Wyniki wyszukiwania - Biblioteka Nauki

Ten serwis zostanie wyłączony 2025-02-11.

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: marginal distribution

Sortuj według:

Ogranicz wyniki do:

Ecological interaction networks: prospects and pitfalls

100%

Ulrich W.

Ecological Questions

2009

tom 11

Interaction networks are a tool to visualize and to study the relationships between interacting species across and within trophic levels. Recent research uncovered many properties of such networks that remained undetected in previous food web studies. These patterns could be related to evolutionary and ecological processes. The study of interaction networks promises therefore progress in the study of constraints that act on the coevolution of interacting species and on food webs. However, there are still many pitfalls associated with the statistical analysis, the properties of the metrics involved and the appropriate null model choice. Here I review the mechanisms that shape interaction matrices, the possible internal structures and their ecological interpretation, and the analytical tools to identify matrix structure. Progress in the field needs critical meta-analytical and comparative studies that indentify the best suited null models (low type I and II error probabilities and high power to disentangle statistical from ecological processes) and clarify the interdependence of different concepts and metrics associated with network approaches. It is not improbable that many patterns recently associated with ecological and evolutionary processes might turn out to be simple side effects of the sampling from the underlying metacommunity distributions.

Flood risk analysis based on nested copula structure in Armand Basin, Iran

100%

Amini S. , Bidaki R. Z. , Mirabbasi R. , Shafaei M.

Acta Geophysica

2022

tom Vol. 70, no 3

1385--1399

Merging different food characteristics in a distribution function is provided by copula structures. In this study, the nested copula structure was used to construct a triradiate distribution of food duration (D), peak (P), and volume (V). The required data were obtained by screening the food events recorded at Armand Gauging Station, Iran. The characteristics of selected 63 food events (1993–2018) were extracted and the best marginal distribution function of each was determined by Kolmogorov–Smirnov test. Then the fitness of six different copula functions (Frank, Clayton, Joe, Gumbel–Hougaard, Gaussian and Student’s t were examined for creating the joint distribution function. The best fitted marginal distribution is Johnson SB, for food duration, and Lognormal (3p), for food peak and food volume. The best-fitted function for creating bivariate and trivariate distributions of food characteristics in Armand Basin is Frank copula. In the next phase, the bivariate and trivariate joint return periods (at two states of AND, OR), Kendall return period and conditional return periods were calculated. The results revealed that the conditional return period of one food variable given two other food variables is greater than the corresponding values for the conditional return period of two food variables given the third food variable.