled ultrasound applications when piezoelectric transducer design is used. Two identical N-channel MOSFETs are used together with 1200 V high and low side driver IC. Simple driving pulses’ delay and skew circuit is used to reduce the cross-conduction. Analysis of switch peak current and channel resistance relation to maximum operation frequency and load capacitance is given. PSPICE simulation was used to analyze the gate driver resistance, gate pulse skew, pulse amplitude influence on energy consumption when loaded by capacitive load. Experimental investigation was verified against simulation and theoretical predictions. For 500 pF capacitance, which is most common for piezoelectric air coupled transducers, pulser consumes 650 μJ at 1 kV pulse and 4 μJ at 50 V. Pulser is capable to produce up to 1 MHz pulse trains with positive 50 V–1 kV pulses with up to 10 A peak output current. When loaded by 200 kHz transducer at 1 kV pulse amplitude rise time is 40 ns and fall time is 32 ns which fully satisfies desired 1 MHz bandwidth.     

Simulating the kinematics of completely faceted surfaces

We fully generalize a previously-developed computational geometry tool [1] to perform large-scale simulations of arbitrary 2 + 1D faceted surfaces z = h(x, y). The method is an explicit front-tracking scheme that uses a compact, three-component facet/edge/junction storage mode. Because it naturally mirrors the intrinsic surface structure, this scheme allows both rapid simulation of large ensembles, and easy extraction of geometrical statistics. To do so, it must overcome the barrier of detecting and resolving a wide variety of topological changes that occur during surface evolution. However, while the variety of topological events is larger than in the case of 2D cellular networks, it is still limited, and our main result is a comprehensive algorithm performing these changes in the code.     

Diode-pumped Nd:YAG ceramic laser at 946 nm passively Q-switched with a Cr4+:YAG saturable absorber

We demonstrate a diode-pumped Nd:YAG ceramic laser with emission at 946 nm that is passively Q-switched by single-crystal Cr⁴⁺:YAG saturable absorber. An average output power of 1.7 W is measured under 18.4 W of incident power using an output mirror with transmission T=4%. The corresponding optical-to-optical efficiency is 9.2%. The laser runs at a pulse repetition rate of 120 kHz and delivers pulses with energy of 14 μJ and duration of 80 ns, which corresponds to a peak power of 175 W.     

Influence of the ZnO nanoparticle sizes and morphology on the photoinduced light reflectivity

We have established a principal possibility of changes of the light reflectivity at the wavelength of 633 nm (He–Ne laser) under influence of the external laser light. The changes are very sensitive to the wavelength of the photoinduced laser. We have chosen two types of the photoinduced lasers: UV nitrogen 7 ns laser at wavelength 371 nm heating near the absorption edge and the 10 ns 1064 nm Nd:YAG laser with wavelength 1064 nm. The power dependences of the reflectivity were studied. Possible explanation of the observed effects is presented following the conception of the nano-trapping levels. These results have been obtained from two ZnO thin films prepared from principally different deposition parameters leading to different particle features and morphologies.     

Soft-X-ray lasing in nickel-like barium at 9.2 nm using the grazing-incidence scheme

We report on the experimental demonstration of saturated X-ray lasing on the 4 d → 4p, J = 0–1 line of nickel-like barium (Ba, Z = 56) at a wavelength of 9.2 nm, using a main pulse energy of 9 J in a 1.5-ps duration pulse from a Nd:glass chirped-pulse amplification laser. Gain saturation was achieved by applying a triple-pulse scheme in which a weak (few-percent) prepulse, preceding the main pulse by 2.4 ns, is followed by a second, relatively intense (16%) prepulse ~ 50 ps before the main pulse. For handling convenience, compound targets of BaF₂ were used, either in the form of windows or coated onto glass slides.     

