"This project, AEROMAP, is designed to produce accurate, daily-updated, global atmospheric aerosol maps. To achieve this, AEROMAP will use the daily, full-Earth coverage of the aerosol optical depth (AOD) provided by the MODerate resolution Imaging Spectrometer (MODIS) satellite instrument in 3 wavelength bands (438-448nm, 673-683nm, 862-877nm), to extrapolate local, ground-based retrievals of aerosol microphysical properties (AMP) from AEROsol robotic NETwork (AERONET) stations to the entire Earth-surface. Extrapolation will be achieved with multiple-input multiple-output universal function-approximating artificial neural networks that will be trained on AERONET data (the AOD at 440nm, 675nm and 870nm) as input and AERONET-AMP retrievals (the aerosol size distribution, the complex refractive index, the effective radius, and the single scattering albedo) as output, in order to learn the inversion function. A portion of the training dataset will be reserved for validation. In the second step, MODIS broad-band AOD data that is spatially co-located with AERONET single-wavelength AOD data will be used to train a second neural network to learn the mathematical relationship between the broad-band and single-wavelength measurements. Then, the worldwide, daily coverage provided by MODIS will be used extrapolate and retrieve the sought-after AMP worldwide. Modern statistical methods of cluster analysis will be used to classify aerosol type regions and global maps of AMP will be used to provide a new full-Earth (near) real-time monitor to globally characterise atmospheric aerosols. Aerosol dispersal during selected events will test the validity of the monitor and a project website/portal will provide users access to global AMP maps, raw data and early-warning alerts of extreme aerosol conditions worldwide."
Field of science
- /natural sciences/computer and information sciences/artificial intelligence/computational intelligence
Call for proposal
See other projects for this call