Does the Madden‐Julian Oscillation influence aerosol variability?

Baijun Tian, Duane E. Waliser, Ralph A. Kahn, Qinbin Li, Yuk L. Yung, Tomasz M. Tyranowski, Igor V. Geogdzhayev, Michael I. Mishchenko, Omar Torres, A. SmirnovJournal of Geophysical Research Atmospheres 2008 · DOI

TL;DR

This study investigates how the Madden-Julian Oscillation (MJO), a major driver of tropical weather variability, influences global aerosol patterns. The findings reveal that the MJO significantly modulates aerosol concentrations, with different satellite retrieval methods showing distinct relationships with rainfall, suggesting that cloud contamination is a key factor in these observed patterns.

What Problem Does This Paper Address?

Atmospheric aerosols play a crucial role in Earth's climate and hydrological cycles by interacting with radiation and influencing cloud formation. However, their spatial and temporal distribution is highly heterogeneous and challenging to document comprehensively due to measurement complexities. While the Madden-Julian Oscillation (MJO) is known to strongly influence tropical weather patterns like convection and rainfall, its impact on aerosol variability, particularly on intraseasonal timescales (30-90 days), remained largely unexplored. Understanding this connection is vital for improving aerosol prediction and its role in climate modeling.

Methods

The researchers employed a composite analysis of MJO events to examine aerosol variability. They utilized multiple global satellite aerosol products:

  • TOMS Aerosol Index (AI): This index measures UV-absorbing aerosols and is sensitive to aerosols in the upper troposphere.
  • MODIS Aerosol Optical Thickness (AOT): This measures the total extinction of light by aerosols in the atmospheric column, sensitive to both absorbing and non-absorbing aerosols.
  • AVHRR AOT: Similar to MODIS AOT, this also measures aerosol extinction.

By compositing data around identified MJO events, they analyzed anomalies in aerosol products and rainfall. They also used ground-based AERONET AOT data for validation. The study specifically investigated the relationships between aerosol anomalies and rainfall anomalies, looking for correlations at different phases of the MJO.

Datasets

The study utilized the following datasets:

  • TOMS Aerosol Index (AI): Daily data from Nimbus-7 (January 1980 - December 1992).
  • MODIS Aerosol Optical Thickness (AOT): L3 MOD08 data product, Version 4, at 0.55 µm on 1°×1° spatial grids (February 2000 - December 2005).
  • AVHRR AOT: Global Aerosol Climatology Project (GACP)/AVHRR product, daily mean AOT at 0.55 µm on 1°×1° grids (January 1982 - June 2005).
  • CMAP Rainfall Data: NOAA CPC Merged Analysis of Precipitation (CMAP) pentad data on a 2.5°×2.5° grid (January 1979 - May 2006) to identify MJO events.
  • AERONET AOT: Ground-based data from Kaashidhoo and Nauru sites for validation.

Key Insights and Ideas

  • MJO's Broad Influence: The MJO isn't just a weather phenomenon; it demonstrably influences aerosol distribution across large regions of the tropics and subtropics.
  • Satellite Retrieval Differences Matter: The distinct relationships observed between TOMS AI and rainfall versus MODIS/AVHRR AOT and rainfall highlight that how aerosols are measured significantly impacts the perceived relationship with meteorological phenomena.
  • Cloud Contamination is a Major Player: The study strongly suggests that cloud contamination in satellite aerosol retrievals is a primary driver of the observed aerosol-rainfall relationships, particularly the positive correlation seen with AOT.
  • Aerosol-Rainfall Relationships are Complex: The paper moves beyond simple correlations to discuss potential physical mechanisms (e.g., aerosol growth, wet deposition, wind effects) and retrieval artifacts, underscoring the multifaceted nature of these interactions.

