scholarly journals Speeding Up the Computation of WRF Double-Moment 6-Class Microphysics Scheme with GPU

2013 ◽  
Vol 30 (12) ◽  
pp. 2896-2906 ◽  
Author(s):  
J. Mielikainen ◽  
B. Huang ◽  
H.-L. A. Huang ◽  
M. D. Goldberg ◽  
A. Mehta
Keyword(s):  
General Circulation ◽  
Graphics Processing Unit ◽  
Cloud Condensation Nuclei ◽  
General Circulation Models ◽  
Processing Unit ◽  
Cloud Droplets ◽  
Microphysics Scheme ◽  
Mixing Ratios ◽  
Computational Performance ◽  
Double Moment

Abstract The Weather Research and Forecasting model (WRF) double-moment 6-class microphysics scheme (WDM6) implements a double-moment bulk microphysical parameterization of clouds and precipitation and is applicable in mesoscale and general circulation models. WDM6 extends the WRF single-moment 6-class microphysics scheme (WSM6) by incorporating the number concentrations for cloud and rainwater along with a prognostic variable of cloud condensation nuclei (CCN) number concentration. Moreover, it predicts the mixing ratios of six water species (water vapor, cloud droplets, cloud ice, snow, rain, and graupel), similar to WSM6. This paper describes improving the computational performance of WDM6 by exploiting its inherent fine-grained parallelism using the NVIDIA graphics processing unit (GPU). Compared to the single-threaded CPU, a single GPU implementation of WDM6 obtains a speedup of 150× with the input/output (I/O) transfer and 206× without the I/O transfer. Using four GPUs, the speedup reaches 347× and 715×, respectively.

Development of an Effective Double-Moment Cloud Microphysics Scheme with Prognostic Cloud Condensation Nuclei (CCN) for Weather and Climate Models

2010 ◽  
Vol 138 (5) ◽  
pp. 1587-1612 ◽  
Author(s):  
Kyo-Sun Sunny Lim ◽  
Song-You Hong
Keyword(s):  
Climate Models ◽  
Cloud Condensation Nuclei ◽  
Computational Cost ◽  
Cloud Microphysics ◽  
Test Bed ◽  
Condensation Nuclei ◽  
Microphysics Scheme ◽  
Mixing Ratios ◽  
Stratiform Region ◽  
Double Moment

Abstract A new double-moment bulk cloud microphysics scheme, the Weather Research and Forecasting (WRF) Double-Moment 6-class (WDM6) Microphysics scheme, which is based on the WRF Single-Moment 6-class (WSM6) Microphysics scheme, has been developed. In addition to the prediction for the mixing ratios of six water species (water vapor, cloud droplets, cloud ice, snow, rain, and graupel) in the WSM6 scheme, the number concentrations for cloud and rainwater are also predicted in the WDM6 scheme, together with a prognostic variable of cloud condensation nuclei (CCN) number concentration. The new scheme was evaluated on an idealized 2D thunderstorm test bed. Compared to the simulations from the WSM6 scheme, there are greater differences in the droplet concentration between the convective core and stratiform region in WDM6. The reduction of light precipitation and the increase of moderate precipitation accompanying a marked radar bright band near the freezing level from the WDM6 simulation tend to alleviate existing systematic biases in the case of the WSM6 scheme. The strength of this new microphysics scheme is its ability to allow flexibility in variable raindrop size distribution by predicting the number concentrations of clouds and rain, coupled with the explicit CCN distribution, at a reasonable computational cost.

scholarly journals On the Efficiency of OpenACC-aided GPU-Based FDTD Approach: Application to Lightning Electromagnetic Fields

2020 ◽  
Vol 10 (7) ◽  
pp. 2359
Author(s):  
Sajad Mohammadi ◽  
Hamidreza Karami ◽  
Mohammad Azadifar ◽  
Farhad Rachidi
Keyword(s):  
Electromagnetic Fields ◽  
Large Scale ◽  
Electromagnetic Compatibility ◽  
Programming Model ◽  
Graphics Processing Unit ◽  
Fdtd Method ◽  
Computation Time ◽  
Critical Factor ◽  
Processing Unit ◽  
Computational Performance

