In this document is described the OceanVAR software (a three- dimensional variational data assimilation system, 3dvar) developed at the CMCC for use with the NEMO modeling system and SLA, ARGO, XBT and GLIDER observations.OceanVar software was developed using Fortran90 and it uses the NetCDF...

The Kalman filter (KF) dates back to 1960, when R. E. Kalman provided a recursive algorithm to compute the solution of a (linear) data filtering and prediction problem, proving to be much more efficient than the N. Wiener's approach, introduced in 1949.Data filtering is a simple example of Data...

This paper addresses scientific challenges related to high level implementation strategies that leads NEMO to effectively use of the opportunities of exascale systems. We consider two software modules as proof-of-concept: the Sea Surface Height equation solver and the Variational Data...

In this document is described a L-BFGS subroutine for large-scale constrained optimization algorithm implemented in GPU-CUDA. Here is provided a brief description of the L-BFGS algorithm developed and a brief tour around the main steps of the subroutine is provided. Significant efforts have been...

We propose a new gradient method for quadratic programming, named SDC, which alternates some steepest descent (SD) iterates with some gradient iterates that use a constant steplength computed through the Yuan formula. The SDC method exploits the asymptotic spectral behaviour of the Yuan...