If it is an image that goes along with data, like in a paper for example, I think that it has to be additive and supportive to the text. An image can be very powerful to make (complex) data clear and understandable. In this case the image can be an abstraction and simplification of the subject matter. I think a good way to think about it is to break down the concept you want to depict to its most basic and simple form. This piece of information you need to illustrate as correct as possible. That way you will end up with a scientifically correct but simplified image that will convey your message in the most optimal way.

On the other hand, scientific visualization is also used quite a lot without supporting text. In that case you will need more 'context' in my opinion to make your point. Unless it is scientific art, but i think you are not referring to this.

Success with your efforts, and keep us posted :)

cheers,

luk

thank you for the reply. I will be posting my findings. Best regards. Jose

The scope of Advanced or Scientific Visualization is very large. The NSF (National Science Foundation) in the Visualization in Scientific Computing Workshop report (1987), describes this topic in the following terms:

“Visualization is a method of computing. It transforms the symbolic into the geometric, enabling researchers to observe their simulations and computations. Visualization offers a method for seeing the unseen. It enriches the process of scientific discovery and fosters profound and unexpected insights. In many fields it is already revolutionizing the way scientists do science . . . The goal of visualization is to leverage existing scientific methods by providing new scientific insight through visual methods”

Thus, Advanced Visualization must be linked to a more efficient management of data repositories, modelling tools and scientic computing related to each scientific area. A good approach to this topic can be seen in C.D.Hansen and C.R.Johnson (eds): “The Visualization Handbook” (Elsevier, 2005), where you can find an introduction to modelling tools which are based on different kinds of (scalar, vector and tensor) fields, Geometric Modeling (and their extensions to Virtual/Augmented Environments), Large-Scale Data Visualization and other topics in this knowledge domain. Thus, Advanced Visualization does integrates models, tools and results arising most of physical and mathematical tools. The most spectacular results concern to simulation of Fluids Mechanics, but advanced visualization of biomedical images contains perhaps the most known results. It includes static applications (bones, internal organs, etc) and dynamical modelling (viruses attacks, simualtion of pharmaceutical responses, etc). Remark that visualization is not necessarily realistic; thus the resource to images is allowed, but sometimes unrealistic visualization is even more interesting and used very often in modern biomedical and/or engineering applications.

cheers

Javier Finat

Scientific illustration (drawing, painting, sketch, diagram, or picture in a book, magazine article, etc.) scientific visualization (imagine visually), scientific art (human creative skill or its application), scientific 3-d representation (same as fore mentioned descriptions but in three dimensions). Scientific illustration usually requires a large degree of scientific knowledge with respect to taxonomy and visual accuracy eliminating anomalies and atypical characteristics (unless those characters are the target of the illustration). Scientific visualization has a looser interpretation and is used mostly for subjects that are not available for direct comparison (e.g. paleontology, astronomy, atomic interactions, theories, etc.) but have as much accuracy as research allows. Scientific art is just that, art that is inspired by science (e.g. Alex Grey) and is not subject to the same degree of accuracy and is more personal to the artist. I am a scientist, an artist, and a teacher with over 40 years of experience in museums, universities, zoological institutions, and art organizations.

Thank you Javier and Glenn for your input. I really want to shake this tree and see if what we (as computer scientists) have come to believe is visualization is really what users (with a high level of knowledge on a domain) are expecting and need or they just cope with what we provide them and suffer in silence.

I don't think that scientific illustration is any different than it is in any other field. Illustration exists because we gain an understanding of the structure of the world through our eyes, better than through ideas attached to words. It's a lot easier to understand what an elephant, or a sunset or a DNA molecule, looks like through a picture than through a word description. It's pretty basic.

there are artefacts that exist in the real world that we can easily identify in a picture, such as you point out with the pictorial representation of an animal or a sunset. We we take a large dataset and assign arbitrary representations in order to " ease" the analysis and visualization of the underlying data, some other issues might appear. The use of colours, arrows, ribbons, streamlines might or might not be conveying enough information. usually, we assume that, for instance, when a CFD scientist is looking at the behaviour of some fluid and it represented with a lot of 3d arrows and area sin red, blue, green, etc, we assume that it is straight forward to understand the behaviour from the fluid, based on those visual cues. Objectively, that representation is just a group of shapes, which might not make any sense for a lay man. A question there is, will a person with no CFD training arrive to the same interpretations that the scientist? Will they look at it in the same way? Is it really easier to understand it in that way, than, for example, through a sample of values for the main variables?

I've seen scientific diagrams that made no sense to me but might have made sense to someone in that field. Sometimes it's because I'm not familiar with the labels and the underlying hypotheses, but in a lot of cases I think they're just bad graphics, bad in that they don't clarify the relationship of the things being diagrammed and labeled. The solution to this is in the application of design principles and visual communication theories that are based in the psychology of how the mind interprets what it sees. A good designer should be able to make an arcane data set understandable to even a layperson, especially by making the patterns of the data easy to perceive. Edward Tufte is a leading proponent of this type of data visualization.

Addressing your original question, are you studying the use of graphics by scientists in order to understand the data patterns themselves or to explain the data to others?

The most important task of scientific illustration is the interpretation of the data obtained. Illustration should be understood as the written research report, not as the technical image. Illustration is rarely a document. The illustration aims to the interpretation of data in terms of research. Illustration uses human intelligence an can something that no technical device can replace. Researchers should appreciate this quality again and learn to use.