Spatial alignment evolution of self-assembled In0.4Ga0.6As island arrays grown on GaAs (3 1 1)B surface by atomic hydrogen-assisted molecular beam epitaxy

Self-assembled In_0.4Ga_0.6As island arrays have been grown on (3 1 1)B GaAs substrates by using atomic hydrogen-assisted molecular beam epitaxy (H-MBE). The evolution process of surface morphology with deposition has been analyzed by atomic force microscopy (AFM) and the development of lateral ordering has been highlighted by two-dimensional fast Fourier transformation (2DFFT) analysis of the AFM images. It is revealed that the InGaAs islands are arranged in nearly perfect two-dimensional (2D) square-like lattice with two sides parallel to [0 1 −1] and [−2 3 3] azimuths. Such an alignment of islands is coincident with the anisotropy of bulk elastic modulus of the GaAs (3 1 1)B substrate.     

Radio frequency source power-induced ion energy impact on SiN films deposited by using a pulsed-PECVD in SiH4–N2 plasma at room temperature

Using a radio frequency (rf) pulsed-plasma enhanced chemical vapor deposition system, silicon nitride (SiN) films were deposited in a SiH₄–N₂ inductively coupled plasma. Effect of duty ratio and rf source powers on deposition rate at room temperature were investigated in the ranges 50–90% and 600–900 W, respectively. Plasma diagnostics on ion energy was conducted and rf source power-induced ion energy impact on SiN films were studied as well as some correlations between deposition rate and ion energy. High and low energies ranged from 17.8 to 22.6 eV, and from 23.6 to 33.8 eV, respectively. Higher ion energies observed at lower duty ratios or lower rf powers was attributed to a lower plasma density. Ion energy flux variation was opposite to that for ion energy. Meanwhile, the deposition rate increased with decreasing the duty ratio at all powers but 900 W. This was not clear as a function of rf source power. The deposition rate ranged from 17.0 to 26.5 nm/min.     

Electrical characterization of dielectric barrier discharge driven by repetitive nanosecond pulses in atmospheric air

Dielectric barrier discharge (DBD) is an important method to produce non-thermal plasma, which has been widely used in many fields. In the paper, a repetitive nanosecond-pulse generator is used for the excitation of DBD. Output positive pulse of the generator has a rise time of about 15 ns and a full width at half maximum of 30–40 ns, and pulse repetition frequency varies from single shot to 2 kHz. The purpose of this paper is to experiment the electrical characteristics of DBD driven by repetitive nanosecond pulses. The variables affecting discharge conditions, including air gap spacing, dielectric thickness, barrier fashion, and applied pulse repetition frequency, are investigated. The relationship between electric field, discharge current, instantaneous discharge power across air gap, and estimated electron density with the length of air gap, dielectric thickness, barrier fashion, and pulse repetition frequency is obtained respectively, and the experimental results are also discussed. In addition, two typical images exhibiting homogeneous and filamentary discharges are given with different experimental conditions.     

Iterative diagonalization in augmented plane wave based methods in electronic structure calculations

Due to the increased computer power and advanced algorithms, quantum mechanical calculations based on Density Functional Theory are more and more widely used to solve real materials science problems. In this context large nonlinear generalized eigenvalue problems must be solved repeatedly to calculate the electronic ground state of a solid or molecule. Due to the nonlinear nature of this problem, an iterative solution of the eigenvalue problem can be more efficient provided it does not disturb the convergence of the self-consistent-field problem. The blocked Davidson method is one of the widely used and efficient schemes for that purpose, but its performance depends critically on the preconditioning, i.e. the procedure to improve the search space for an accurate solution. For more diagonally dominated problems, which appear typically for plane wave based pseudopotential calculations, the inverse of the diagonal of (H ? ES) is used. However, for the more efficient “augmented plane wave + local-orbitals” basis set this preconditioning is not sufficient due to large off-diagonal terms caused by the local orbitals. We propose a new preconditioner based on the inverse of (H ? λS) and demonstrate its efficiency for real applications using both, a sequential and a parallel implementation of this algorithm into our WIEN2k code.     

