Thursday, October 28, 2010

Applying actual science to blood splatter evidence

A recurring theme on this blog has been the lack of actual science behind many so-called forensic "sciences," as highlighted last year in a major study by the National Academy of Sciences, and the need for primary research in many fields where flawed evidence is commonly relied upon in court. In that vein, here's an interesting article from Inside Science News Service about primary research into blood-splatter evidence which debunks past techniques while pointing to possible, actually science-based methodologies for using such evidence:
Sanjeev Chandra, an engineer at the University of Toronto who helps General Motors develop better ways to spray paint its cars. "A lot of the physics is exactly the same for blood."
In soon to be published research, Chandra and his team have revisited the techniques and software packages that forensic experts have developed over the years to reconstruct the origin of blood splatters. By testing this software scientifically on splatters of pig blood in the lab, they've shown that there is significant room for improvement in the models, which typically use straight lines to trace the path of blood droplets a surface back to their point of origin.

"They aren't very accurate," said Chandra. "They don't consider the effects of gravity on blood droplets. They ignore air drag, which can be very significant."

Calculating the speed at which drops of blood leave the body during an attack is an important measurement for blood pattern analysis. The physics behind the velocity and size of a blood drop gives investigators an idea of what kind of wound was inflicted.

A drop of blood falling from a cut finger, for instance, is a battle primarily between the force of surface tension, which keeps it stuck to the body, and gravity, which pulls it downwards. Solve the equations, and you'll find that a typical drop released this way has a volume of less than one percent of a teaspoon.

Blood released from a wound by a violent impact -- such as a bullet -- tends fly in even smaller drops. That's because the force of a bullet is much stronger than gravity and easily overcomes the surface tension, flinging tiny drops away at high speeds.

Forensic investigators look at the size of blood spots created on surfaces at the crime scenes to get an idea of the volume and velocity of blood drops produced by a wound. But this doesn't always provide a clear picture.

"Often, the same pattern can be produced in many ways," said Daniel Attinger, an expert in fluid dynamics at Columbia University in New York who is working with the Department of Justice to strengthen the scientific framework behind blood pattern analysis.

A large spot could be the result of a small drop moving very quickly that spreads out on impact, or a large drop that contains more liquid but hits the surface at a slower speed.

That's why Chandra and his group are taking a closer look at not only the size but the shape of blood spots. They're developing a new method for calculating the speed and volume of a drop of blood by measuring the tiny spines that extend from blood drops like the tentacles of an octopus.

In a paper published in 2005, they showed that counting these spines could help to differentiate different velocities. A forensic investigator asked to analyze a spatter of pig blood in the lab using this new method had "fairly reasonable success," said Chandra.

One limitation of this technique, though, is that a spot of blood must be well-preserved to maintain these spines. This depends on the surface: wood and drywall preserve these shapes, but glass and tile are too smooth.
Applying scientific rigor to forensics is a win-win for everybody. These critiques of traditional blood-splatter evidence - particularly the false assumption that blood travels in a straight line and the failure to account for the effects of surface tension - are among the reasons I recently included it in the "Top Ten Forensic Sciences That Aren't Really Science." Until this research is complete and new techniques (which are still being developed) have been tested and replicated by others, the questions arise: Will the old techniques continue to be used in court, and will past convictions where false forensic conclusions were presented to juries as fact be revisited to make sure innocent people weren't convicted based on junk science? My guess, regrettably, is the answer to those two questions will be "yes" and "no."


Prison Doc said...

A win-win it will be, and sorely needed. As a scientist I probably have a bias toward "scientific" evidence but DNA seems to clearly be a winner (Except when perverted as in the first OJSimpson case. Hopefully we are seeing the end of "eyewitness" identification and the newer findings on inaccuracy of fingerprint evidence, etc. will surely help. Good article.

Anonymous said...

It is not "blood splatter" but "blood spatter."

Gritsforbreakfast said...

10:26, They're synonyms, and "splatter" was the word used in the article.

Anonymous said...

What if actual science gets in the way of lockin' somebody up (or sticking a needle in their arm)?

I think some people on this blog would benefit from reading "The New Jim Crow: Mass Incarceration in the Age of Color Blindness" by Michelle Alexander.

Rev. Charles

Anonymous said...

Yes, blood does splatter, but in the Forensice context, it is spatter. Sorry, one of my pet peeves. Still love the blog, though!!

Anonymous said...

And of course I just lost all credibility by misspelling Forensic. My bad!!