Tracking Environmental Change Using Lake Sediments
H. John B. Birks, André F. Lotter, Steve Juggins, John P. Smol
additive modelling, additive monotone regression splines, bayesian hierarchical models, bayesian inference, biological dynamics, computing, constrained gaussian ordination, data-mining, ecological, mechanistic modelling, model selection, Modelling, multi-proxy studies
Quantitative palaeolimnology has made great advances in the last 20 years. The subject is not static, however, and as more and more demanding questions are asked of palaeolimnology in the future, there will be more and more future numerical challenges to be addressed and subjects to be explored. These include the problems of model selection, trait analysis, data-mining, time-warp analysis, quantile regression, additive modelling, new techniques for temporal- series analysis, and increasing use of Bayesian inference. The practical problems of computing and of available software are also discussed and it is clear that future developments in quantitative palaeolimnology will depend on researchers becoming proficient in the use of R and its innumerable packages relevant to palaeolimnological data-analysis.