Aqueous polymer–nitrate solution deposition of YBCO films

High critical current density YBa₂Cu₃O_7?x (YBCO) films were prepared by solution deposition of aqueous non-fluorine precursors. Non-fluorine polymer-assisted deposition (PAD) processes utilizing rheology modifiers and chelating agents were used to produce 50 nm films with a critical current density (J_c) over 3 MA/cm² and 400 nm films with J_c > 1 MA/cm²_. T_c measurements indicated that films have T_c values near 90 K. The total heat treatment time to produce these high performance films was less than 4 h. Rheology modifiers such as polyvinyl alcohol (PVA) and hydroxyethyl cellulose (HEC) were used to increase the thickness of deposited films independent of the solution cation concentration. Chelating agents such as polyethylene glycol (PEG) and sucrose increased the barium ion solubility. Nitrate crystallization during deposition was controlled through rapid drying with vacuum and coating with hot solutions.     

Laser ablation fast pulse discharge plasma spectroscopy analysis of Pb,Mg and Sn in soil

Laser ablation fast pulse discharge plasma spectroscopy (LA-FPDPS) technique is a recently developed atomic emission analytical technique that is analogous to dual pulse laser induced breakdown spectroscopy (DP-LIBS). LA-FPDPS, however, uses a periodical oscillating discharge plasma generation method on samples instead of the second laser beam in DP-LIBS. Here we describe the electric characteristics and its application to the analysis of Pb, Mg and Sn in soil. Due to the fast discharge process, the peak power deposition rate is up to 1.5 MW, although the discharge energy is relatively small. The main energy deposition process only last for ~ 4 μs. From the measured spectra, calibration curves for Pb, Mg and Sn in soil were derived and the limits of detection were 1.5 μg/g, 34 μg/g and 0.16 μg/g respectively.     

Growth of the Au-induced c(8 × 2) structure on vicinal Ge(001)

We have studied the formation of Ge(001) c(8 × 2)–Au surfaces on vicinal samples by scanning tunneling microscopy. The vicinal samples are tilted 1° toward [110]. The c(8 × 2)–Au surface is prepared by depositing 0.75 ± 0.05 ML of Au onto a germanium surface held at 800 K. The anisotropy introduced by the atomic steps of the vicinal surface and the preferential etching of S_B steps during Au deposition is sufficient to introduce a preferred growth direction for the c(8 × 2)–Au phase. The result is a sample on which 78% of the surface is populated by Au-induced chains oriented parallel to the step direction. These parallel Ge(001) c(8 × 2)–Au domains are separated by single or multiple height D_A steps (0.28 nm high).     

Tissue gadolinium deposition in renally impaired rats exposed to different gadolinium-based MRI contrast agents: Evaluation with inductively coupled plasma mass spectrometry (ICP-MS)

ObjectivesTo quantify tissue gadolinium (Gd) deposition in renally impaired rats exposed to Gd-EOB-DTPA and other Gd-based MRI contrast agents by means of inductively coupled plasma mass spectrometry (ICP-MS), and to compare the differences in distribution among major organs as possible triggers for nephrogenic systemic fibrosis (NSF).MethodsA total of 15 renally impaired rats were injected with Gd-EOB-DTPA, Gd-DTPA-BMA and Gd-HP-DO3A. Gd contents of skin, liver, kidney, lung, heart, spleen, diaphragm and femoral muscle were measured by inductively coupled plasma mass spectrometry (ICP-MS). Histological assessment was also conducted.ResultsTissue Gd deposition in all organs was significantly higher (P = 0.005 ~ 0.009) in the Gd-DTPA-BMA group than in the Gd-HP-DO3A and Gd-EOB-DTPA groups. In the Gd-DTPA-BMA group, Gd was predominantly deposited in kidney (1306 ± 605.7 μg/g), followed by skin, liver, lung, spleen, femoral muscle, diaphragm and heart. Comparing Gd-HP-DO3A and Gd-EOB-DTPA groups, Gd depositions in the kidney, liver and lung were significantly lower (P = 0.009 ~ 0.011) in the Gd-EOB-DTPA group than in the Gd-HP-DO3A group although no significant differences were seen for any other organs.ConclusionsGd-EOB-DTPA is a stable and safe Gd-based contrast agent (GBCA) showing lower Gd deposition in major organs in renally impaired rats, compared with other GBCAs. This fact suggests that the risk of NSF onset would be low in the use of Gd-EOB-DTPA.     

