Within ICES, a lot data is collected on abundances of marine organisms, e.g. in trawl surveys, acoustic surveys, and visual surveys of e.g. cetaceans. Generally these data are collected over a large number of locations, and the spatial and temporal heterogeneity of the marine habitat causes spatial and temporal variation in the observed abundances.

Analyses of such data may reveal abundance trends of marine populations, but the concurrent effects of abiotic factors, gear selectivity, and spatial and temporal variation is challenging. Bayesian inference by means of Integrated Nested Laplace Approximation (INLA) is has proven to be a powerful numeric technique to analyse spatial and temporal variation in combination with other covariates. INLA is available as a package in R. inlabru is an extension to INLA specifically designed to make it easier to work with spatial data and to fit models with complex observation processes.

In this course, we aim to teach ecologists and stock assessors how to analyse spatial data as it is often collected in marine research. Attendees will gain an understanding of the relevant statistical concepts/methods and how to implement them using the INLA/inlabru software. Examples will be based on fitting models to observations to investigate spatial and temporal correlations and possible associations with abiotic factors using real data, e.g. as extracted from the DATRAS data base.

The objective is to teach all participants how to independently analyse spatial data collected at sea using INLA/inlabru and to understand the statistical underpinnings of their models.

Participants

Anyone who desires improving analyses of data collected at sea, including those data that are used as inputs to fishery stock assessments.

This includes survey planning expert groups, stock assessment expert groups, marine mammal ecologists (WGMME), and memebrs of the Joint OSPAR/HELCOM/ICES Working Group on Seabirds.

Course Instructors:

Janine Illian, School of Mathematics and Statistics, University of Glasgow, UK
Sara Martino, Norwegian University of Science and Technology, Norway

Prerequisites

Participants should be familiar with the R environment and general statistical approaches for modelling such as regression, analysis of covariance, and general linear models.