EurekaMag.com logo
+ Translate

Can ARMA models be used reliably in ecology? Les modeles ARMA peuvent-ils etre utilises avec confiance en ecologie?


, : Can ARMA models be used reliably in ecology? Les modeles ARMA peuvent-ils etre utilises avec confiance en ecologie?. Acta Oecologica. July; 184: 427-447

Several characteristics in ecological time series (limited size, non-normal distribution, missing data) are usually put forward to justify that statistical methods as ARMA modeling (AutoRegressive Moving Average) are not used unlike in economics or in hydrology. However, none study of sensitivity has been published on applying ARMA method to ecological data. We tested its robustness by simulation and considered four factors: the nature of the generating process (AR, MA or ARMA), the length of the time series, the error model and the criterion used to select a model. We studied the probability of correctly identifying the generating process in the simulation for all combinations of the four factors. This probability is high (about 908) only if time series length exceeds 50 points, if the generating process is simple (pure AR or MA) with a parameter having a high modulus (about 0.8) and if SBC is the selecting criterion; a mixed process (ARMA) was correctly identified at best in 10% to 20% of the simulations. Among the three criterions tested, AICC leads to the best selection of a model, especially with low time series length. All models are not sensitive to asymmetry of the error model in our simulations. The results are discussed according to the goals of the ecologists when analysing time series. Limitations of an automatic identification of a model are underlined.

(PDF 0-2 workdays service)

Accession: 038040231

Submit PDF Full Text: Here


Submit PDF Full Text

No spam - Every submission is manually reviewed

Due to poor quality, we do not accept files from Researchgate

Submitted PDF Full Texts will always be free for everyone
(We only charge for PDFs that we need to acquire)

Select a PDF file:
Close
Close

Related references

Anonymous, 1997: Les modeles ARMA peuvent-ils etre utilises avec confiance en ecologie?

Malgras, J.; Debouzie, D., 1997: Can ARMA models be used with confidence in ecology?. Several characteristics in ecological time series (limited size, non-normal distribution, missing data) are usually put forward to justify that statistical methods as ARMA modeling (AutoRegressive Moving Average) are not used unlike in economics o...

Zou, D.; Coudron, T.A.; Liu, C.; Zhang, L.; Wang, M.; Chen, H., 2013: Nutrigenomics in Arma chinensis: transcriptome analysis of Arma chinensis fed on artificial diet and Chinese oak silk moth Antheraea pernyi pupae. The insect predator, Arma chinensis, is capable of effectively controlling many pests, such as Colorado potato beetle, cotton bollworm, and mirid bugs. Our previous study demonstrated several life history parameters were diminished for A. chinensi...

Megnin, P., 1881: Sur de petite Helminthes agames enkysiees qui peuvent etre confondus et qui l'ont e te avec la Trichina spiralis. Attention is directed to Megnin"s paper on Trichina (34).

Megnin, P., 1880: Les Parasites et les maladies parasitaires chez l'homme, les animaux domestiques et les animaux sauvages, avec lesquels ils peuvent etre on contact. Insectes, Arachnides, Crustacea. Avec 63 figures dans le texte, et un Atlas de 26 planches dessinees par l'auteur. A useful treatise, divided into six chapters, dealing with Diptera, Hemiptera, Aphaniptera and Coleoptera, Epizoa, Acari, and Crustacea.

Kestens, E., 1998: Time series and (S)ARMA models. Journal of Diarrhoeal Diseases Research 16(2): 84-86

Chang M.K.; Kwiatkowski J.W.; Nau R.F.; Oliver R.M.; Pister K.S., 1982: ARMA models for earthquake ground motions. This paper outlines the use of discrete, autoregressive/moving-average (ARMA) models for identification and estimation of parameters in models derived from analysis of uniformly digitized earthquake ground motion acceleration data. Such models are...

Aurass H.; Kurths J., 1979: Description of discreet measurements with ARMA models. Gerlands Beitraege zur Geophysik: , Pages 249-260. 1979.

Etchison, T.; Brownie, C.; Pantula, S.G., 1995: A portmanteau test for spatial ARMA models. For assessing the fit of separable autoregressive moving average (ARMA) models, Cullis and Gleeson (1991, Biometrics 47, 1449-1460) considered a portmanteau test for lack of fit which is an extension of the Ljung-Box statistics used in time series...

Salas Jose D.; Boes Duane C.; Smith Ricardo A., 1982: Estimation of ARMA models with seasonal parameters. The Yule-Walker equations for ARMA (p, q) models with periodic parameters are derived from which moment estimates can be obtained. Specifically,for the case of ARMA (p, 1) models, the periodic autoregressive parameters can be found by solving a sy...