Inside this Issue

• Bioinformatics
• Next Generation Tsunami Modeling
• Space Weather Forecasting
• Navigating Information Space
• Imagining Without Constraints
• Internal Tides of the Ocean
• High-Flying at Mach 8

Story by Dr. Sergei Maurits and Jeff McAllister

The National Space Weather Program, established during the last decade by several U.S. government agencies, defines space weather as “conditions on the Sun and in the thermosphere that can influence the performance and reliability of space-borne and ground-based technological systems and can endanger human life or health.” The University of Alaska Fairbanks Eulerian Parallel Polar Ionosphere Model (UAF EPPIM) is a computer model of the northern polar ionosphere that helps scientists understand and predict space weather events.

Space weather variability and intensity are closely related to solar activity. The most pronounced solar pattern is its eleven-year cycle, which governs the number of active solar regions. The latest peak of this cycle occurred in 2000-2002. Now, in 2004, the Sun is entering a period of decreasing activity. The post-peak solar activity “season” is characterized by less frequent but potentially more powerful coronal mass ejections (CMEs). CMEs dramatically affect the solar wind, which in turn alters Earth’s magnetosphere and ionosphere. They can generate powerful geomagnetic storms and spectacular auroral displays.

The series of solar eruptions occurring in October/November 2003 was named the Halloween storm. One of the largest magnetic storms on record, the Halloween storm, brought an enormous amount of geophysical information to the space weather scientific community. The UAF EPPIM has a long record of contributing to storm study efforts. The model, which was developed and refined at the Arctic Region Supercomputing Center (ARSC) and UAF’s Geophysical Institute, produced findings for the Fall 2004 AGU presentation “Comparison of the UAF Ionosphere Model with Incoherent-Scatter Radar Data.” Results of this study are also being prepared for future publication in the Journal of Geophysical Research.

In addition to past-analysis, the UAF EPPIM also accomplishes ionospheric forecasting. Real-time forecasting involves management of the model’s continuous run to keep the model time in synch with the system clock and current solar wind propagation velocities. To simulate conditions approximately two hours in the future, current forecasts use inputs from the National Oceanic and Atmospheric Administration Space Environment Center in Boulder, CO. The inputs include upstream measurements of the approaching solar wind by the ACE (Advanced Composition Explorer) satellite.

Validation of the produced forecast versus data is an important direction of the model’s development as a space weather tool. Comparisons with ground-based ionosondes and ionospheric tomography data are currently used. Massive (>100,000 per year) statistical comparisons with ionosonde critical frequency foF2 (the highest frequency that will reflect from the main F2-layer of the ionosphere for vertical propagation) indicate that achievable accuracies range from 10 to 20% of RMS (relative root-mean square) for the summer season daytime and 30 to 50% for winter nighttime (http://www.arsc.edu/SpaceWeather). Accuracy is improving as the model matures.

The UAF EPPIM can be run at arbitrary grid resolutions. ARSC HPC tools were used to optimize the code’s performance. The model has been ported to ARSC’s Cray, SGI, IBM, and Linux computers. The result is a robust parallel code, which can run on a wide variety of platforms at gradient-resolving resolutions exceeding 10 x 10 x 10 kilometers. ARSC’s HPC resources facilitate explorations of periods of special interest at such extreme fidelity, while computational runs at the regular resolutions routinely cover model years. The ability to perform these long runs on ARSC HPC platforms significantly speeds validations and improvements to the model.

The EPPIM development is supported in part by the Department of Defense’s High Performance Computing Modernization Program and a University Partnering for Operational Support (UPOS) grant. UPOS is a joint program of UAF and the Applied Physics Lab of John Hopkins University for development of prototype applications for the U.S. Air Force Weather Agency (AFWA) and other agencies. Ionosphere and related studies of radio signal propagation are one of the major science emphasis areas supported by ARSC. EPPIM is a collaborative project with Dr. Sergei Maurits and Jeff McAllister, of ARSC, and Professor Brenton Watkins and Dr. Anton Kulchitsky of the UAF Geophysical Institute.

The graphs above show typical comparison results with ionosonde data. A wintertime period (January 16-19, 2004) was selected to illustrate a wide range of day/night ionospheric variations. While mid-latitude and sub-polar curves show pronounced diurnal trends (lower and mid-panels), the high-latitude graph demonstrates the effects of highly-structured electron density patterns. The passage of such ionospheric structures is recorded as a noticeable modulation of the ionosonde data (red triangles, upper panel). The EPPIM (solid line) reasonably follows the actual ionosphere in terms of modulation depth, which is a strong indicator of the model’s capability to simulate ionospheric motion and structuring with high-resolution inputs. Future transition to data-enhanced drivers (such as velocity patterns Super Dual Auroral Radar Network—SuperDARN) is expected to improve the predictions further.