Key Results

  • Spatial Patterns: Significant MJO-associated aerosol anomalies were found in the tropical Indian and western Pacific Oceans (where MJO convection is active) and in tropical Africa and the Atlantic Ocean (where convection is weaker but background aerosols are high).
  • TOMS AI vs. Rainfall: A strong inverse linear relationship was observed between TOMS AI anomalies and rainfall anomalies. This means higher rainfall was associated with lower TOMS AI, and vice versa.
  • MODIS/AVHRR AOT vs. Rainfall: A weaker, less coherent positive correlation was found between MODIS/AVHRR AOT anomalies and rainfall anomalies. Higher rainfall was generally associated with higher AOT.
  • AERONET Validation: Ground-based AERONET data from Kaashidhoo and Nauru showed a strong positive correlation with rainfall, aligning with the MODIS/AVHRR AOT results and supporting the idea that cloud contamination might be influencing these satellite retrievals.
  • Cloud Contamination Hypothesis: Further analysis using a more stringent cloud-screening method for AERONET data showed a lack of coherent positive correlation between AOT and rainfall, strongly supporting the hypothesis that cloud contamination is a significant contributor to the observed positive AOT-rainfall relationship.

Limitations and Future Work

  • Cloud Contamination: The authors acknowledge that cloud contamination is a major contributor to the observed aerosol-rainfall relationships, particularly the positive correlation seen with MODIS/AVHRR AOT. While they present preliminary analysis suggesting its importance, they state that other physical mechanisms cannot be entirely excluded.
  • Physical Mechanisms: While several physical mechanisms (aerosol humidification, wet deposition, wind effects, biological production) are discussed as potential contributors, their exact roles and relative importance in explaining the observed relationships remain unclear.
  • Future Research: The paper emphasizes the need for further research to fully understand these complex aerosol-rainfall relationships. This will likely require a synergistic approach combining multisensor satellite data, in situ observations, targeted aircraft campaigns, and advanced chemistry-transport modeling. The potential for MJO-based prediction of aerosols and air quality, given the MJO's predictability, is highlighted as a key motivation for this future work.

Summary updated

Abstract

We investigate the modulation of aerosols by the Madden‐Julian Oscillation (MJO) using multiple, global satellite aerosol products: aerosol index (AI) from the Total Ozone Mapping Spectrometer (TOMS) on Nimbus‐7, and aerosol optical thickness (AOT) from the Moderate Resolution Imaging Spectroradiometer (MODIS) on Terra and Aqua and the Advanced Very High Resolution Radiometer (AVHRR) on NOAA satellites. A composite MJO analysis indicates that large variations in the TOMS AI and MODIS/AVHRR AOT are found over the equatorial Indian and western Pacific Oceans where MJO convection is active, as well as the tropical Africa and Atlantic Ocean where MJO convection is weak but the background aerosol level is high. A strong inverse linear relationship between the TOMS AI and rainfall anomalies, but a weaker, less coherent positive correlation between the MODIS/AVHRR AOT and rainfall anomalies, were found. The MODIS/AVHRR pattern is consistent with ground‐based Aerosol Robotic Network data. These results indicate that the MJO and its associated cloudiness, rainfall, and circulation variability systematically influence the variability in remote sensing aerosol retrieval results. Several physical and retrieval algorithmic factors that may contribute to the observed aerosol‐rainfall relationships are discussed. Preliminary analysis indicates that cloud contamination in the aerosol retrievals is likely to be a major contributor to the observed relationships, although we cannot exclude possible contributions from other physical mechanisms. Future research is needed to fully understand these complex aerosol‐rainfall relationships.
BibTeX
@article{tian2008does,
  author = {Baijun Tian and Duane E. Waliser and Ralph A. Kahn and Qinbin Li and Yuk L. Yung and Tomasz M. Tyranowski and Igor V. Geogdzhayev and Michael I. Mishchenko and Omar Torres and A. Smirnov},
  title = {{Does the Madden‐Julian Oscillation influence aerosol variability?}},
  journal = {Journal of Geophysical Research Atmospheres},
  year = {2008},
  doi = {10.1029/2007jd009372},
}
Figure