Using Singularity in Local (Approach 4) for the reproducibility of SUMMA modeling

Instruction

1. Install Singularity in Ubuntu os.

• Installation description: https://sylabs.io/guides/3.5/admin-guide/installation.html#installation-on-linux

2. Create a Definition file to create a Singularity image

• Use a SUMMA and pySUMMA definition file in this GitHub
• https://github.com/DavidChoi76/SUMMA_Singularity_In_Rivanna/blob/master/summa_singularity.def

3. Create the Singularity image

sudo singularity build summa3_singularity.sif summa_singularity.def

• Sinuglarity image in HydroShare: https://www.hydroshare.org/resource/6e8b3991a6dc46ba97c00688f7cb67de/

4. You can find summa_singularity.sif singularity image in the HS resource and download it in your local environment.

5. Download kernel.json file from this GitHub to your local environment

• https://github.com/DavidChoi76/SUMMA_Singularity_In_Rivanna/blob/master/kernel.json

6. Create SUMMA Jupyter Kernel

mkdir -p ~/.local/share/jupyter/kernels/summa
mv kernel.json ~/.local/share/jupyter/kernels/summa

7. Start JupyterLab and select summa Kernel

If you don't want to create a Singularity image, you can use kernel_copy.json in this GitHub

• https://github.com/DavidChoi76/SUMMA_Singularity_In_Rivanna/blob/master/kernel_copy.json

1. Download kernel_copy.json in your local computer.

2. Change the name of kernel_copy.json to kernel.json

3. After that, follow step 6 above