Results 1 
4 of
4
2013: A hybrid MPI/OpenMP parallel algorithm and performance analysis for an ensemble square root filter suitable for dense observations
 J
"... A hybrid parallel scheme for the ensemble square root filter (EnSRF) suitable for parallel assimilation of multiscale observations, including those from dense observational networks such as those of radar, is developed based on the domain decomposition strategy. The scheme handles internode communi ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
A hybrid parallel scheme for the ensemble square root filter (EnSRF) suitable for parallel assimilation of multiscale observations, including those from dense observational networks such as those of radar, is developed based on the domain decomposition strategy. The scheme handles internode communication through a message passing interface (MPI) and the communication within sharedmemory nodes via Open Multiprocessing (OpenMP) threads. It also supports pure MPI and pure OpenMPmodes. The parallel framework can accommodate highvolume remotesensed radar (or satellite) observations as well as conventional observations that usually have larger covariance localization radii. The performance of the parallel algorithm has been tested with simulated and real radar data. The parallel program shows good scalability in pure MPI and hybrid MPI–OpenMP modes, while pure OpenMP runs exhibit limited scalability on a symmetric sharedmemory system. It is found that inMPImode, better parallel performance is achieved with domain decomposition configurations in which the leading dimension of the state variable arrays is larger, because this configuration allows for more efficient memory access. Given a fixed amount of computing resources, the hybrid parallel mode is preferred to pure MPI mode on supercomputers with nodes containing sharedmemory cores. The overall performance is also affected by factors such as the cache size, memory bandwidth, and the networking topology. Tests with a real data case with a large number of radars confirm that the parallel data assimilation can be done on amulticore supercomputer with a significant speedup compared to the serial data assimilation algorithm. 1.
Submitted to Monthly Weather Review
, 2013
"... In recent studies, the authors have successfully demonstrated the ability of an ensemble Kalman filter (EnKF), assimilating real radar observations, to produce skillful analyses and subsequent ensemblebased probabilistic forecasts for a tornadic mesoscale convective system (MCS) that occurred over ..."
Abstract
 Add to MetaCart
In recent studies, the authors have successfully demonstrated the ability of an ensemble Kalman filter (EnKF), assimilating real radar observations, to produce skillful analyses and subsequent ensemblebased probabilistic forecasts for a tornadic mesoscale convective system (MCS) that occurred over Oklahoma and Texas on 9 May 2007. The current study expands upon this prior work, performing experiments for this case on a larger domain using a nestedgrid EnKF which accounts for mesoscale uncertainties through the initial ensemble and lateral boundary condition perturbations. In these new experiments, conventional observations (including surface, wind profiler, and upperair observations) are assimilated in addition to the WSR88D and CASA radar data used in the previous studies, better representing meso and convectivescale features. The relative impacts of conventional and radar data on analyses and forecasts are examined, and biases within the ensemble are investigated. Compared to prior results, the radarassimilating experiments accounting for mesoscale uncertainties produce superior forecasts based on both subjective and objective verification metrics. The new experiments produce a substantiallyimproved forecast, including better
Corresponding author address:
, 2013
"... A hybrid parallel scheme for the ensemble square root filter (EnSRF) suitable for parallel assimilation of multiscale observations including those from dense observational networks such as those of radar is developed based on the domain decomposition strategy. The scheme handles internode communic ..."
Abstract
 Add to MetaCart
A hybrid parallel scheme for the ensemble square root filter (EnSRF) suitable for parallel assimilation of multiscale observations including those from dense observational networks such as those of radar is developed based on the domain decomposition strategy. The scheme handles internode communication through message passing interface (MPI), and the communication within sharedmemory nodes via Open MultiProcessing (OpenMP) threads; it also supports pure MPI and pure OpenMP modes. The parallel framework can accommodate highvolume remotesensed radar (or satellite) observations as well as conventional observations that usually have larger covariance localization radii. The performance of the parallel algorithm has been tested with simulated and real radar data. The parallel program shows good scalability in pure MPI and hybrid MPI/OpenMP modes, while pure OpenMP runs exhibit limited scalability on a symmetric sharedmemory system. It is found that in MPI mode, better parallel performance is achieved with domain decomposition configurations in which the leading dimension of the state variable arrays is larger, because this configuration allows for more efficient memory access. Given a fixed amount of computing
Towards a theory of optimal localisation
, 2014
"... All practical ensemblebased atmospheric data assimilation (DA) systems use localisation to reduce the damaging impact of spurious longrange correlations arising from the finite ensemble size. However, the form of the localisation function is generally adhoc, and requires expensive tuning to optim ..."
Abstract
 Add to MetaCart
All practical ensemblebased atmospheric data assimilation (DA) systems use localisation to reduce the damaging impact of spurious longrange correlations arising from the finite ensemble size. However, the form of the localisation function is generally adhoc, and requires expensive tuning to optimise the system. For the case of a single observation and known true background error correlation, we derive an expression for the localisation factor that minimises the expected rootmeansquare (RMS) analysis error. Idealised tests show this formulation performs well for multiple observations provided their density is not too high. The width of the optimal localisation function scales with the width of the underlying correlation, but does not have the same shape. The optimal observationspace localisation for a single spatially integrating observation depends on the observationtogridpoint background error correlation, making it broader than the optimal localisation for point observations and potentially competitive with modelspace localisation. A new form of hybrid DA is proposed in which localisation damps the sample correlations towards their climatological mean rather than zero, reducing the RMS error and potentially improving the dynamic balance of the analysis. The presence of variance errors causes the optimal localisation factor to depend on the ratio of observation to background error variance, and raises the possibility that a small amount of variance damping may be beneficial. For dense observations, a more elaborate theory is required, which will almost certainly depend on the observation network. We present some preliminary analysis of the features of the multiobservation problem, which for instance suggests that the optimal solution may involve different localisation factors in the numerator and denominator of the Kalman filter equation. We note that even optimal localisation gives an expected RMS error which exceeds that of perfect DA, contrary to the assumption made by ‘deterministic ’ ensemble filters.