Vacancy for post-doctoral Ecological Statistician at Centre for Research
into Ecological and Environmental Modelling (CREEM), University of St
Andrews
Vacancy Description
School of Mathematics and Statistics, Salary: £33,199 - £36,261 per
annum, Start Date: 1 September 2019 or as soon as possible thereafter,
Fixed Term for 18 months
This is an 18-month fixed-term post. We are looking for an ecological
statistician to work as part of an inter-disciplinary project to develop
new quantitative models and analytical methods for inferring behavioural
response of marine mammal species to US Navy sonar (see Harris et al.
2018 for a recent review). The post will involve working on two tasks
within the larger project: (1) developing algorithms for fitting
multi-species dose-response functions and applying them to experimental
data to determine which species can be grouped together in terms of
their responsiveness to sonar; (2) comparing methods for inferring
behavioural response from long-term passive acoustic data collected
during observational studies in the vicinity of Navy sonar exercises.
The first task builds upon established Bayesian methods for modelling
probability of behavioural response as a function of sonar dose (Miller
et al. 2014, Antunes et al. 2015). These focused on single-species
analyses. Extensions have been made that allow for multi-species
modelling, including model selection for which species should share
common dose-response functions and which should be different. However,
the current implementation is computationally expensive and hence a
bespoke algorithm, likely based on reversible jump Markov chain Monte
Carlo (RJMCMC), is desired. Analyses of real-world experimental data
already collected is also part of this task. The second task will
involve, at least partly, a simulation study comparing two approaches
that have already been applied to long-term passive acoustic data (e.g.,
Oswald et al. 2016): Generalized Estimating Equations (GEEs) and Hidden
Markov Models (HMMs).
We welcome applications from candidates with a PhD in Statistics or a
closely-related discipline, who have an interest in using and developing
statistical methods to solve real-world problems in ecology. Experience
with and/or research interests in Bayesian analysis, RJMCMC, HMMs and
GEEs would be highly advantageous, but is not a requirement. The
overall project is a collaboration between ecologists, statisticians,
oceanographers and acousticians and so the ability to communicate
effectively between disciplines is essential.
The successful candidate will be based in the world-leading Centre for
Ecological and Environmental Modelling (CREEM) under the supervision of
Professor Len Thomas and Dr. Catriona Harris. CREEM has an excellent
record of retaining research staff, so while this is a fixed-term
18-month post there may be prospects for continuation beyond that period.
For informal enquiries, we encourage those considering applying to
contact Professor Len Thomas ([log in to unmask]) and/or Dr.
Catriona Harris ([log in to unmask]).
Applications are particularly welcome from women, who are
under-represented in Science posts at the University. You can find out
more about Equality and Diversity at http://www.st-andrews.ac.uk/hr/edi/.
The University is committed to equality for all, demonstrated through
our working on diversity awards (ECU Athena SWAN/Race Charters; Carer
Positive; LGBT Charter; and Stonewall). More details can be found at
http://www.st-andrews.ac.uk/hr/edi/diversityawards/.
Post: Research Fellow - AR2228HM
Closing Date: 5 July 2019
Please quote ref: AR2228HM
Summary info: https://tinyurl.com/y3najjvx
Further Particulars: https://tinyurl.com/yxo2yvb4
--
Len Thomas [log in to unmask] lenthomas.org @len_thom
Centre for Research into Ecological and Environmental Modelling
The Observatory, University of St Andrews, Scotland KY16 9LZ
Office: UK+1334-461801 Admin: UK+1334-461842
The University of St Andrews is a charity
registered in Scotland, No SC013532.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|