A multi-platform online magazine for multi-disciplinary and in-depth content. http://perspiceremagazine.org/

• Taught discussion/lab sections for classes Air Pollution, Ocean Biogeochemistry and Field Method. • Gave lectures for key review parts and when lecturer is absent.• Practice verbal communications and explaining science to non-science students.• Average evaluation from students: 3.7 out of 4.

• Used statistical methods (partial regression, analysis of variance, resampling techniques, etc.) to evaluate fire risk at global scale. Performed by open source Python packages.• Analyzed remote sensing data and social-economic data with 1 km resolution and global coverage, total size in terabytes. Performed on UCI Linux cluster and national lab supercomputer.• Built models for post-fire vegetation growth, produced a Python package for model building and fire analysis. Coupled this model with National Center for Atmospheric Research Earth system model.• Written and maintained data science lab note for teaching other group members. • Acquired excellent verbal and written skills by collaboration with multi-discipline group and writing proposals.

• Used machine learning (Artificial Neural Network) on paleo-pollen data to reconstruct vegetation of the past, acquired better results than other traditional methods. • Create new statistic method for extreme detection in remote sensing data, identified key regions for vegetation degradation.• Paper writing and data visualization for 4 published papers.