Forensic Hydrology?

December 1, 2011 7:33 AM

How can you tell where water came from? Lets say I give you a bucket of water taken from a local water source (river, lake, etc). There's some plants and maybe a fish in it as well. What could you tell me about where this water came from, could you pinpoint which body it came from, and how would you come to this conclusion?

Yes, this is for fiction.

posted by The Whelk to Science & Nature (21 answers total)

Mot exactly water, but "Balloons Of War" by John McPhee breaks down geology in just such a way, on sand if I recall? They figured out exactly where the balloons were coming from based on the sand, it's "forensic geology" I think? Seems to be such a thing as "Forensic Hydrology" as well, so I'd guess that's the way to go.
posted by Blake at 7:40 AM on December 1, 2011

Isotope analysis would be a good start. I've read somewhere it is powerful enough to determine in which region some plant was grown to detect subsidy fraud. With the plants, fish and other debris in your water it should be possible to pinpoint the aquifier.

Provided you have a list of isotope ratios of the aquifiers you want to compare it with.

With the plants and fish you could probably discern whether it was flowing or stagnant water. If there are some industry plants or cities releasing pollutants into your hypothetical river, you could find out in which part of it the water was taken.

How accurately do you need to localize the source?
posted by Triton at 7:43 AM on December 1, 2011 [2 favorites]

If you had reference samples from other places, you could match DNA for close relatives of the creatures in your sample. You could profile all the matter, but that wouldn't get you too far in a court of law without reference samples, both matching and non-matching.

You could look at satellite imagery, and locate, for instance, nearby dairy farms, or fertilized crops, and predict those chemical signatures. You could use satellite imagery and known growing zones to identify likely coincident flora, and look for the associated pollen.

Spring water might have dissolved minerals that could be predicted from an understanding of the underground conditions.

With spring water, you may be able to determine from dissolved gas ratios how long the stream has been above ground. I think there would be more CO2 underground, and more O2 above ground. Also, if you captured an airtight sample, the dissolved gasses might reflect climate and nearby industrial activities.
posted by StickyCarpet at 7:48 AM on December 1, 2011

Here's a detailed presentation from the USGS on forensic hydrology - looks like they're studying isotopes and minute traces of other compounds. Fascinating stuff.
posted by jquinby at 7:51 AM on December 1, 2011 [2 favorites]

Stable isotope analysis isn't exact enough to determine what aquifer. It's better at continental scale comparisons.

If you're lucky enough to be writing about the SE United States, you could have pulled up a fish or crayfish endemic to one river or stream.
posted by hydrobatidae at 7:56 AM on December 1, 2011

Response by poster: This is about New York State (should have mentioned that)
posted by The Whelk at 7:57 AM on December 1, 2011

Lets say I give you a bucket of water taken from a local water source (river, lake, etc).

The temperature of the water at the source might be relevant.
posted by StickyCarpet at 8:03 AM on December 1, 2011

On preview, not relevant to forensics for New York, but interesting tidbit:

I had a conversation with the owner/president of a local artisan brewery, where he told me that to make the beer taste authentic to the region its style originated from, a key component was the right water chemistry. He had compiled a recipe book of sorts for adding minerals to the water, for different geographical locations like Belgium or Ireland. He said that as a general characteristic, going from eastern Europe to the West, the water goes from soft to hard. This translated into beers tasting better as lighter or darker beers, so a general trend was soft water --> light beer, and hard water --> darker beer. Hence, Ireland --> Guinness.
posted by lizbunny at 8:04 AM on December 1, 2011

This is not really so much a contribution as a suggestion, but what if an herb with potent roots (like ginseng) was growing by the river? Would enough of the chemicals in the root be released in the water and be identified by gas chromatography or a similar technique?

I'm not a scientist, please ignore this if it's far-fetched.
posted by mccarty.tim at 8:08 AM on December 1, 2011

Here's a list of nuclear plant locations in NY state:

http://www.eia.gov/cneaf/nuclear/state_profiles/new_york/ny.html

Perhaps you could look for a very slight increase in radioactivity? I realize that cooling water doesn't pick up much, but maybe you'd have some funny isotopes in slightly higher concentrations if the water came from near one of those spots?
posted by jenkinsEar at 8:12 AM on December 1, 2011

As someone who done similar things, it's fairly difficult to do.

There's no one analytical technique that would give you an exact answer.

