About this Research Topic
Modern developments in geoscience are due in part to the availability of large amounts of data, which in most cases are acquired in time domain. Indeed, time dependent data are ubiquitous in diverse fields of the Earth Science such as atmosphere, hydrosphere, cryosphere, solid earth geophysics, volcanology, natural hazards and so on. In all these domains, different kinds of time series analyses and modelling are contributing to improving our understanding and prediction of the mechanisms behind the complex system “Earth”.
A non-exhaustive list of such linear and nonlinear analysis concepts includes time-frequency analysis, correlation and variogram analysis, autoregressive models, forecasting approaches, univariate or multivariate statistical analysis methods, denoising algorithms, stationarity and seasonality analysis, fractals and multifractals. In this context, the recent exponential growth of machine learning developments and applications have opened new perspectives in the advanced analysis of data time series for pattern discovery and recognition. Related techniques are also often used in data pre-processing and reduction the dimensionality of data. They help to find hidden patterns within data, contributing to answering emerging questions in the different fields of geoscience and overcoming persistent difficulties in analysing and interpreting of real-world complex case studies. Applications of recent advances in time series analysis methodology covering important problems in geoscience range from climate change to geophysics, seismic and volcanic monitoring and early warning, exploration, geochemistry, pollution and natural hazards forecasting and so on.
This Research Topic is devoted to integrating the different aspects of advanced time series techniques in geosciences from fundamental theory to the applications, putting the main emphasis on new results, either methodological or using unique data sets.
We expect to gather showcases of recent advances and novel applications of time series analysis in the full variety of research topics in Earth Sciences. The ultimate goal is to highlight the cutting edge analysis and modelling approaches to identify most relevant approaches and common trends across different areas of geosciences. Accepted manuscripts may cover one or several of the following topics:
Advanced exploratory analysis and visualisation of geoscientific time series;
Analysis and quantification of time series complexity;
Applications of machine learning techniques for time series modelling and forecasting;
Information retrieval from time series data mining; and
Innovative time domain denoising techniques for data exploration.
All article types are welcome, but in particular we encourage Original Research, Methods, Reviews and Mini Reviews, Brief Research Reports, Technology and Code.
Keywords: geophysical time series modelling, nonlinear time series in geoscience, time series prediction in geoscience, machine learning and pattern recognition in time series, spectral analysis and Kalman filtering, neural networks in geoscience, time-frequency analysis of non-stationary time series in geoscience
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.