Senior Scientist
Space Science and Engineering Center
University of Wisconsin-Madison
1225 West Dayton St.

E-mail: babaum at
Fax: (608) 262-5974
Phone: (608) 263-3898


Active projects:

NASA Suomi-NPP Atmosphere Team

Data Fusion Team: Constructing high spatial resolution IR absorption bands for VIIRS based on fusion of VIIRS and CrIS data

I am serving as both a PI on a cloud retrieval project and the Atmospheres Discipline Lead for the NASA Suomi NPP Program. My own team's immediate science goal is as follows. MODIS has three bands sensitive to carbon dioxide around 4.3 µm  and another four around 15 µm, 2 bands sensitive to water vapor near 6.7 µm, and a band sensitive to ozone near 9.7 µm. VIIRS has none of these IR bands affected by atmospheric absorption, which results in a degradation of the accuracy of the cloud products (cloud mask, cloud top pressure/height and thermodynamic phase) and the moisture products (total precipitable water vapor, upper tropospheric humidity). The lack of at least one IR absorption channel on VIIRS degrades the accuracy of the cloud mask, cloud top pressure/height, and cloud thermodynamic phase products. However, the suite of MODIS IR absorption bands has a variety of other uses in both operational products (such as total precipitable water, or TPW) and applications such as volcanic plume tracking, inference of polar winds, and more.

Our solution to this problem: we developed an innovative way to construct all of the MODIS IR absorption bands (i.e., bands 23, 24, 25, 27, 28, 30, and 33-36) at the imager's nominal high spatial resolution using a data fusion technique involving an imager (such as MODIS, VIIRS, or AVHRR) and an interferometer (such as AIRS, CrIS, or IASI). While our original goal was to construct a 13.3-µm carbon dioxide band) for VIIRS, we recently found out during the course of our work that we could also construct other IR bands too. While the constructed fusion band radiances are not quite as accurate as actual measurements, the radiance differences are within 1% and may be adequate for improving products and applications. For further description, please visit our project site on data fusion.


PyroCumulonimbus (PyroCb) Blog

When a wildfire complex becomes pyroconvective, the smoke plume can reach the upper troposphere and directly impact ice clouds. Of particular scientific interest to me: what is the impact of smoke on ice cloud properties? This blog serves to catalogue such extreme wildfire events occurring around the world since 2013 and provides some background for each event. With contributions from major scientific groups (NASA, Naval Research Lab, NOAA, Canadian Meteorological Centre, others), this blog provides geostationary imagery animations, Aerosol Index maps from OMPS, cloud imagery from AVHRR, MODIS, and VIIRS, and more. Sometimes spectacular photos are sent in by interested viewers - these really help tell the story. The 2017 fire season will be our last, as funding winds down.


Spectral Ice Cloud Bulk Scattering Models Available

This long-term effort has largely wrapped up at this point. My colleagues (Dr. Ping Yang at Texas A&M University and Dr. Andrew Heymsfield at NCAR) and I hope these models are useful and hope to hear from you about what you have done with them.

Spectral models from the UV to the Far-IR in NetCDF: Provides access to ice cloud bulk scattering models at each of 445 discrete wavelengths ranging from 0.2 µm to 100 µm, both with and without the full phase matrix. Individual sets of spectral models are provided based on (1) solid columns only, (2) the aggregate of solid columns only, and (3) a general habit mixture that employs nine ice habits (droxtals, plates, solid/hollow columns, solid/hollow bullet rosettes, small/large aggregate of plates and an aggregate of columns). These models are based on randomly oriented particle calculations for severely roughened ice particles. Details are provided in Baum et al. (2014, JQSRT - preprint available on my publication page).

A netCDF file for each set of models is available that provides microphysical and single-scattering properties including the full phase matrix; each is about 55 MB compressed.

Separate files are also available that provide the spectral models over the full wavelength range but without the phase matrix (i.e., only the asymmetry parameter is provided). These files are quite small, about 250 KB each.

Imager Models: Provides access to bulk scattering models for about 36 different polar-orbiting and geostationary imagers; models include integration over the imager-specific channel spectral response functions and are based on severely roughened ice particles. Models provide the intensity (i.e., the scattering phase function) but not the full phase matrix.

Microphysical Data: Provides access to the in situ data for more than 14,000 particle size distributions from 11 field campaigns (and increasing over time).

Shortwave Spectral Models: Provides access to ice cloud bulk scattering models at individual wavelengths from 0.4 µm to 3 µm in wavelength at 0.01 µm resolution; based on severely roughened ice particles. Models provide the intensity (i.e., the scattering phase function) but not the full phase matrix.

Old Version 2 models: still available but strongly recommend Version 3 models