Data Mining of Atmospheric Releases
I am working with my collegues in the Laboratory for Atmospheric Hazard Modeling on reducing the uncertainty of atmospheric release simulations using machine learning clustering and simulation. The goal is to find patterns of high risk areas, and determine the atmospheric conditions which are associated with each area. Such patterns can be helpful in near real time hazard response, in order to determine what are the areas most at risk, how to cohordinate relief efforts, and ultimately how to best protect people and properties.
CQuake
CEOSR Earthquake Monitoring and Forecasting System
CQuake is a real time data mining program to monitor and forecast earthquakes using remote sensing and model data.
Within 30 minutes after an earthquake occurs, CQuake performs a systematic analysis of remote sensing and model data using data mining algorithms. The results are made publicably accessible at http://crete.gmu.edu/cquake. Additionally, once a day, CQuake downloads data for several predefined regions of the world, and performs an analysis to find anomalies which might give early warning information about impending earthquakes.
The goal is build a better understanding of the processes involved in the earthquake preparation stage, and to find patterns that can be used to mitigate earthquake risks.
I developed CQuake under the supervision of Dr. Menas Kafatos and Dr. Ramesh Singh, and in collaboration with several US and international research groups.
Latest Earthquakes -- Updated every 30 minutes
Near real time image of the weather over the USA and Mexico.
SWC
CEOSR Severe Weather Center
The goals of the severe weather center are to provide real-time or near real time monitoring of severe weather events, such as hurricanes and tornadoes for the Mid Atlantic region of the USA. In addition, it will provide analyses, comparisons, and assessments of current versus past severe weather events.
Our products will include model runs, online dynamic analysis for satellite observations, including three-hour, daily, and monthly precipitation reports derived from the TRMM observations, sea surface winds from the other satellites, e.g. QuikSCAT, model results and online satellite observations.
Analyses will be combined with satellite images, such as visible, enhanced infrared, and water vapor images from the GOES observations, cloud images from the TRMM and MODIS.