An open accelerator (OpenACC)-aided graphics processing unit (GPU)-based finite difference time domain (FDTD) method is presented for the first time for the 3D evaluation of lightning radiated electromagnetic fields along a complex terrain with arbitrary topography. The OpenACC directive-based programming model is used to enhance the computational performance, and the results are compared with those obtained by using a CPU-based model. It is shown that OpenACC GPUs can provide very accurate results, and they are more than 20 times faster than CPUs. The presented results support the use of OpenACC not only in relation to lightning electromagnetics problems, but also to large-scale realistic electromagnetic compatibility (EMC) applications in which computation time efficiency is a critical factor.

scholarly journals Snow-induced buffering in aerosol–cloud interactions

2020 ◽  
Author(s):  
Takuro Michibata ◽  
Kentaroh Suzuki ◽  
Toshihiko Takemura
Keyword(s):  
Radiative Forcing ◽  
General Circulation ◽  
State Of The Art ◽  
General Circulation Models ◽  
Regional Pattern ◽  
Cloud Droplets ◽  
Microphysical Process ◽  
Aerosol Cloud ◽  
Effective Radiative Forcing ◽  
Aerosol Cloud Interactions

Abstract. Complex aerosol–cloud–precipitation interactions lead to large differences in estimates of aerosol impacts on climate among general circulation models (GCMs) and satellite retrievals. Typically, precipitating hydrometeors are treated diagnostically in most GCMs, and their radiative effects are ignored. Here, we quantify how the treatment of precipitation influences the simulated effective radiative forcing due to aerosol–cloud interactions (ERFaci) using a state-of-the-art GCM with a two-moment prognostic precipitation scheme that incorporates the radiative effect of precipitating particles, and investigate how microphysical process representations are related to macroscopic climate effects. Prognostic precipitation substantially weakens the magnitude of ERFaci (by approximately 75 %) compared with the traditional diagnostic scheme, and this is the result of the increased longwave (warming) and weakened shortwave (cooling) components of ERFaci. The former is attributed to additional adjustment processes induced by falling snow, and the latter stems largely from riming of snow by collection of cloud droplets. The significant reduction in ERFaci does not occur without prognostic snow, which contributes mainly by buffering the cloud response to aerosol perturbations through depleting cloud water via collection. Prognostic precipitation also alters the regional pattern of ERFaci, particularly over northern mid-latitudes where snow is abundant. The treatment of precipitation is thus a highly influential controlling factor of ERFaci, contributing more than other uncertain tunable processes related to aerosol–cloud–precipitation interactions. This change in ERFaci caused by the treatment of precipitation is large enough to explain the existing difference in ERFaci between GCMs and observations.

scholarly journals Impact of mixing state and hygroscopicity on CCN activity of biomass burning aerosol in Amazonia

2016 ◽  
Author(s):  
Madeleine Sánchez Gácita ◽  
Karla M. Longo ◽  
Julliana L. M. Freire ◽  
Saulo R. Freitas ◽  
Scot T. Martin
Keyword(s):  
Biomass Burning ◽  
General Circulation ◽  
Cloud Model ◽  
Cloud Condensation Nuclei ◽  
General Circulation Models ◽  
Mixing State ◽  
Activation Kinetics ◽  
Cloud Properties ◽  
Aerosol Mass ◽  
Biomass Burning Aerosol

