Craig DeForest

Solar physicist

 
 

I am interested in the formation and heating of the solar corona, and in the tools we use to probe that mystery.  The fundamental puzzle of solar physics is: why is the solar corona about 300 times hotter than the photosphere of the star?  Answering that question leads to many twisting paths through the intellectual woods.


The answer is tied up in transport of energy from the core of the star to its outer atmosphere.  Heat from nuclear fusion is converted into mechanical energy by convection in the Sun’s outer layers; some of that mechanical energy cascades into a rich, complex magnetic field that forms, shapes, and heats the solar corona.  The exact mechanisms of magnetic energy storage and release are not (yet) well understood, though I have hope that they will be in my lifetime.  Probing them requires developing new types of instrument and new data analysis techniques.

Current research topics and interests

Deep-field solar wind imaging: Using the rather excellent data sets available from the STEREO mission, we’ve been able to extract the best images to date of the Thomson-scattered light from solar wind features in the heliosphere.  You can download two early presentations here in PowerPoint format [AAS/SPD 2011 (204 MB); Boulder Solar Day 2012 (582 MB)].  Since 2011, we’ve  reported several groundbreaking new results in the literature, including: the technique we use; identity of pre-CME cavities in the Sun with magnetic clouds near Earth; distortion of the flux rope by ram effects in transit; the derived rate of open magnetic flux disconnection from the Sun (which affects the strength of the interplanetary magnetic field); the relative importance of pickup and source material to ICME propagation; the importance of turbulence in solar wind propagation; and the onset of turbulence at the top of the solar corona itself.  You can see my papers page for each of these results.


MHD modeling: Magnetic reconnection and its consequences are critical to understanding the solar corona.  Magnetohydrodynamics is too complex to be addressed analytically in nontrivial cases, so everyone uses numerical simulation to understand it, but numerical codes typically cannot simulate the extremely high conductivities present in the solar corona.  This is important, because conductivity and reconnective energy release are intimately connected in non-intuitive ways.  Together with several others including Charles Kankelborg and my graduate student Laurel Rachmeler, I developed a new approach -- fluxon modeling -- that allows simulation of magnetic systems in the complete absence of unwanted magnetic reconnection (DeForest & Kankelborg 2007).  We are using it to conduct the world’s first controlled numerical experiments on the effects of reconnection (Rachmeler et al. 2008; Rachmeler et al. 2009).  Our fluxon code, FLUX, is now capable of predicting steady solar wind flow.  We are working on deploying it to the CCMC.


Space weather prediction: The largest magnetic instabilities on the Sun drive coronal mass ejections (CMEs), which can propagate across the entire inner solar system and impact Earth, with consequences for radio communication, power grids, spacecraft, aircraft, and other parts of our technological web.  Predicting space weather requires understanding both the origins of CMEs and the evolution of their interplanetary counterparts, ICMEs.  Solar wind imaging (“heliospheric imaging”) has the potential to revolutionize space weather prediction, much as tracking weather satellites revolutionized terrestrial weather prediction.


Computer vision: As data become easier to acquire, data set size has grown immensely.  the Solar Dynamics Observatory is expected to produce several terabytes of data per day, with a total volume measured in petabytes.  Systematically extracting useful information requires automated feature recognition and other computer vision techniques.  Our magnetic feature tracking code, SWAMIS, helped to standardize computer vision techniques in the solar physics community, and is currently deployed as part of the SDO/HMI analysis software, detecting emerging flux regions in the solar photosphere.


Spectral data compression: I developed and patented a hybrid spectral wavelet compression algorithm that is being used to compress spectral image data from ESA’s upcoming Solar Orbiter mission.  Solar Orbiter’s SPICE instrument will downlink UV spectral images with just 16 bits per line profile.


Perl Data Language: PDL is a fast, flexible, free vector data analysis language based on Perl. It has a lot of advantages, one of which is that it is open source, and another of which is that it is incorporated into all major Linux and Mac open-source repositories.  It is more concise and powerful than any of the commercial alternatives or the scientific Python dialects, and permits easy inbound and outbound calling for interoperation with other languages (e.g. C).

Karen Harvey’s Dissertation: “Magnetic Bipoles on the Sun”

Karen Harvey was a pioneer of solar magnetic image analysis, and much beloved within the solar physics community.  In 2002, I found that her important dissertation had become nearly unavailable.  Mandy Hagenaar lent me her copy of the dissertation, and I scanned it for posterity.  With permission from Jack Harvey, here is a copy of the complete dissertation, as a PDF: kharvey_dissertation.pdf (420MB)