@Article{CamposVelhoFSBWSCC:2022:PaIm,
author = "Campos Velho, Haroldo Fraga de and Furtado, Helaine Cristina
Morais and Sambatti, Sabrina B{\'e}rgoch Monteiro and Barros,
Carla Osthoff Ferreira de and Welter, Maria Eugenia Sausen and
Souto, Roberto Pinto and Carvalho, Diego and Cardoso, Douglas O.",
affiliation = "{Instituto Nacional de Pesquisas Espaciais (INPE)} and
{Universidade Federal do Oeste do Par{\'a} (UFOPA)} and {} and
{Laborat{\'o}rio Nacional de Computa{\c{c}}{\~a}o
Cient{\'{\i}}fica (LNCC)} and {Laborat{\'o}rio Nacional de
Computa{\c{c}}{\~a}o Cient{\'{\i}}fica (LNCC)} and
{Laborat{\'o}rio Nacional de Computa{\c{c}}{\~a}o
Cient{\'{\i}}fica (LNCC)} and {Centro Federal de
Educa{\c{c}}{\~a}o Tecnol{\'o}gica Celso Suckow da Fonseca} and
{Centro Federal de Educa{\c{c}}{\~a}o Tecnol{\'o}gica Celso
Suckow da Fonseca}",
title = "Data Assimilation by Neural Network for Ocean Circulation:
Parallel Implementation",
journal = "Supercomputing Frontiers and Innovations",
year = "2022",
volume = "9",
number = "1",
pages = "74--86",
keywords = "Data assimilation, Articial Neural Network, Shallow water
equations, Parallel processing.",
abstract = "Data assimilation (DA) is an essential issue for operational
prediction centers, where a com-puter code is applied to simulate
physical phenomena by solving differential equations. The
pro-cedure to determine the best initial condition combining data
from observation and previousforecasting (background) is carried
out by a data assimilation method. The Kalman filter (KF) isa
technique for data assimilation, but it is computationally
expensive. An approach to reduce thecomputational effort for DA is
to emulate the KF by a neural network. The multi-layer
perceptronneural network (MLP-NN) is employed to emulate the
Kalman in a 2D ocean circulation model,and algorithmic complexity
to KF and NN is presented. A shallow-water system models the
oceandynamics. Synthetic measurements are used for evaluating the
MLP-NN for the data assimilationprocess. Here, a parallel version
for the DA procedure by the neural network is described andtested,
showing the performance improvement for a parallel version of the
NN-DA.",
doi = "10.14529/jsfi220105",
url = "http://dx.doi.org/10.14529/jsfi220105",
issn = "2409-6008",
label = "lattes: 5142426481528206 1 CamposVelhoFSBWSCC:2022:PaIm",
language = "en",
targetfile = "superfri-2022-1-74-86.pdf",
url = "https://superfri.org/index.php/superfri/issue/view/33",
urlaccessdate = "10 maio 2024"
}