I’m currently developing stochastic models for temperature data from the Cosmic Microwave Background (CMB), collected during the  WMAP mission. I downloaded datasets from the LAMBDA site (Legacy Archive for Microwave Background Data Analysis) and then developed software tools in Matlab to parse and analyze the data.

From my previous work in statistical geodesy, I recognized that the temperature anisotropies of the CMB can be analyzed and modeled using the same techniques I found useful for studying the fine structure of the earth’s gravity field when I worked at TASC under  contracts with the Navy and NASA.  Check out this table, which compares earth’s gravity with the CMB. For an introduction to my analysis techniques for spectrum estimation, see these slides [pdf].

An example of a power spectrum I computed from WMAP data along the prime meridian appears here. I used state-space techniques [pdf] originally developed for studying the fine structure of the earth’s gravity field. These techniques also provide the autocorrelation function, as well as sensitive statistical tests for Gaussian behavior based on Kalman filtering. If you are interested in the math behind the state-space techniques I use, please consult the report Stochastic State-space Modeling with Applications.pdf