Welcome to the clim-recal repository!

Welcome to clim-recal, a specialized resource designed to tackle systematic errors or biases in Regional Climate Models (RCMs). As researchers, policy-makers, and various stakeholders explore publicly available RCMs, they need to consider the challenge of biases that can affect the accurate representation of climate change signals.

clim-recal provides both a broad review of available bias correction methods as well as software, practical tutorials and guidance that helps users apply these methods methods to various datasets.

clim-recal is an extensive software library and guide to application of Bias Correction (BC) methods:

• Contains accessible information about the why and how of bias correction for climate data
• Is a software library for for the application of BC methods (see our full pipeline for bias-correction of the ground-breaking local-scale (2.2km) Convection Permitting Model (CPM). clim-recal brings together different software packages in python and R that implement a variety of bias correction methods, making it easy to apply them to data and compare their outputs.
• Was developed in partnership with the MetOffice to ensure the propriety, quality, and usability of our work
• Provides a framework for open additions of new software libraries/bias correction methods (in planning)

Table of Contents

1. Overview: Bias Correction Pipeline
2. Documentation
3. The Datasets
4. Why Bias Correction?
5. Contributing
6. Future Plans
7. License

Overview: Bias Correction Pipeline

clim-recal is a debiasing pipeline, with the following steps:

1. Set-up & data download We provide custom scripts to facilitate download of data
2. Preprocessing This includes reprojecting, resampling & splitting the data prior to bias correction
3. Apply bias correction Our pipeline embeds two distinct methods of bias correction
4. Assess the debiased data We have developed a way to assess the quality of the debiasing step across multiple alternative methods

For a quick start on bias correction, refer to our comprehensive analysis pipeline guide.

Documentation

We are in the process of developing comprehensive documentation for our code base to supplement the guidance provided in this and other README.md files. In the interim, there is documentation available in the following forms:

• Comments within R scripts
• Command line --help documentation for some of our python scripts
• python function and class docstrings
• Local render of documentation via quarto

R

For R scripts, please refer to contextual information and usage guidelines, and feel free to reach out with any specific queries.

python

For many of our python command line scripts, you can use the --help flag to access a summary of the available options and usage information:

$ python resampling_hads.py --help

usage: resampling_hads.py [-h] --input INPUT [--output OUTPUT] [--grid_data GRID_DATA]

options:
-h, --help            show this help message and exit
--input INPUT         Path where the .nc files to resample is located
--output OUTPUT       Path to save the resampled data data
--grid_data GRID_DATA Path where the .nc file with the grid to resample is located

This will display all available options for the script, including their purposes.

Quarto

We also hope to provide comprehensive documentation via quarto. This is a work in progress, but if you would like to render documentation locally you can do so via quarto and conda:

1. Ensure you have a local installation of conda or anaconda .
2. Checkout a copy of our git repository
3. Create a local conda environment via our environment.yml file. This should install quarto.
4. Activate that environment
5. Run quarto preview.

Below are example bash shell commands to render locally after installing conda:

$ git clone https://github.com/alan-turing-institute/clim-recal
$ cd clim-recal
$ conda create -n clim-recal -f environment.yml
$ conda activate clim-recal
$ quarto preview

We appreciate your patience and encourage you to check back for updates on our ongoing documentation efforts.

The Datasets

UKCP18

The UK Climate Projections 2018 (UKCP18) dataset offers insights into the potential climate changes in the UK. UKCP18 is an advancement of the UKCP09 projections and delivers the latest evaluations of the UK's possible climate alterations in land and marine regions throughout the 21st century. This crucial information aids in future Climate Change Risk Assessments and supports the UK’s adaptation to climate change challenges and opportunities as per the National Adaptation Programme.

HADS

HadUK-Grid is a comprehensive collection of climate data for the UK, compiled from various land surface observations across the country. This data is organized into a uniform grid to ensure consistent coverage throughout the UK at up to 1km x 1km resolution. The dataset, spanning from 1836 to the present, includes a variety of climate variables such as air temperature, precipitation, sunshine, and wind speed, available on daily, monthly, seasonal, and annual timescales.

Geographical Dataset

The geographical dataset can be used for visualising climate data. It mainly includes administrative boundaries published by the Office for National Statistics (ONS). The dataset is sharable under the Open Government Licence v.3.0 and is available for download via this link. We include a copy in the data/Geofiles folder for convenience. In addition, the clips for three cities' boundaries from the same dataset are copied to three.cities subfolder.

Why Bias Correction?

Regional climate models contain systematic errors, or biases in their output [1]. Biases arise in RCMs for a number of reasons, such as the assumptions in the general circulation models (GCMs), and in the downscaling process from GCM to RCM [1,2].

Researchers, policy-makers and other stakeholders wishing to use publicly available RCMs need to consider a range of "bias correction” methods (sometimes referred to as "bias adjustment" or "recalibration"). Bias correction methods offer a means of adjusting the outputs of RCM in a manner that might better reflect future climate change signals whilst preserving the natural and internal variability of climate [2].

Part of the clim-recal project is to review several bias correction methods. This work is ongoing and you can find our initial taxonomy here. When we've completed our literature review, it will be submitted for publication in an open peer-reviewed journal.

Our work is however, just like climate data, intended to be dynamic, and we are in the process of setting up a pipeline for researchers creating new methods of bias correction to be able to submit their methods for inclusion on in the clim-recal repository.

1. Senatore et al., 2022, https://doi.org/10.1016/j.ejrh.2022.101120
2. Ayar et al., 2021, https://doi.org/10.1038/s41598-021-82715-1

Contributing

We hope to bring together the extensive work already undertaken by the climate science community and showcase a range of libraries and techniques. If you have suggestions on the repository, or would like to include a new method (see below) or library, please raise an issue or get in touch!

Adding to the conda environment file

To use R in anaconda you may need to specify the conda-forge channel:

$ conda config --env --add channels conda-forge

Some libraries may be only available through pip, for example, these may require the generation / update of a requirements.txt:

$ pip freeze > requirements.txt

and installing with:

$ pip install -r requirements.txt

Future plans

• More BC Methods: Further bias correction of UKCP18 products. This is planned for a future release and is not available yet.
• Pipeline for adding new methods: This is planned for a future release and is not available yet.