Electronic properties of the polycrystalline tin dioxide interface with conjugated organic layers

The results on the electronic structure of the unoccupied electronic states of the polycrystalline SnO₂ in the energy range from 5 eV to 25 eV above the Fermi level are presented. The modification of the electronic structure and of the surface potential upon deposition of the ultrathin films of copper phthalocyanine (CuPc) and of perylene tetracarboxylic acid dianhydride (PTCDA) film onto the SnO₂ surface were studied using the very low energy electron diffraction (VLEED) method and the total current spectroscopy (TCS) measurement scheme. A substantial attenuation of the TCS signal coming from the SnO₂ surface was observed upon formation of a 1.5–2 nm thick organic deposit layer while no new spectral features from the deposit were distinguishable. It was observed that the electronic structure typical for the organic films was formed within the organic deposit thickness range from 2 nm to 7 nm. The interfacial charge transfer was characterized by the formation of the polarization layer up to 5 nm thick in the organic films. The PTCDA deposition on SnO₂ was accompanied by the negative charge transfer onto the organic layer and to the 0.65 eV increase the surface work function. At the CuPc/SnO₂ interface, the negative charge was transferred to the SnO₂ surface and the overall surface work function decreased by 0.15 eV.     

Nanosecond-pulsed erbium-doped fiber lasers with graphene saturable absorber

We demonstrate graphene mode-locked nanosecond erbium-doped fiber laser in an all-fiber ring cavity. The clean and robust pulse train was generated at 27 mW pump power. Resultant central wavelength, repetition rate and pulse width was 1560 nm, 388 kHz and 6 ns, respectively. With two stage fiber amplifier, the output power was 553 mW, corresponding to single pulse energy of 1.4 μJ. In addition, the pulse-width can be varied ranging from 3 ns to 20 ns at repetition rate between 200 kHz and 1.54 MHz by changing the length of the laser cavity.     

Effect of oxide on carbon deposition behavior of CH4 fuel on Ni/ScSZ cermet anode in high temperature SOFCs

This work investigated the effect of oxide in Ni-zirconia cermets on the carbon deposition behavior in internal reforming SOFCs. Within 800–1000 °C, carbon deposition was found to decrease with increasing temperature on Ni/ScSZ cermet anodes at a low oxygen / carbon ratio (O / C = 0.03) during anodic oxidation of methane. On the contrary, an opposite trend was observed on Ni/YSZ under the same conditions, consisting with the temperature dependence of carbon deposition predicted by a thermodynamic equilibrium calculation. Results of temperature-programmed-reduction (TPR) of NiO mixed with YSZ or ScSZ indicated that interaction of Ni with ScSZ is stronger than that with YSZ. The stronger interaction was corroborated by observed tendency of inhibiting Ni agglomeration by both BET specific surface area analysis and SEM observation. It was also found that the dependence of CO₂ production rate monitored by GC on current density showed a similar dependence trend of the equilibrium CO₂ content on O / C ratio. A model in which H₂O_ad enrichment effects on Ni surface by anodic current depend on the interaction between Ni and the oxide in Ni cermet was proposed to explain the different carbon deposition behaviors between Ni/YSZ and Ni/ScSZ cermets.     

Self-limited growth of the CaF nanowire on the Si(5 5 12)-2 × 1 template

The atomic structure and interfacial bonding of the ordered-and-isolated CaF nanowires on Si(5 5 12)-2 × 1 have been disclosed by scanning tunneling microscopy and synchrotron photoemission spectroscopy. Initially, CaF molecules dissociated from thermally deposited CaF₂ molecules are adsorbed preferentially on the chain structures of Si(5 5 12)-2 × 1 held at 500 °C. With increasing CaF₂ deposition amount, one-dimensional (1D) CaF nanowires composed of (113) and (111) facets are formed. The line density of these CaF nanowires increases as a function of deposition amount. Finally, at a submonolayer coverage, the surface is saturated with these 1D nanowires except for the (225) subunit, while the original period of Si(5 5 12)-2 × 1, 5.35 nm, is preserved. It has been deduced by the present studies that, owing to these preferential adsorption of CaF and facet-dependent growth of a CaF layer within a unit periodic length of Si(5 5 12)-2 × 1, such a self-limited growth of the CaF nanowire with a high aspect ratio becomes possible.