GENEST (GENERALIZED ESTIMATOR)

Este es un contenido que creo que merece la pena dejarlo localizable en un post. Se trata de GenEst (Generalized Estimator), un software escrito en R y diseñado para estimar la mortalidad de aves y murciélagos en parques eólicos y centrales solares, pero también aplicable en otro tipo de proyectos.

Captura de pantalla de GenEst. Tomado de GenEst User Guide.

Como el abstract del User Guide resume muy bien la base teórica general y las características de aplicación, me ahorro aburrir y os lo pego a continuación tal cual ;) 

ABSTRACT

GenEst (Generalized Estimator) is a software package designed for use by anyone analyzing data associated with estimating bird or bat fatalities at renewable-energy facilities, such as wind and solar facilities, but it has applicability in many other situations, as well. It is designed to addresses the general problem of estimating the size of a population when not all animals are present on all survey occasions—or a superpopulation (Williams and others 2011)—when the probability of detection is generally less than one. The population is not closed, and the probability of detection can vary among individuals due to physical characteristics of the individual, such as size, or on the environmental conditions in which the individual exists, such as vegetation or season. To estimate the number of fatalities, carcasses are usually collected during distinct searches repeated at (generally) constant intervals through time, and counts of carcasses are adjusted for imperfect detection. Imperfect detection may be due to any of several possible detection biases, for example: (1) search teams fail to find carcasses that are present in the searched area at the time of the search, (2) scavengers remove carcasses before searches are conducted, (3) carcasses fall outside the searched area, or (4) fatalities occur outside the monitored period. In parallel with the search process, investigators typically conduct field trials to estimate the effects of the first two components, and use observed locations of carcasses as well as knowledge regarding the sampling fraction to estimate the third. The fourth is often a matter of educated guess. Accurate estimation of the detection biases is critical to accurate estimation of total mortality. Because of imperfect detection, the simple count of observed carcasses does not accurately represent the actual population of animals killed by turbines and cannot be used as an experimental field trials. Included in the software are example datasets for analyses, standard R package help files, this user guide, and vignettes detailing use at the command-line.

• Acceso a GENEST