Preprints
https://doi.org/10.5194/gmd-2016-204
https://doi.org/10.5194/gmd-2016-204

Submitted as: development and technical paper 21 Sep 2016

Submitted as: development and technical paper | 21 Sep 2016

Review status: this preprint was under review for the journal GMD. A revision for further review has not been submitted.

Technical Note: Improving the computational efficiency of sparse matrix multiplication in linear atmospheric inverse problems

Vineet Yadav1 and Anna M. Michalak2 Vineet Yadav and Anna M. Michalak
  • 1Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, 91011, USA
  • 2Department of Global Ecology, Carnegie Institution for Science, Stanford, California, 94305, USA

Abstract. Matrix multiplication of two sparse matrices is a fundamental operation in linear Bayesian inverse problems for computing covariance matrices of observations and a posteriori uncertainties. Applications of sparse-sparse matrix multiplication algorithms for specific use-cases in such inverse problems remain unexplored. Here we present a hybrid-parallel sparse-sparse matrix multiplication approach that is more efficient by a third in terms of execution time and operation count relative to standard sparse matrix multiplication algorithms available in most libraries. Two modifications of this hybrid-parallel algorithm are also proposed for the types of operations typical of atmospheric inverse problems, which further reduce the cost of sparse matrix multiplication by yielding only upper triangular and/or dense matrices.

Vineet Yadav and Anna M. Michalak

 
Status: closed (peer review stopped)
Status: closed (peer review stopped)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
Printer-friendly Version - Printer-friendly version Supplement - Supplement
 
Status: closed (peer review stopped)
Status: closed (peer review stopped)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
Printer-friendly Version - Printer-friendly version Supplement - Supplement

Vineet Yadav and Anna M. Michalak

Vineet Yadav and Anna M. Michalak

Viewed

Total article views: 1,274 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
851 356 67 1,274 90 63 84
  • HTML: 851
  • PDF: 356
  • XML: 67
  • Total: 1,274
  • Supplement: 90
  • BibTeX: 63
  • EndNote: 84
Views and downloads (calculated since 21 Sep 2016)
Cumulative views and downloads (calculated since 21 Sep 2016)

Viewed (geographical distribution)

Total article views: 1,164 (including HTML, PDF, and XML) Thereof 1,160 with geography defined and 4 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 26 Oct 2021
Download
Short summary
Multiplication of two matrices that consists of few non-zero entries is a fundamental operation in problems that involve estimation of greenhouse gas fluxes from atmospheric measurements. To increase computational efficiency of estimating these quantities, modification of the standard matrix multiplication algorithm for multiplying these matrices is proposed in this research.