Spatial source contribution and interannual variation in deposition of dust aerosols over the Chinese Loess Plateau

This is the workflow used to generate the figures published in AGU: Spatial source contribution and interannual variation in deposition of dust aerosols over the Chinese Loess Plateau The workflow is customized for the NIRD toolkit enivorment, and uses absolute path for the data. Therefore these need to be updated if the workflow is to be run on other machine where paths to the data files could be different, if note the docker container is used. The paths used in the workflow is configure in the Master_thesis_UiO_workflow/config directory.

The computing enviroment can be installed using conda, but there is also a docker enviroment which can be used. The easiest would be to use the docker container, data files are available in this zenodo repository. The workflow can configured to work on any computing system you the same thesisdocker container if the path are updated.

Setting up computing evniroment.

To pull the docker image:

docker pull thesisdocker:latest

Download the data from the Zenodo repository.

Create a volume for the home directory:

docker volume create home

To Start the container, but rember to change the path to where the Zenodo data is stored.

docker run -it -p 8888:8888 -v home:/home/notebook -v /<path>/<to>/>data-folder>/data:/home/notebook/data thesisdocker:latest

The docker container will start a juptyer lab instance with all conda enviroment and packages installed.

Running the workflow

To run the workflow simply within the thesis docker container enviroment or on nird toolkit, simply activite the enviroment by source activate dust. Then run the workflow you need to be in the AGU_JGR_CLP_SOURCE_WORKFLOW/Master_thesis_UiO_workflow directory, the first step of runnning the workflow is done by first checking if all the files are there:

snakemake -n plot_all_agu 

This will excute a dry run, if thtis work without any errors then the paths have been set correctly. Running the workflow is then just a matter of typing

snakemake -j2 plot_all_agu --rerun-triggers mtime

The -j2 flag request 2 processors. The --rerun-triggers mtime can be added avoid generating all the intermidiate files for making the plots.

Some of the figure are generated by notebooks which is not contained in the workflow AGU_JGR_CLP_SOURCE_WORKFLOW/notebooks