Hollaway, M.J.

Fuzzy changepoint application to evaluate numerical model ability to capture important shifts in environmental time series

This application is an implementation of a Fuzzy changepoint based approach to evaluate how well numerical models capture local scale temporal shifts in environmental time series. A changepoint in a time series represents a change in the statistical properties of the time series (either mean, variance or mean and variance in this case). These can often represent important local events of interest that numerical models should accurately capture. The application detects the locations of changepoints in two time series (typically one representing observations and one representing a model simulation) and estimates uncertainty on the changepoint locations using a bootstrap approach. The changepoint locations and associated confidence intervals are then converted to fuzzy numbers and fuzzy logic is used to evaluate how well the timing of any changepoints agree between the time series. The app returns individual similarity scores for each changepoint with higher scores representing a better performance of the numerical model at capturing local scale temporal changes seen in the observed record.

To use this application, the user will upload a csv file containing the two time series to be compared.

This work was supported by Engineering and Physical Sciences Research Council (EPSRC) Data Science for the Natural Environment (DSNE) project (EP/R01860X/1) and the Natural Environment Research Council (NERC) as part the UK-SCAPE programme (NE/R016429/1).

Publication date: 2021-03-01

Provenance & quality

This application comprises the source code of an R Shiny app (app.R) that implements the method described in https://doi.org/10.1016/j.envsoft.2021.104993. The application also provides access to the standalone functions used to execute the analysis (Fuzzy_CPT_functions.R). It can be run on any machine with R and the required packages installed. It has been tested up to R version 3.6.1 with more testing information and package dependencies found in Session_info.txt.


Hollaway, M.J.; Henrys, P.A.; Killick, R.; Leeson, A.; Watkins, J. (2021). Evaluating the ability of numerical models to capture important shifts in environmental time series: A Fuzzy change point approach. Environmental Modelling & Software. https://doi.org/10.1016/j.envsoft.2021.104993

Correspondence/contact details

Dr Michael Hollaway
UK Centre for Ecology & Hydrology
Lancaster Environment Centre, Library Avenue, Bailrigg


Hollaway, M.J.
UK Centre for Ecology & Hydrology

Other contacts

NERC EDS Environmental Information Data Centre
NERC Environmental Information Data Centre
Rights Holder
UK Centre for Ecology & Hydrology

Additional metadata

Topic categories
Changepoints,  Data analytics,  Data Labs Data science,  Evaluation framework,  Fuzzy logic,  UK-SCAPE,  UK Status, Change and Projections of the Environment,  Uncertainty
Engineering and Physical Sciences Research Council Award: EP/R01860X/1
Natural Environment Research Council Award: NE/R016429/1
Last updated
18 May 2022 12:29