Sensitivity study of a large-scale air pollution model by using high-performance computations and Monte Carlo algorithms

Large-scale Monte Carlo simulations require high-quality random number generators to ensure correct results. The contrapositive of this statement is also true — the quality of random number generators can be tested by using them in large-scale Monte Carlo simulations. We have tested many commonly-used random number generators with high precision Monte Carlo simulations of the 2-d Ising model using the Metropolis, Swendsen-Wang, and Wolff algorithms. This work is being extended to the testing of random number generators for parallel computers. The results of these tests are presented, along with recommendations for random number generators for high-performance computers, particularly for lattice Monte Carlo simulations.

Download Full-text

Building Sensor Grid Architecture for Large-Scale Air Pollution Data Management

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.831.276 ◽

2013 ◽

Vol 831 ◽

pp. 276-281

Author(s):

Ya Jie Ma ◽

Zhi Jian Mei ◽

Xiang Chuan Tian

Keyword(s):

Air Pollution ◽

Data Management ◽

High Performance ◽

Large Scale ◽

Low Cost ◽

Physical World ◽

Sensor Nodes ◽

Grid Architecture ◽

Autonomous Sensor ◽

Efficient Data

Large-scale sensor networks are systems that a large number of high-throughput autonomous sensor nodes are distributed over wide areas. Much attention has paid to provide efficient data management in such systems. Sensor grid provides low cost and high performance computing to physical world data perceived through sensors. This article analyses the real-time sensor grid challenges on large-scale air pollution data management. A sensor grid architecture for pollution data management is proposed. The processing of the service-oriented grid management is described in psuedocode. A simulation experiment investigates the performance of the data management for such a system.

Download Full-text

Validation of a 3-D hemispheric nested air pollution model

Atmospheric Chemistry and Physics Discussions ◽

10.5194/acpd-3-3543-2003 ◽

2003 ◽

Vol 3 (4) ◽

pp. 3543-3588 ◽

Cited By ~ 12

Author(s):

L. M. Frohn ◽

J. H. Christensen ◽

J. Brandt ◽

C. Geels ◽

K. M. Hansen

Keyword(s):

Air Pollution ◽

Weather Prediction ◽

Sensitivity Study ◽

Horizontal Resolution ◽

Environmental Research ◽

Pollution Transport ◽

Model Calculations ◽

Mean Values ◽

Chemical Scheme ◽

Air Pollution Model

Abstract. Several air pollution transport models have been developed at the National Environmental Research Institute in Denmark over the last decade (DREAM, DEHM, ACDEP and DEOM). A new 3-D nested Eulerian transport-chemistry model: REGIonal high resolutioN Air pollution model (REGINA) is based on modules and parameterisations from these models as well as new methods. The model covers the majority of the Northern Hemisphere with currently one nest implemented. The horizontal resolution in the mother domain is 150 km × 150 km, and the nesting factor is three. A chemical scheme (originally 51 species) has been extended with a detailed description of the ammonia chemistry and implemented in the model. The mesoscale numerical weather prediction model MM5v2 is used as meteorological driver for the model. The concentrations of air pollutants, such as sulphur and nitrogen in various forms, have been calculated, applying zero nesting and one nest. The model setup is currently being validated by comparing calculated values of concentrations to measurements from approximately 100 stations included in the European Monitoring and Evalutation Programme (EMEP). The present paper describes the physical processes and parameterisations of the model together with the modifications of the chemical scheme. Validation of the model calculations by comparison to EMEP measurements for a summer and a winter month is shown and discussed. Furthermore, results from a sensitivity study of the model performance with respect to resolution in emission and meteorology input data is presented. Finally the future prospects of the model are discussed. The overall validation shows that the model performs well with respect to correlation for both monthly and daily mean values.

Download Full-text

Optimized stochastic methods for sensitivity analysis for large-scale air pollution model

10.15439/2021f51 ◽

2021 ◽

Author(s):

Venelin Todorov ◽

Tzvetan Ostromsky ◽

Ivan Dimov ◽

Rayna Georgieva

Keyword(s):

Air Pollution ◽

Sensitivity Analysis ◽

Large Scale ◽

Stochastic Methods ◽

Air Pollution Model

Download Full-text