Steve Drasco (University of St.Andrews)
Seminar Room, The Observatory: 2:00 PM, 27 Jan 2021

:Gravitational waves are a recently observed natural phenomenon predicted a century ago by Einstein's theory of gravity, general relativity.  They are currently one of the hottest fields of physics and astronomy, generating numerous scientific discoveries and accolades over the past five years.  Gravitational waves are produced by nearly any motion of matter, but only the most extreme scenarios produce an effect strong enough to be observed.  To date, all such scenarios have involved the merger of compact stellar remnants (black holes and neutron stars).  These observations demanded new technologies in engineering, optics, theoretical physics and astrophysics, and data analysis.  I will describe gravitational waves in general, as well as two example sources that I have worked on. One is deterministic in nature (the merger of two black holes of vastly different size), while the other is fundamentally probabilistic (a stochastic background that could be produced by a variety of different astrophysical events).  For both sources, I will focus on the role of statistics.

RefID: 2008 click to edit (admin only)

Andy Seaton (University of St.Andrews)
, The Observatory: 2:00 PM, 03 Feb 2021

I will present a chapter of my thesis on fitting a spatial point process model with a point transect distance sampling observation process.  I'll present distance sampling from a spatial point process perspective and draw links with the classical distance sampling literature.  The point process model is assumed to be a log-Gaussian Cox process, which allows the use of spatially structured random effects to describe the density of birds.  Here I use a Gaussian Markov random field approach based on a stochastic partial differential equation representation of the random field.  Both the spatial model and the observation model are fitted jointly using integrated nested Laplace approximations (INLA).  I will introduce a novel approach to dealing with model components which are non-linear in their parameters - in this case the distance sampling detection function.  This is based on a linearisation approach that optimises over "non-linear" parameters, fitting the model with INLA for each estimate of the non-linear model component.  This linearisation scheme is implemented in the R package inlabru - which extends the functionality of R-INLA and provides useful wrappers that simplify the INLA code required specify, fit, and interrogate models fitted using INLA.  I will show results from a case study applied to a single-year of a point transect distance sampling survey of a tropical bird endemic to Hawai'i.  The model is readily extendible to multiple years.  I will also briefly discuss some issues with summarising and visualising results from spatial models with random fields.  To end I will mention some general ideas where I think this linearisation approach could be useful to fitting models in statistical ecology more generally - this is not completed work, just some ideas for the future.

RefID: 2012 click to edit (admin only)

Dr Gideon Bradfurd (Michigan State University)
, The Observatory: 2:00 PM, 24 Feb 2021RefID: 2011 click to edit (admin only)

Timo Adam (CREEM, University of St.Andrews.)
, The Observatory: 2:00 PM, 12 May 2021RefID: 2016 click to edit (admin only)