The above is a visually enhanced scientific simulation of a tsunami wave “hitting” the harbor of Heraclion in Crete. The animation data were produced using a methodology for the photorealistic scientific visualization of tsunami waves, i’ve been working on, with the Technical University of Crete, sporadically over the last 5 years. The video was presented, along with a scientific paper, at the IEEE International Conference of Imaging Systems & Techiques, last October in Santorini.
The Abstract of the paper is following :
“A methodology is proposed for the photorealistic scientific visualization of a tsunami wave. The methodology is applied to the case of the tsunami occurring after the Mediterranean earthquake in 365 AD, provoking sea level rise in Heraklion port in Crete, Greece. The work presented in this paper puts forward a semi-automated workflow for parsing modeled data embedded in animation software of choice. By scripting the manual work needed for the import and processing of geo-referenced time arrays, a 4 minutes animation showcasing the automated workflow was created. The script implemented directs the automatic loading of geotif data and produces the required shapekeys needed for the final 3d animation. It minimized production costs for the sea level animation to a minimum and aided the quick visualization of large arrays of numerical models in geo-referenced 3D representing the impact zone. The 3D & animation data can then be handed out to the post processing team focusing on special effects and photorealistic rendering, maintaining scientific accuracy in the final animation.”
The co-authors of the paper are :
- Katerina Mania (Associate Professor, Department of Electronic and Computer Engineering of the Technical University of Crete, Greece )
- Giannis Giakoumidakis (Department of Electronic and Computer Engineering of the Technical University of Crete, Greece)
Download the paper from the IEEE Explore, digital library:
If you don’t have a IEEE membership just ask me for a full copy! I’ll be happy to mail it to you.
In a next post i will share the Blender3d Python script that was used to automate the animation of the sea level based on scientific data.
The following video was produced, a few years ago, by importing (just) 20 time arrays of heightdata and setting up the animation keys manually :
For comparison, the first animation has a total of 480 time arrays (keys) and was completed in a fracture of the time.
The goal of this work was to produce reliable animation data, that can be enhanced with visual effects, while maintaining scientific accuracy.