Space Weather refers to conditions in the near-Earth space environment that can negatively impact the performance and reliability of technological systems or can endanger human life or health. Significant space weather impacts can be both costly and dangerous. The impacts of space weather include satellite memory errors, phantom commands, loss of communication with deployed troops and aircraft that are over the horizon, significant errors in GPS position determination, and failures of power grids. The development of a reliable space weather forecast capability is a high national priority for both civilian and national security purposes. Overcoming the limited availability of in situ and remote sensing data sources and finding ways to compensate for missing physics even in the most advanced Sun-to-Earth space weather models are fundamental challenges that must be solved to make advances to space weather forecasting models. This project employs computational thinking to carry out transformative research that addresses both of these challenges. It will pioneer new methods for space weather modeling that will be transformational and eventually lead to a reliable space weather forecast capability. The project is highly interdisciplinary and will pursue these important goals in space weather modeling by developing and utilizing frontier image reconstruction ideas from theoretical computer science together with novel adaptive model correction methodologies from control systems engineering.
The results of this project will aid in establishing a reliable space weather forecasting capability for the nation. Present methods have very limited reliability beyond now-casting. This project would help achieve approximately 18-hour forecast capability, which would constitute a tremendous advance. Through collaboration with the NASA/NSF Community Coordinated Modeling Center the project will enable validation and transfer of the resulting knowledge, systems and tools to real space weather operations. Computer codes, algorithms, and tools will also be made freely available to other researchers, allowing the broader research community to utilize the tools that are developed for their own research needs. Four graduate students will be employed and trained under this project. In addition, the research results will be incorporated in undergraduate and graduate curricula at University of Michigan. The project will also strengthen and expand an ongoing collaboration with a local children's museum by creating exhibits about space exploration and space weather.
1. The terrestrial uper atmosphere (thermosphere) is a strongly driven system in which the global state can be rapidly altered by the drivers. One of the main drivers of the thermosphere is the solar flux in the X-ray and EUV bands. The solar flux in these bands is proxied by ground-based measurements of F10.7, which is the flux solar irradiance at the wavelength of 10.7 cm. We considered the problem of estimating F10.7 and physical states in the ionosphere-thermosphere using the Global Ionosphere-Thermosphere Model (GITM) and retrospective-cost adaptive input and state estimation. This non-Bayesian method retrospectively optimizes the input to the estimator using the output from the physical system in order to drive the estimator output toward the output of the physical system. This technique is used to estimate F10.7 using simulated data as well as real satellite data. 2. Magnetic loops are building blocks of the closed-field corona. While active region loops are readily seen in images taken at EUV and X-ray wavelengths, quiet-Sun loops are seldom identifiable and are therefore difficult to study on an individual basis. The first analysis of solar minimum (Carrington Rotation 2077) coronal loops utilizing a novel technique called the Michigan Loop Diagnostic Technique(MLDT) is presented. This technique combines Differential Emission Measure Tomography and a potential field source surface (PFSS) model, and consists of tracing PFSS field lines through the tomographic grid on which the local differential emission measure is determined. As a result, the electron temperature Te and density Ne at each point along each individual field line can be obtained. Using data from STEREO/ EUVI and SOHO/ MDI, the MLDT identifies two types of QS loops in the corona: so-called up loops in which the temperature increases with height and so-called down loops in which the temperature decreases with height. Up loops are expected, however, down loops are a surprise, and furthermore, they are ubiquitous in the low-latitude corona. Up loops dominate the QS at higher latitudes. The MLDT allows independent determination of the empirical pressure and density scale heights, and the differences between the two remain to be explained. The down loops appear to be a newly discovered property of the solar minimum corona that may shed light on the physics of coronal heating. The results are shown to be robust to the calibration uncertainties of the EUVI instrument.
