Dr. Zhibo Zhang, Department of Physics.
Desert dust is abundant and recognized as an integral component of the Earth system that influences weather and climate via a suite of complex interactions with the energy, water, and carbon cycles. Dust storms cause detrimental losses of human life and economic activities through degrading air quality, spreading diseases, disrupting transportation, and reducing efficiency of solar power generation. There has been growing attention in the past decades to advancing the research of dust cycle – a chain of processes involving emissions, transport, transformation, and deposition, and its tight coupling with other bio-geo-chemical components of the Earth system and human dimension. However, many knowledge gaps remain.
Dust varies greatly across a wide range of temporal and spatial scales, resulting from a combination of sporadic nature of dust events, large variability of transport and removal processes, and the tight and complex coupling of dust with other components of the Earth system. Although data assimilation has been widely used in the satellite era to impose strong constraint on aerosol optical depth (AOD), dust is at most weakly constrained due to large uncertainty in model simulations of aerosol components and vertical distribution. Dust deposition from the reanalysis would suffer from a mass imbalance issue. It is thus critical to develop a comprehensive, remote-sensing observations based, and self-consistent global dust data record over multi-decadal time scales.
Built upon the team’s extensive experience in satellite remote sensing, machine learning (ML), aerosol modeling, and integrated data analysis, we propose to create a Comprehensive and Augmented Multi-decadal Remote-sensing Observations of Dust (CAMRO-Dust) data record by integrating multiple satellites and developing novel approaches. Satellite sensors include MODIS, VIIRS, and CALIOP. Specifically, we will perform the following major tasks towards the creation of dynamically consistent data sets of dust:
1) Improving and enhancing aerosol retrievals in both the mid-visible and thermal infrared wavelengths over global oceans from MODIS and VIIRS through implementing a ML-based dust detection algorithm and accounting for dust non-sphericity.
2) Deriving global dust optical depth (DOD) from MODIS and VIIRS retrievals by separating dust from other components based on size and absorption of dust particles.
3) Producing global 3-D distributions of fine and coarse dust extinction based on CALIOP observations of aerosol backscatter and depolarization ratio.
4) Estimating dust deposition fluxes into oceans by using the 3-D dust distributions from satellites.
5) Developing and validating a dust PM2.5 data record for accurately assessing the health impacts of dust.
With these concerted efforts we will develop a comprehensive, coherent, and multi-decadal data record of global dust for broad earth science research and applications. We will analyze and address data error and uncertainties. This unique dataset will be hosted at and distributed by one of NASA Distributed Active Archive Centers (DAACs) designated by the program management to the research community and other stakeholders, along with data readers and adequate documentation of retrieval methods, data production, and product quality. The so-produced dynamically consistent data record of dust can be used to advance the observational understandings of the interannual variability and trend of dust over recent decades, dust direct radiative effects on both solar and terrestrial radiation, air quality and health impacts of dust, and dust fertilizing effect on oceanic ecosystems. The comprehensive dust data set can also be used to systematically evaluate global and regional air quality, climate, and earth system models and effectively guide the reduction of modeling uncertainties.