Cannon et al. (2025) CanCPLD: Convective Parameters and Lightning Data to Support Future Thunderstorm Projections in North America

Identification

• Journal: Scientific Data
• Year: 2025
• Date: 2025-10-10
• Authors: Alex J. Cannon, Karen E. Ramsey, Alessio C. Spassiani
• DOI: 10.1038/s41597-025-05924-7

Research Groups

• Climate Research Division, Environment and Climate Change Canada, Victoria, British Columbia, Canada
• Department of Physics & Astronomy, University of British Columbia, Vancouver, British Columbia, Canada
• Climate Research Division, Environment and Climate Change Canada, Toronto, Ontario, Canada

Short Summary

This paper introduces CanCPLD, a comprehensive multi-decade dataset of lightning flash totals and 201 convective parameters for North America, compiled from reanalysis and high-resolution climate model simulations. The dataset aims to support research on future thunderstorm projections by providing precomputed environmental conditions, thereby reducing computational barriers for researchers.

Objective

• To compile and provide a comprehensive, multi-decade dataset of lightning flash totals and convective parameters for North America to support studies on the future evolution of thunderstorms, particularly in high-latitude regions, under various global warming scenarios.

Study Configuration

• Spatial Scale:
  • North America (15°N to 90°N latitude, 180°W to 40°W longitude).
  • Lightning data (CLDN): North of 40°N latitude on a 0.1° × 0.1° grid (grid cell area approximately 9.5 × 10^7 m² at 40°N and 6.2 × 10^7 m² at 60°N).
  • Convective parameters (ERA5, HighResMIP): All of North America on a 0.25° × 0.25° grid (approximately 31,000 meters).
  • Climate model grid spacings: Ranging from approximately 0.19° × 0.19° (MRI-AGCM3-2-S) to 0.56° × 0.56° (MRI-AGCM3-2-H).
  • Vertical resolution: 18 pressure levels (surface to 10,000 Pa) for ERA5 and some HighResMIP models; 8 pressure levels (surface to 5,000 Pa) for other HighResMIP models.
• Temporal Scale:
  • Historical period: 1998–2024 for CLDN lightning data and ERA5 reanalysis.
  • Future projections: Four 20-year periods from HighResMIP models, corresponding to global warming levels (GWLs) of +1°C (2001-2020, historical reference), +2°C, +3°C, and +4°C above preindustrial levels.
  • Temporal resolution: 3-hourly for CLDN and ERA5; 6-hourly for HighResMIP simulations.

Methodology and Data

• Models used:
  • HighResMIP global climate models (CMIP6): CMCC-CM2-VHR4 (AOGCM), EC-Earth3P-HR (AOGCM/AGCM), MRI-AGCM3-2-H (AGCM), MRI-AGCM3-2-S (AGCM).
  • thundeR R package (v1.1.3): Used for calculating 201 convective parameters from 3D atmospheric state variables.
  • Logistic Regression: Employed to demonstrate the link between convective parameters and lightning occurrence (e.g., EB21-LR model).
• Data sources:
  • Observation: Canadian Lightning Detection Network (CLDN) for cloud-to-ground (CG) and intra-cloud/cloud-to-cloud (CC) lightning flash totals.
  • Reanalysis: European Centre for Medium-Range Weather Forecasts reanalysis version 5 (ERA5) for historical 3D atmospheric state variables.
  • Climate Model Simulations: HighResMIP simulations from the Earth System Grid Federation (ESGF) climate data archive for historical and future 3D atmospheric state variables under SSP5-8.5 scenario.

Main Results

• CanCPLD Dataset Compilation: A comprehensive dataset (CanCPLD) was compiled, including 3-hourly lightning flash totals (1998-2024) and 201 convective parameters derived from ERA5 (1998-2024) and HighResMIP climate models (20-year periods for +1°C, +2°C, +3°C, +4°C GWLs).
• Technical Validation of Lightning Data: Replicated previous studies, showing consistent spatial distribution of CG flash density, extreme daily lightning activity, and annual/monthly flash counts over Canada. CG flashes are recommended for climatological analyses due to stable detection efficiency (>90%).
• Technical Validation of ERA5 Convective Parameters: Parameters calculated using a reduced set of 17 pressure levels (surface to 10,000 Pa) are highly correlated (Pearson r ≥ 0.95 for most parameters) and unbiased compared to those calculated with all 37 ERA5 levels, with minor episodic deviations for MUCIN and MULFC_TEMP.
• Technical Validation of HighResMIP Convective Parameters: HighResMIP models capture broad spatial patterns of convective parameters relative to ERA5 for the historical period (2001-2020). EC-Earth3P-HR models showed the highest performance, while helicity fields exhibited greater relative error and lower spatial correlation, partly due to coarser vertical sampling in some models.
• Future Projections: HighResMIP models consistently project increases in instability (MUCAPE) and stronger convective inhibition (MUCIN) across North American land areas under +2°C, +3°C, and +4°C GWLs. Humidity (RH05km) increases and lifting condensation level height (MULCLHGT) lowers for MRI-AGCM3-2 variants. Deep wind shear (BS06km) and storm-relative helicity (SRH_3km) generally decrease with warming, though with considerable inter-model spread.
• Link between Parameters and Lightning: A logistic regression model trained on ERA5 convective parameters and CLDN lightning data demonstrated a high spatial correlation (0.86) in predicting lightning occurrence, confirming the utility of CanCPLD for statistical modeling.

Contributions

• Novel Dataset Creation: Provides CanCPLD, a unique, comprehensive, and precomputed multi-decade dataset of lightning and convective parameters for North America, specifically designed to support future thunderstorm projections.
• Reduced Computational Barrier: By distributing precomputed convective parameters, the dataset significantly lowers the data transfer and computational burden for researchers, facilitating broader access to high-resolution data for thunderstorm research.
• Consistent Framework for Climate Change Studies: Integrates high-resolution reanalysis (ERA5) and HighResMIP climate model simulations with observed lightning data, offering a consistent spatiotemporal framework for assessing convective storm environments under historical and future climate scenarios.
• Focus on High Latitudes: Addresses the gap in studies focusing on high latitudes of North America, a region experiencing rapid warming and potential expansion of convective environments.
• Technical Validation: Provides thorough technical validation of the dataset components, including the impact of reduced vertical levels and inter-model comparisons, guiding users on data suitability and limitations.

Funding

• Environment and Climate Change Canada (ECCC)
• World Climate Research Programme (through its Working Group on Coupled Modelling for CMIP6)
• Earth System Grid Federation (ESGF)
• Multiple funding agencies supporting CMIP6 and ESGF

Citation

@article{Cannon2025CanCPLD,
  author = {Cannon, Alex J. and Ramsey, Karen E. and Spassiani, Alessio C.},
  title = {CanCPLD: Convective Parameters and Lightning Data to Support Future Thunderstorm Projections in North America},
  journal = {Scientific Data},
  year = {2025},
  doi = {10.1038/s41597-025-05924-7},
  url = {https://doi.org/10.1038/s41597-025-05924-7}
}