Which system variables carry robust early signs of upcoming phase transition? An ecological example

Página do item simplificado
dc.creatorEhsan Negahbani
dc.creatorD. Alistair Steyn-Ross
dc.creatorMoira L. Steyn-Ross
dc.creatorLuis Antonio Aguirre
dc.date.accessioned2025-03-25T16:53:10Z
dc.date.accessioned2025-09-09T00:33:47Z
dc.date.available2025-03-25T16:53:10Z
dc.date.issued2016
dc.identifier.doihttps://doi.org/10.1371/journal.pone.0163003
dc.identifier.issn1932-6203
dc.identifier.urihttps://hdl.handle.net/1843/80913
dc.languageeng
dc.publisherUniversidade Federal de Minas Gerais
dc.relation.ispartofPlos One
dc.rightsAcesso Restrito
dc.subjectMatemática para engenharia - Processamento de dados
dc.subjectMATLAB (Programa de computador)
dc.subject.otherecological model describes interaction dynamics between a predator and an age-structured prey composed of juvenile and adult
dc.subject.otherMATLAB, extending numerical experiments to close vicinity of SN point
dc.titleWhich system variables carry robust early signs of upcoming phase transition? An ecological example
dc.typeArtigo de periódico
local.citation.issue9
local.citation.volume11
local.description.resumoGrowth of critical fluctuations prior to catastrophic state transition is generally regarded as a universal phenomenon, providing a valuable early warning signal in dynamical systems. Using an ecological fisheries model of three populations (juvenile prey J, adult prey A and predator P), a recent study has reported silent early warning signals obtained from P and A populations prior to saddle-node (SN) bifurcation, and thus concluded that early warning signals are not universal. By performing a full eigenvalue analysis of the same system we demonstrate that while J and P populations undergo SN bifurcation, A does not jump to a new state, so it is not expected to carry early warning signs. In contrast with the previous study, we capture a significant increase in the noise-induced fluctuations in the P population, but only on close approach to the bifurcation point; it is not clear why the P variance initially shows a decaying trend. Here we resolve this puzzle using observability measures from control theory. By computing the observability coefficient for the system from the recordings of each population considered one at a time, we are able to quantify their ability to describe changing internal dynamics. We demonstrate that precursor fluctuations are best observed using only the J variable, and also P variable if close to transition. Using observability analysis we are able to describe why a poorly observable variable (P) has poor forecasting capabilities although a full eigenvalue analysis shows that this variable undergoes a bifurcation. We conclude that observability analysis provides complementary information to identify the variables carrying early-warning signs about impending state transition.
local.publisher.countryBrasil
local.publisher.departmentENG - DEPARTAMENTO DE ENGENHARIA ELETRÔNICA
local.publisher.initialsUFMG
local.url.externahttps://journals.plos.org/plosone/article?id=10.1371/journal.pone.0163003

Arquivos

Licença do pacote

Agora exibindo 1 - 1 de 1
Imagem de Miniatura
Nome:
License.txt
Tamanho:
1.99 KB
Formato:
Plain Text
Descrição:
Baixar

Coleções

Artigo de Periódico

Logo UFMG