Abstract. Smoke aerosols prevail throughout Amazonia because of widespread biomass burning during the dry season. External mixing, low variability in the particle size distribution and low particle hygroscopicity are typical. There can be profound effects on cloud properties. This study uses an adiabatic cloud model to simulate the activation of smoke particles as cloud condensation nuclei (CCN) and to assess the relative importance of variability in hygroscopicity, mixing state, and activation kinetics for the activated fraction and maximum supersaturation. The analysis shows that use of medium values of hygroscopicity representative of smoke aerosols for other biomass burning regions on Earth can lead to significant errors, compared to the use of low hygroscopicity reported for Amazonia. Kinetic limitations, which can be significant for medium and high hygroscopicity, did not play a strong role for CCN activation of particles representative of Amazonia smoke aerosols, even when taking into account the large aerosol mass and number concentrations typical of the region. Internal compared to external mixing of particle components of variable hygroscopicity resulted in a significant overestimation of the activated fraction. These findings on uncertainties and sensitivities provide guidance on appropriate simplifications that can be used for modeling of smoke aerosols within general circulation models.

scholarly journals The impact of the 1783–1784 AD Laki eruption on global aerosol formation processes and cloud condensation nuclei

2010 ◽  
Vol 10 (2) ◽  
pp. 3189-3228
Author(s):  
A. Schmidt ◽  
K. S. Carslaw ◽  
G. W. Mann ◽  
B. M. Wilson ◽  
T. J. Breider ◽  
...  
Keyword(s):  
Boundary Layer ◽  
Northern Hemisphere ◽  
General Circulation ◽  
Cloud Condensation Nuclei ◽  
General Circulation Models ◽  
Aerosol Formation ◽  
Condensation Nuclei ◽  
Upper Troposphere ◽  
Microphysical Processes ◽  
The Impact

Abstract. The 1783–1784 AD Laki flood lava eruption commenced on 8 June 1783 and released 122 Tg of sulphur dioxide gas over the course of 8 months into the upper troposphere and lower stratosphere above Iceland. Previous studies have examined the impact of the Laki eruption on sulphate aerosol and climate using general circulation models. Here, we study the impact on aerosol microphysical processes, including the nucleation of new particles and their growth to cloud condensation nuclei (CCN) using a comprehensive Global Model of Aerosol Processes (GLOMAP). Total particle concentrations in the free troposphere increase by a factor ~16 over large parts of the Northern Hemisphere in the 3 months following the onset of the eruption. Particle concentrations in the boundary layer increase by a factor 2 to 5 in regions as far away as North America, the Middle East and Asia due to long-range transport of nucleated particles. CCN concentrations (at 0.22% supersaturation) increase by a factor 65 in the upper troposphere with maximum changes in 3-month zonal mean concentrations of ~1400 cm−3 at high northern latitudes. 3-month zonal mean CCN concentrations in the boundary layer at the latitude of the eruption increase by up to a factor 26, and averaged over the Northern Hemisphere, the eruption caused a factor 4 increase in CCN concentrations at low-level cloud altitude. The simulations show that the Laki eruption would have completely dominated as a source of CCN in the pre-industrial atmosphere. The model also suggests an impact of the eruption in the Southern Hemisphere, where CCN concentrations are increased by up to a factor 1.4 at 20° S. Our model simulations suggest that the impact of an equivalent wintertime eruption on upper tropospheric CCN concentrations is only about one-third of that of a summertime eruption. The simulations show that the microphysical processes leading to the growth of particles to CCN sizes are fundamentally different after an eruption when compared to the unperturbed atmosphere, underlining the importance of using a fully coupled microphysics model when studying long-lasting, high-latitude eruptions.

scholarly journals GPU acceleration of a model-based iterative method for Digital Breast Tomosynthesis

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
R. Cavicchioli ◽  
J. Cheng Hu ◽  
E. Loli Piccolomini ◽  
E. Morotti ◽  
L. Zanni
Keyword(s):  
Graphics Processing Unit ◽  
Digital Breast Tomosynthesis ◽  
Projection Algorithm ◽  
Processing Unit ◽  
Data Set ◽  
Breast Tomosynthesis ◽  
Model Based ◽  
Computational Performance ◽  
Gradient Based ◽  
Time Required

