Mekong-Energy-Sediment

Introduction

This repository contains data and code accompanying the manuscript:

Xu et al., Strategizing Renewable Energy Transitions to Preserve Sediment Transport Integrity (under review)

The study investigates optimal renewable energy deployment strategies in the Mekong River basin that balance carbon emission reduction targets with the preservation of critical sediment transport processes.

Requirements

Core Environment

• Python 3.10.16 (Recommended via Miniconda)
• Gurobi Optimizer (with academic/commercial license)

Installation Options

Option 1: Pip

pip install numpy==1.26.4 pandas==2.0.3 scipy==1.11.4 pyomo==6.9.1 xarray==2023.6.0 gurobipy==12.0.1

Option 2: Conda (Recommended)

conda create -n mekong python=3.10.16
conda activate mekong
conda install -c conda-forge numpy=1.26.4 pandas=2.0.3 scipy=1.11.4 pyomo=6.9.1 xarray=2023.6.0
conda install -c gurobi gurobi

Usage

Single Run Execution

Run a single scenario with specified parameters:

python run_mekong_gurobi.py --carbon=<c> --sediment=<s> --limit=<l>

Where:

--carbon (<c>): Carbon emission constraint level (1-4, representing different carbon policy scenarios)

--sediment (<s>): Minimum sediment transport requirement (14-54, with 0.2 increments)

--limit (<l>): Transmission line scenario (1, 2, 4, or 6 representing different infrastructure configurations)

Parallel Execution

For comprehensive parameter space exploration, use the provided parallel execution script:

./parallel.sh

The script implements a parallelized sweep through:

• 4 transmission scenarios (limit)

• 4 carbon policy levels (carbon)

• 201 sediment transport constraints (sediment from 14 to 54 in 0.2 increments)

The parallel.sh script uses GNU parallel to run N=10 concurrent processes for efficient computation.

Reproducing Figures in the Manuscript

We have uploaded all the output files and code used to reproduce the figures in the manuscript in the folder code_data_reproduce_figures.

Please note that some original output data files in NetCDF (.nc) format are large in size, so they have been compressed to meet GitHub's size limitations. You can download and extract them locally to access the full data.

To view and process the .nc files, we recommend using the xarray package.

Reproducing Uncertainty Scenario Results

Before reproducing the results of the scenario analysis, you need to generate the inputs for the uncertainty scenarios first. Since the uncertainty analysis involves 72,900 scenarios, we did not upload all input files. Please follow the steps below to reproduce uncertainty scenario results:

1. Run ./input/generate_uncertainty_input.py to generate the scenario inputs. You can modify the start_no, end_no, and exclude_no parameters in the script to flexibly define the range and specific scenarios you wish to generate.
2. After running the script of generating inputs, you will obtain a series of input files named input_$no.xlsx (where no depends on your specified settings), as well as a summary file named uncertainty_scenarios_$start_no_$end_no.xlsx. The summary file includes a no column, followed by the parameter values corresponding to the candidate inputs in the input/uncertainty_input_data folder.
3. Once the uncertainty inputs are generated, run the following command to simulate one or multiple scenarios:

   python run_mekong_gurobi-uncertainty.py $start_no $end_no

4. To run multiple scenarios in batch mode on a supercomputer, use the uncertainty_script_55289_61964.sh script.
   • This script is designed for Aspire2a, which uses the PBS job scheduler. You will need to modify the script according to your computing platform.
5. After running the scripts, results will be saved under the ./output/Uncertainty directory.

Additional Resources:

• Summary of all inputs: input/uncertainty_all_scenarios.xlsx
• All output data for uncertainty scenarios is available at: https://zenodo.org/records/15679035

Contact

For questions regarding the code or methodology, please contact Zhanwei Liu.