Shikwambana et al. (2026) Analyzing the Effect of the 2015/16 Catastrophic El Niño Event on Wildfire Emissions in Southern Africa Using Lagged Correlation and Interrupted Time-Series Causal Impact Technique

Identification

• Journal: Earth
• Year: 2026
• Date: 2026-03-06
• Authors: Lerato Shikwambana, Kganyago Mahlatse, Xi Zhang
• DOI: 10.3390/earth7020042

Research Groups

• Earth Observation Directorate, South African National Space Agency, Pretoria, South Africa
• Unit for Environmental Sciences and Management, School of Geo- and Spatial Science, North-West University, Potchefstroom, South Africa
• Department of Geography, Environmental Management & Energy Studies, University of Johannesburg, Johannesburg, South Africa
• National Engineering Research Center of Geographic Information System, China University of Geosciences, Wuhan, China

Short Summary

This study analyzed the impacts of the 2015/16 El Niño on Southern African wildfire emissions, vegetation, and meteorological conditions, revealing that the event significantly amplified fire emissions and degraded ecosystem functioning through strong climate-fire-vegetation feedback.

Objective

• To analyze the lagged correlations and causal impact of the 2015/16 El Niño on meteorological parameters and wildfire-related emissions in Southern Africa.

Study Configuration

• Spatial Scale: Southern Africa, defined by 7.71° S to 45.16° S latitude and 36.00° W to 0.38° E longitude.
• Temporal Scale: 2010–2017 for time-series analysis; pre-El Niño period: 2010–2014; post-El Niño period: 2015–2017.

Methodology and Data

• Models used:
  • Time-lagged cross-correlation analysis (Pearson r correlation)
  • Bayesian causal impact analysis (using Interrupted Time-Series (ITS) technique via CausalPy version 0.7.0 Python package)
  • Exploratory analysis (descriptive statistics, boxplots, Cohen’s d effect sizes, percentage change in standard deviation)
• Data sources:
  • Satellite observations (Moderate Resolution Imaging Spectroradiometer - MODIS): Normalized Difference Vegetation Index (NDVI), Canopy Water (CW), Burned Area (BA), Aerosol Optical Depth (AOD), Particulate Matter (PM2.5).
  • Reanalysis data (Modern-Era Retrospective Analysis for Research and Applications, version 2 - MERRA-2): Black Carbon (BC), Organic Carbon (OC), Carbon Monoxide (CO), Sulfur Dioxide (SO2), Sulfates (SO4), Precipitation (PCPN), Soil Moisture (SM), Evapotranspiration (ET), Temperature (Temp), wind speed, wind direction.

Main Results

• Vegetation greenness (NDVI) shows strong positive contemporaneous correlations with soil moisture (r = 0.64, lag = 0) and canopy water (r = 0.59, lag = 0).
• Precipitation (PCPN) exhibits negative correlations with PM2.5 (r = -0.37, lag = -1 month), ET (r = -0.57, lag = -1 month), and AOD (r = -0.29, lag = -2 months), indicating delayed aerosol scavenging effects.
• Carbonaceous aerosols (OC, BC), PM2.5, and AOD are strongly synchronously coupled (e.g., OC-BC r = 0.98, lag = 0; AOD-PM2.5 r = 0.90, lag = 0), confirming biomass burning as a dominant source.
• The 2015/16 El Niño caused statistically significant and sustained post-2015 increases in fire-related emissions (CO, BC, OC, PM2.5, AOD), particularly during austral winter and dry seasons.
• Cumulative causal impacts show increases: approximately 7.5 standardized kilograms per cubic meter for CO, 4 standardized micrograms per cubic meter for BC, 7 standardized micrograms per cubic meter for OC, 15 standardized micrograms per cubic meter for PM2.5, and 15 (unitless) for AOD.
• Burned area (BA) showed a steady but relatively marginal cumulative increase, reaching approximately 3 standardized square kilometers by the end of 2017.
• Precipitation, soil moisture, evapotranspiration, and vegetation greenness displayed persistent negative anomalies, reflecting widespread drought stress and reduced ecosystem functioning.
• Temperature showed a positive anomaly post-2015, with a Cohen's d effect size marginally greater than 0.2.
• Aerosol-related parameters (AOD, OC, PM2.5, BC) exhibited the largest increase in volatility (approximately 7–17%) and positive Cohen’s d effect sizes, indicating both greater variability and an upward shift in mean conditions.
• Sulfur dioxide (SO2) and soil moisture (SM) showed the highest declines in volatility (approximately -10%).

Contributions

• First study to comprehensively evaluate both lagged correlations and causal impacts of an El Niño event on wildfire emissions and associated meteorological and vegetation dynamics in Southern Africa.
• Utilizes a robust dual approach combining time-lagged cross-correlation and Bayesian interrupted time-series (ITS) causal impact analysis to provide a multi-dimensional understanding of extreme ENSO event impacts.
• Demonstrates that the 2015/16 El Niño intensified combustion efficiency and aerosol emissions, rather than solely expanding burned area, leading to significant air quality degradation.
• Offers critical insights for developing mitigation planning, early warning systems, and long-term climate adaptation strategies in Southern Africa.

Funding

• National Research Foundation of South Africa (Ref Number: BRIC231103160523)
• South African National Space Agency (SANSA) (for Article Processing Charge)
• Natural Science Foundation of China (42461144214)

Citation

@article{Shikwambana2026Analyzing,
  author = {Shikwambana, Lerato and Mahlatse, Kganyago and Zhang, Xi},
  title = {Analyzing the Effect of the 2015/16 Catastrophic El Niño Event on Wildfire Emissions in Southern Africa Using Lagged Correlation and Interrupted Time-Series Causal Impact Technique},
  journal = {Earth},
  year = {2026},
  doi = {10.3390/earth7020042},
  url = {https://doi.org/10.3390/earth7020042}
}