AbstractDigital Breast Tomosynthesis (DBT) is a modern 3D Computed Tomography X-ray technique for the early detection of breast tumors, which is receiving growing interest in the medical and scientific community. Since DBT performs incomplete sampling of data, the image reconstruction approaches based on iterative methods are preferable to the classical analytic techniques, such as the Filtered Back Projection algorithm, providing fewer artifacts. In this work, we consider a Model-Based Iterative Reconstruction (MBIR) method well suited to describe the DBT data acquisition process and to include prior information on the reconstructed image. We propose a gradient-based solver named Scaled Gradient Projection (SGP) for the solution of the constrained optimization problem arising in the considered MBIR method. Even if the SGP algorithm exhibits fast convergence, the time required on a serial computer for the reconstruction of a real DBT data set is too long for the clinical needs. In this paper we propose a parallel SGP version designed to perform the most expensive computations of each iteration on Graphics Processing Unit (GPU). We apply the proposed parallel approach on three different GPU boards, with computational performance comparable with that of the boards usually installed in commercial DBT systems. The numerical results show that the proposed GPU-based MBIR method provides accurate reconstructions in a time suitable for clinical trials.

An Efficient Algorithm of Wiener–Hopf Method With Graphics Processing Unit for Duct Acoustics

2017 ◽  
Vol 139 (5) ◽  
Author(s):  
Hanbo Jiang ◽  
Alex Siu Hong Lau ◽  
Xun Huang
Keyword(s):  
Graphics Processing Unit ◽  
Aircraft Engine ◽  
Low Noise ◽  
Superior Performance ◽  
Processing Unit ◽  
Duct Acoustics ◽  
Design And Optimization ◽  
Computational Performance ◽  
Hopf Method ◽  
Graphics Processing

Acoustic liner optimization calls for very efficient simulation methods. A highly efficient and straightforward algorithm is proposed here for the Wiener–Hopf solver, which also takes advantage of the parallel processing capability of the emerging graphics processing unit (GPU) technology. The proposed algorithm adopts a simple concept that re-arranges the formulations of the Wiener–Hopf solver to appropriate matrix forms. This concept was often overlooked but is surprisingly succinct, which leads to a stunningly efficient computational performance. By examining the computational performance of two representative setups (lined duct and duct radiations), the current study shows the superior performance of the proposed algorithm, particularly with GPU. The much improved computational efficiency further suggests the potential of the proposed algorithm and the use of GPU for practical low-noise aircraft engine design and optimization.

scholarly journals Droplet Activation and Mixing in Large-Eddy Simulation of a Shallow Cumulus Field

2012 ◽  
Vol 69 (2) ◽  
pp. 444-462 ◽  
Author(s):  
Joanna Slawinska ◽  
Wojciech W. Grabowski ◽  
Hanna Pawlowska ◽  
Hugh Morrison
Keyword(s):  
Cloud Condensation Nuclei ◽  
Cloud Base ◽  
Subgrid Scale ◽  
Cloud Droplets ◽  
Microphysics Scheme ◽  
Free Air ◽  
Aircraft Observations ◽  
The Mean ◽  
Cloud Activation ◽  
The Impact

Abstract This paper presents the application of a double-moment bulk warm-rain microphysics scheme to the simulation of a field of shallow convective clouds based on Barbados Oceanographic and Meteorological Experiment (BOMEX) observations. The scheme predicts the supersaturation field and allows secondary in-cloud activation of cloud droplets above the cloud base. Pristine and polluted cloud condensation nuclei (CCN) environments, as well as opposing subgrid-scale mixing scenarios, are contrasted. Numerical simulations show that about 40% of cloud droplets originate from CCN activated above the cloud base. Significant in-cloud activation leads to the mean cloud droplet concentration that is approximately constant with height, in agreement with aircraft observations. The in-cloud activation affects the spatial distribution of the effective radius and the mean albedo of the cloud field. Differences between pristine and polluted conditions are consistent with the authors’ previous study, but the impact of the subgrid-scale mixing is significantly reduced. Possible explanations of the latter involve physical and numerical aspects. The physical aspects include (i) the counteracting impacts of the subgrid-scale mixing and in-cloud activation and (ii) the mean characteristics of the environmental cloud-free air entrained into a cloud. A simple analysis suggests that the entrained cloud-free air is on average close to saturation, which leads to a small difference between various mixing scenarios. The numerical aspect concerns the relatively small role of the parameterized subgrid-scale mixing when compared to mixing and evaporation due to numerical diffusion. Although the results are consistent with aircraft observations, limitations of the numerical model due to low spatial resolution call for higher-resolution simulations where entrainment processes are resolved rather than mostly parameterized as in the current study.

