Simulation of global and hemispheric temperature variations and signal detection studies using neural networks
The concept of neural network models is a statistical strategy which can be used if a superposition of any forcing mechanisms leads to any effects and if a sufficient related observational data base is available. In comparison to multiple regression analysis, the main advantages are that neural netw...
Published in: | Meteorologische Zeitschrift |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Borntraeger
1998
|
Subjects: | |
Online Access: | https://doi.org/10.1127/metz/7/1998/171 https://doaj.org/article/86548523532c4fe1bb048eb05187e363 |
id |
ftdoajarticles:oai:doaj.org/article:86548523532c4fe1bb048eb05187e363 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:86548523532c4fe1bb048eb05187e363 2024-09-15T18:24:19+00:00 Simulation of global and hemispheric temperature variations and signal detection studies using neural networks A. Walter M. Denhard C.-D. Schönwiese 1998-08-01T00:00:00Z https://doi.org/10.1127/metz/7/1998/171 https://doaj.org/article/86548523532c4fe1bb048eb05187e363 EN eng Borntraeger http://dx.doi.org/10.1127/metz/7/1998/171 https://doaj.org/toc/0941-2948 0941-2948 doi:10.1127/metz/7/1998/171 https://doaj.org/article/86548523532c4fe1bb048eb05187e363 Meteorologische Zeitschrift, Vol 7, Iss 4, Pp 171-180 (1998) neuron layers kohonen layer grossberg layer neutral networks regression analysis neuronale netzwerkmodelle regressionsanalyse Meteorology. Climatology QC851-999 article 1998 ftdoajarticles https://doi.org/10.1127/metz/7/1998/171 2024-08-05T17:50:04Z The concept of neural network models is a statistical strategy which can be used if a superposition of any forcing mechanisms leads to any effects and if a sufficient related observational data base is available. In comparison to multiple regression analysis, the main advantages are that neural network models are an appropriate tool also in the case of non-linear cause-effect relations and that interactions of the forcing mechanisms are allowed. In comparison to more sophisticated methods like general circulation models, the main advantage is that details of the physical background, like feedbacks, can be unknown. Neural networks learn from observations which reflect feedbacks implicity. The disadvantage, of course, is that the physical background is neglected. In addition, the results prove to be sensitively dependent from the network architecture like the number of hidden neurons or the initialisation of learning parameters. We used a supervised learning backpropagationnetworkwith three neuron layers, as well as an unsupervised learning counterpropagation network. This latter network architecture consists of a so-called Kohonen layer, where the internal features of the data set in question are "learned", which is followed by a so-called Grossberg layer, where the "learned" internal representations are adjusted to the observed time series to be simulated. Both concepts (backpropagation, counterpropagation) are tested in respect to their ability to simulate the observed global as well as hemispheric mean surface air temperature annual variations 1874-1993 if parameter time series of the following forcing mechanisms are incorporated: equivalent CO2 concentrations, tropospheric sulfate aerosol concentrations (both anthropogenic), volcanism, solar activity, and ENSO (all natural). It arises that in this way up to 83 % of the observed temperature variance can be explained, significantly more than by multiple regression analysis. The implication of the North Atlantic Oscillation does not improve these results. On a ... Article in Journal/Newspaper North Atlantic North Atlantic oscillation Directory of Open Access Journals: DOAJ Articles Meteorologische Zeitschrift 7 4 171 180 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
neuron layers kohonen layer grossberg layer neutral networks regression analysis neuronale netzwerkmodelle regressionsanalyse Meteorology. Climatology QC851-999 |
spellingShingle |
neuron layers kohonen layer grossberg layer neutral networks regression analysis neuronale netzwerkmodelle regressionsanalyse Meteorology. Climatology QC851-999 A. Walter M. Denhard C.-D. Schönwiese Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
topic_facet |
neuron layers kohonen layer grossberg layer neutral networks regression analysis neuronale netzwerkmodelle regressionsanalyse Meteorology. Climatology QC851-999 |
description |
The concept of neural network models is a statistical strategy which can be used if a superposition of any forcing mechanisms leads to any effects and if a sufficient related observational data base is available. In comparison to multiple regression analysis, the main advantages are that neural network models are an appropriate tool also in the case of non-linear cause-effect relations and that interactions of the forcing mechanisms are allowed. In comparison to more sophisticated methods like general circulation models, the main advantage is that details of the physical background, like feedbacks, can be unknown. Neural networks learn from observations which reflect feedbacks implicity. The disadvantage, of course, is that the physical background is neglected. In addition, the results prove to be sensitively dependent from the network architecture like the number of hidden neurons or the initialisation of learning parameters. We used a supervised learning backpropagationnetworkwith three neuron layers, as well as an unsupervised learning counterpropagation network. This latter network architecture consists of a so-called Kohonen layer, where the internal features of the data set in question are "learned", which is followed by a so-called Grossberg layer, where the "learned" internal representations are adjusted to the observed time series to be simulated. Both concepts (backpropagation, counterpropagation) are tested in respect to their ability to simulate the observed global as well as hemispheric mean surface air temperature annual variations 1874-1993 if parameter time series of the following forcing mechanisms are incorporated: equivalent CO2 concentrations, tropospheric sulfate aerosol concentrations (both anthropogenic), volcanism, solar activity, and ENSO (all natural). It arises that in this way up to 83 % of the observed temperature variance can be explained, significantly more than by multiple regression analysis. The implication of the North Atlantic Oscillation does not improve these results. On a ... |
format |
Article in Journal/Newspaper |
author |
A. Walter M. Denhard C.-D. Schönwiese |
author_facet |
A. Walter M. Denhard C.-D. Schönwiese |
author_sort |
A. Walter |
title |
Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
title_short |
Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
title_full |
Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
title_fullStr |
Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
title_full_unstemmed |
Simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
title_sort |
simulation of global and hemispheric temperature variations and signal detection studies using neural networks |
publisher |
Borntraeger |
publishDate |
1998 |
url |
https://doi.org/10.1127/metz/7/1998/171 https://doaj.org/article/86548523532c4fe1bb048eb05187e363 |
genre |
North Atlantic North Atlantic oscillation |
genre_facet |
North Atlantic North Atlantic oscillation |
op_source |
Meteorologische Zeitschrift, Vol 7, Iss 4, Pp 171-180 (1998) |
op_relation |
http://dx.doi.org/10.1127/metz/7/1998/171 https://doaj.org/toc/0941-2948 0941-2948 doi:10.1127/metz/7/1998/171 https://doaj.org/article/86548523532c4fe1bb048eb05187e363 |
op_doi |
https://doi.org/10.1127/metz/7/1998/171 |
container_title |
Meteorologische Zeitschrift |
container_volume |
7 |
container_issue |
4 |
container_start_page |
171 |
op_container_end_page |
180 |
_version_ |
1810464638949654528 |