Code related to "Climate Change Will Increase High Temperature Risks, Degradation, and Costs of Rooftop Photovoltaics Globally"

# SolarPVDegradationClimateChange This is the Code repository for publication on accelerating degradation of rooftop solar photovoltaics under climate change. Download the dateset through the link as described in 'data section' (not here due to size of 140 **Terabytes (TB)** ). Key contributors: Haochi Wu, Michael Craig at the University of Michigan, & Qinqin Kong at Purdue Univerisity. Updated: 7/23/2024 HW Contact: [e-mail] (Haochi Wu)  ## Python Environment The python environment is deployed in Linux and pipeline managed using slurm computing cluster platform at Rosen Center for Advanced Computing at Purdue University and Advanced Research Computing at the University of Michigan. ## calculation code The calculation code in `/calculation` folder contains core process of the research, including: - **01_slurm_command.py** - Slurm command script for submitting batch jobs. - **01_Slurm_PVTemp_Vectorized_Calculation_CMIP6_GCMs.py** - Python script for vectorized calculation of PV temperature using CMIP6 GCMs data. - Accompanied by `01_Slurm_PVTemp_Vectorized_Calculation_CMIP6_GCMs.sbat` for Slurm job submission. - **02_slurm_command.py** - Second Slurm command script for batch jobs. - **02_Slurm_PVTemp_Vectorized_Calculation_ERA5.py** - Python script for vectorized calculation of PV temperature using ERA5 data. - Accompanied by `02_Slurm_PVTemp_Vectorized_Calculation_ERA5.sbat` for Slurm job submission. - **03_slurm_command.py** - Third Slurm command script for batch jobs. - **03_Slurm_T98_Calculation_ERA5.py** - Python script for T98 temperature calculation using ERA5 data. - Accompanied by `03_Slurm_T98_Calculation_ERA5.sbat` for Slurm job submission. - **04_Create_Climatology_and_Abnormal_Signal_CMIP6_GCM.py** - Script to create climatology and identify abnormal signals using CMIP6 GCM data. - **05_Add_Abnormal_Signal_to_ERA5.py** - Script to add identified abnormal signals to ERA5 data. - **06_Plot_T98_for_ERA5_add_CMIP6_GCMs.py** - Script for plotting T98 temperatures combining ERA5 and CMIP6 GCMs data. - **07_calculate_LCOE_Changes.py** - Script to calculate changes in Levelized Cost of Electricity (LCOE). - **tas_abnormal_from_qin.csv** - CSV file containing abnormal temperature data for 20 CMIP6 GCMs. ## visualization code The visualization code in `/visualization` folder contains visualization code for our research, including: - **Capacity_under_risk.ipynb** - **Increased_risk_area.ipynb** - **Regional_T98_map.ipynb** - **Roofpotential_capacity.ipynb** And more relevant pre/post processing code and visualization script is coming soon. ## Data Availability CMIP6 GCMs temperature and energy generation dataset: https://transfer.rcac.purdue.edu/file-manager?origin_id=e7eecae4-ab46-4016-b8f8-ba0379421b0b&origin_path=%2F CMIP6 temperature dataset downscaled and bias-corrected by ERA5 dataset: https://transfer.rcac.purdue.edu/file-manager?origin_id=64916c8c-1c24-4716-ba90-a06fc412e778&origin_path=%2F The dataset for future rooftop PV temperature is about **130 Terabytes (TB)** in size and is freely available at Purdue Fortress long-term archive, accessible via Globus: ## Data Description ### Rooftop PV under CMIP6 GCMs Climate Data: The file size for the CMIP6 GCMs temperature raw series and energy generation series ranges from 0.3GB to 1.5GB per file. We have stored data for the years 1950 to 2100, covering 150 years for 20 different GCMs. total of 150 * 20 = 3,080 years of data. ### Rooftop PV under CMIP6 GCMs Climate with ERA5 Bias-Corrected Data: The size of the temperature files for CMIP6 after bias correction is approximately 33GB. This dataset includes hourly temperature data for 27 ERA5 sample years, 20 GCM models, and 7 warming targets, resulting in a total of 7 * 20 * 27 = 3,780 years of data. ### Derived 98 quantile temperature (T98) Data: Based on the above dataset, we have calculated the typical T98 temperature. The global map data is provided at a resolution of 0.25° * 0.25°, with a file size of approximately 7MB. with total of 7 * 20 * 27 = 3,780 years of data. ### the 20 CMIP6 GCMs are listed below: | Index | Model Name | |-------|-------------------------| | 1 | ACCESS-CM2 | | 2 | BCC-CSM2-MR | | 3 | CanESM5 | | 4 | CMCC-CM2-SR5 | | 5 | CMCC-ESM2 | | 6 | CNRM-CM6-1 | | 7 | EC-Earth3_r1i1p1f1 | | 8 | EC-Earth3_r3i1p1f1 | | 9 | EC-Earth3_r4i1p1f1 | | 10 | GFDL-ESM4 | | 11 | HadGEM3-GC31-LL | | 12 | HadGEM3-GC31-MM | | 13 | KACE-1-0-G | | 14 | KIOST-ESM | | 15 | MIROC-ES2L | | 16 | MIROC6 | | 17 | MPI-ESM1-2-HR_r1i1p1f1 | | 18 | MPI-ESM1-2-HR_r2i1p1f1 | | 19 | MPI-ESM1-2-LR | | 20 | MRI-ESM2-0 |