scholarly journals Comparison of Cloud Microphysics Schemes in a Warn-on-Forecast System Using Synthetic Satellite Objects

2018 ◽  
Vol 33 (6) ◽  
pp. 1681-1708 ◽  
Author(s):  
Thomas A. Jones ◽  
Patrick Skinner ◽  
Kent Knopfmeier ◽  
Edward Mansell ◽  
Patrick Minnis ◽  
...  
Keyword(s):  
Prediction Models ◽  
Negative Impact ◽  
Cloud Condensation Nuclei ◽  
Weather Prediction ◽  
Weather Conditions ◽  
Variable Density ◽  
Cloud Microphysics ◽  
Convective Precipitation ◽  
Microphysics Scheme ◽  
Double Moment

AbstractForecasts of high-impact weather conditions using convection-allowing numerical weather prediction models have been found to be highly sensitive to the selection of cloud microphysics scheme used within the system. The Warn-on-Forecast (WoF) project has developed a rapid-cycling, convection-allowing, data assimilation and forecasting system known as the NSSL Experimental WoF System for ensembles (NEWS-e), which is designed to utilize advanced cloud microphysics schemes. NEWS-e currently (2017–18) uses the double-moment NSSL variable density scheme (NVD), which has been shown to generate realistic representations of convective precipitation within the system. However, very little verification on nonprecipitating cloud features has been performed with this system. During the 2017 Hazardous Weather Testbed (HWT) experiment, an overestimation of the areal coverage of convectively generated cirrus clouds was observed. Changing the cloud microphysics scheme to Thompson generated more accurate cloud fields. This research undertook the task of improving the cloud analysis generated by NVD while maintaining its skill for other variables such as reflectivity. Adjustments to cloud condensation nuclei (CCN), fall speed, and collection efficiencies were made and tested over a set of six severe weather cases occurring during May 2017. This research uses an object-based verification approach in which objects of cold infrared brightness temperatures, high cloud-top pressures, and cloud water path are generated from model output and compared against GOES-13 observations. Results show that the modified NVD scheme generated much more skillful forecasts of cloud objects than the original formulation without having a negative impact on the skill of simulated composite reflectivity forecasts.

GPU-ACCELERATED FINITE ELEMENT MODELING OF BIO-HEAT CONDUCTION FOR SIMULATION OF THERMAL ABLATION

2018 ◽  
Vol 18 (07) ◽  
pp. 1840012 ◽  
Author(s):  
JINAO ZHANG ◽  
JEREMY HILLS ◽  
YONGMIN ZHONG ◽  
BIJAN SHIRINZADEH ◽  
JULIAN SMITH ◽  
...  
Keyword(s):  
Finite Element ◽  
Finite Element Modeling ◽  
Thermal Ablation ◽  
Surgical Simulation ◽  
Graphics Processing Unit ◽  
Processing Unit ◽  
Modeling And Analysis ◽  
Element Modeling ◽  
Computational Performance ◽  
High Level

Efficient simulation of heating processes in thermal ablation is of great importance for surgical simulation of thermal ablation procedures. This paper presents a Graphics Processing Unit (GPU) assisted finite element methodology for modeling and analysis of bio-heat transfer processes in the treatment of thermal ablation. The proposed methodology employs finite element method for discretization of the bio-heat equation, and the finite element modeling is implemented using the High-Level Shader Language of the Microsoft Direct3D 11. Simulations and comparison analyses are conducted, demonstrating computational performance improvement of up to 55.3 times using the proposed methodology.

Sign in / Sign up

Export Citation Format

Share Document