Skip to contents

Getting started

R-CMD-check CRAN status License: MIT Lifecycle: experimental

Overview

weathergenr is an R package that implements a semiparametric, multivariate, multisite stochastic weather generator designed for climate risk and stress-testing applications. The approach is conceptually based on the framework of Steinschneider & Brown (2013) and is adapted here for gridded datasets and netcdf-based workflows.

The package is intended for workflows such as:

  • climate risk and stress testing
  • hydrological and water-resources modelling
  • scenario analysis for climate adaptation studies

Methodological framework

The generator represents climate variability across multiple time scales by coupling low-frequency climate dynamics with realistic daily weather sequences. It consists of three components:

1. Low-frequency climate variability (WARM) Interannual to decadal variability is modeled with wavelet autoregressive methods applied to annual climate aggregates. This preserves persistence and spectral structure and defines annual climate states that condition daily weather.

2. Daily weather generation (Markov chain + KNN) Wet-dry persistence is simulated with a multi-state Markov chain, while daily precipitation and temperature values are generated via K-nearest-neighbour resampling. This maintains seasonality, cross-variable dependence, and spatial coherence.

3. Climate perturbation and stress testing Quantile-based perturbations impose controlled changes in means, variability, and extremes, enabling systematic climate stress testing while preserving internal consistency.

Intended use

The resulting framework is intended for bottom-up climate vulnerability assessments to explore system response under a wide range of plausible climate conditions.

Key features

  • netcdf i/o for convenient coupling with other models
  • efficient generation, filtering, and evaluation of large stochastic ensembles
  • integrated diagnostics and visualization for validation and analysis

Installation

Install the latest version from GitHub:

# install.packages("devtools")
# devtools::install_github("Deltares-research/weathergenr")

Getting started

A quick tutorial is available here:
https://deltares-research.github.io/weathergenr/articles/getting_started.html

References

Steinschneider, S., & Brown, C. (2013). A semiparametric multivariate, multisite weather generator with low-frequency variability for use in climate risk assessments. Water Resources Research, 49(11), 7205-7220. (https://doi.org/10.1002/wrcr.20528)

License

Licensed under the MIT License. See LICENSE for details.