Water is difficult because there's comparatively little in it, in terms of extras, suspended sediments, biota, chemical residues. Soils and sediments are a lot easier to work with, forensically.

Isotopic analysis isn't specific enough without unique geological features near by. Isotopes do ages well, but sources, not so much.

Plants and animals can give you sterols and other "bio-marker" chemicals to look for, but again, not hugely specific. The exact mixtures are very sample-collection dependent and I'd be very careful of sampling error.

DNA has the problem of specificity as well; organisms will tend to populate broader areas which makes nailing down a location hard. Forensic DNA is possible, but then you do have to know how wide-spread a population is, and that can be difficult.

I'd bet on being able to tag a watershed, but determining where in that catchment basin? I would not likely be comfortable going to court with that.

What can make this possible is industrial residues and pollutants. Volatile chemicals and mine tailings often have patterns of chemical abundances that are unique to a particular source. In particular, give me a family of compounds like petroleum residues, PCBs, or PCDD/Fs and I can give you a quite specific answer, usually.

Note that this is all comparative. I would need suspect source samples as well. None of this is cheap or quick, by the way. We charge about $2000/sample (for only one of the types of analysis listed above) and it takes us a couple of months to do right.
posted by bonehead at 8:13 AM on December 1, 2011 [7 favorites]

Heavy metal analysis. I know an analytical chemist who has also done forensics work in this area. And yes, heavy metal analysis can be used to identify the source of a substance, assuming you have enough samples from different sources to create profiles. This person has done some interesting work for the DEA using this technology.
posted by ellenaim at 8:14 AM on December 1, 2011

“Sight-unseen” detection of rare aquatic species using environmental DNA
posted by glibhamdreck at 8:26 AM on December 1, 2011

This is about New York State

If the water has been transported from a reservoir, NY State is unique, perhaps, in using giant old oak-lined underground pipes. The oak should have some kind of signature.
posted by StickyCarpet at 8:47 AM on December 1, 2011

This is still a bit science fictiony, but the microbiome of any static body of water, and perhaps any geograhically defined stretch of flowing water is likely to be highly distinctive. You might need a big bucket to capture a sufficient sample, but the pattern of microbial contamination is probably more specific than anything else.
posted by roofus at 9:03 AM on December 1, 2011

Perhaps a tracer has been introduced into the water? A dye or other compound? I know they've done this around my neck of the woods to try to figure out where water in various caves eventually emerges above ground (or if it does at all).
posted by jquinby at 9:13 AM on December 1, 2011

Most water on earth is believed to come from the multitudes of comets which hit the Earth, early in its life. The comets had a lot of water ice on them. But that is not really what you are looking for here.
posted by Danf at 10:45 AM on December 1, 2011

DNA has the problem of specificity as well; organisms will tend to populate broader areas which makes nailing down a location hard.

What about cross-referencing the various biota in The Whelk's bucket? If a fish is found in one area, a tadpole is found in another and seaweed in a third, the small area where those organisms have domain should contain the source.
posted by Terminal Verbosity at 10:52 AM on December 1, 2011

What about cross-referencing the various biota in The Whelk's bucket?

That's a significant sampling problem. Can you prove what the geographic range is for that population? Can you be certain that those particular genetic markers are not present in nearby or even far-flung places? A huge amount of environmental sampling may need to be done to establish that.

It has been done though: there was a case in the early 1990s where a suspect's presence at a crime scene was established by genetically-matching the leaf of a bush at the scene with one found in his vehicle. Sometimes you get lucky.

It also comes down to what a good enough answer is. Is it ok if a few miles is good enough? Does it need to be a few inches?
posted by bonehead at 11:11 AM on December 1, 2011

At eh very basic level (and cheap too) there are field level analyses that could be done using tools as basic as test strips (pH, hardness, etc) to electronic in-situ test tools (for example this Horiba) that could give you quantitative information of things like conductivity, turbidity, and dissolved oxygen. If you gave me a sample and said "lake, river, or ocean" I could guess correctly given the information gathered with these tools.
posted by Big_B at 2:00 PM on December 1, 2011

It's probably easier to figure out where the water didn't come from.

SPOILER:

There was a Bones episode (bear with me) that had a dead guy (of course) who died by drowning. He was a hockey player and they ruled out the ice from the rink and the lake where they actually found the body. They finally figured out that he died from being drowned in his fish tank. All of this was based on tests they made on the water found in the corpse's throat and the other bodies of water involved.
posted by deborah at 5:38 PM on December 1, 2011