Anthony Davis
Document Actions
Anthony Davis joined LANL's Space and Remote Sensing Sciences Group (ISR-2) in 1997 after a 5-year tenure at NASA's Goddard Space Flight Center, Climate and Radiation Branch. PhD in Physics at McGill University on radiative transfer in highly variable multifractal media in 1992; MSc at Université de Montréal and undergraduate at Université Pierre et Marie Curie in Astronomy/Astrophysics. His research background is in steady-state and time-dependent three-dimensional radiative transfer theory and mathematical geophysics (primarily statistical sampling problems in scale-by-scale data analysis, and wavelet/fractal-based stochastic modeling). He applies this expertise to optical remote sensing of cloud and surface properties using both active and passive techniques, driven both by climate modeling needs and national security concerns. Anthony's community engagement includes regular reviewing of manuscripts and proposals, organization of meetings on specialized topics within the framework of the American Geophysical Union (AGU) meetings technical committees or working groups at AGU, American Meteorological Society (AMS), in the International Radiation Commission (IRC) of the International Association of Meteorology and Atmospheric Physics (IAMAP) and DOE's Atmospheric Radiation Measurement (ARM) Program.
In the LDRD project on the Aerosol-Cloud-Radiation-Climate Puzzle, Anthony works on both modeling and observation thrusts. In both cases, he applies his expertise in 3D radiative transfer (RT). The dynamical cloud modeling is done with HiGrad in collaboration with Jon Reisner and Scott Smith and it is primarily to quantify the indirect effect of the aerosol on the climate system, especially when the nature of the aerosol is varied, e.g., from naturally occurring to (say) black carbon. The cloud remote sensing studies are in collaboration with Petr Chylek and are designed to put observational constraints on the HiGrad-based cloud models.
In support of the cloud process modeling, we are leveraging results from an ongoing project funded by the DOE Atmospheric Radiation Measurement (ARM) program in efficient 3D RT for estimation of local heating and cooling rates. This means that we allow for the horizontal fluxes that are explicitly neglected in standard 1D models. By "efficient" it is meant that some approximation technique is invoked that yields accurate enough fluxes for the purposes of the dynamical cloud model, i.e., removing most of the bias associated with the assumption of 1D photon transport. With respect to full (FLOP-hungry) 3D RT, this is the analog of using the analytical two-stream approximation of Schuster (1905) rather than a full discrete-ordinate method of Chandrasekhar-Wick for plane-parallel slabs.
In support of the cloud observation effort, there is a need for capability to predict radiances in 3D RT at the high spatial resolution (5-20 m) of the imagery from the DOE Multispectral Thermal Imager (MTI). At such high resolution, the pixels are indeed radiatively coupled through horizontal photon flow driven by gradients in opacity. This is a significant challenge in comparison with the modeling project which requires of 3D RT only fluxes. Fluxes are angularly integrated radiances. However, these fluxes are required everywhere in the 3D grid while only the 2D distribution of boundary radiances are of immediate interest in remote sensing. Spectrally, the model-driven fluxes need to be integrated across the solar and thermal spectra while in remote sensing wavelength is for all practical purposes discretely sampled.
Approximation techniques such as 3D diffusion theory or adjoint perturbation theory are applied to the heating/cooling rate problem. In contrast, the radiance prediction problem in cloud remote sensing calls for an accurate (but slow) approach such as the Monte Carlo method.
