AD Main Menu

Shocker: Marijuana odor may indicate presence of marijuana

Ben Anderson
Aaron Jansen illustration

A press release Thursday from the Alaska State Troopers revealed the results of a three-year study that determined the answer to a question bothering everyone: Is the odor of marijuana associated with the presence of the drug?

The statement from troopers, titled "Study Finds Marijuana Odors Significantly Associated With Marijuana Grows" was based on research conducted by the University of Alaska Anchorage Justice Center at the request of troopers. Troopers were quick to point out that no tax dollars were use to fund the study -- it was paid for with "federally forfeited illegal drug proceeds," according to troopers, who said the goal was to obtain:

...an empirical estimate of the extent to which AST investigators’ detection of marijuana odors served as a reliable indicator of the presence of illegal quantities of marijuana in structures.

The study examined 200 searches for marijuana conducted by troopers between 2006 and 2008. Of those 200 searches, 197 resulted in the seizure of marijuana. The three cases where no marijuana was seized were not preceded by the detection of marijuana odor.

The study concluded that the likelihood of discovering four ounces or more of marijuana was 2.7 times greater if based on the presence of marijuana odor. In cases where there were 25 or more plants growing and marijuana odor was present, the likelihood of finding marijuana was 2.9 times greater than in cases where no odor was detected, The study said.

Presumably, the study could be used to buttress requests for search warrants in cases where law enforcement officers claim to have smelled marijuana. The courts have proven skeptical of some smell claims.

Return of the dope-sniffing Alaska trooper

In April, a U.S. District Court judge rejected the notion that trooper Kyle S. Young, who made a pot bust in the Matanuska-Susitna Valley in February, sniffed the weed in question while driving down the street in his trooper vehicle. Young subsequently obtained a search warrant based in significance on his sense of smell.

Young made a bust. Questions then arose about the legal justification for the search.

An expert witness testified in court that there was a "zero" chance that Young had initially smelled the marijuana grow operation from the location that he testified to. Due partly to that testimony, the evidence -- the plants and equipment from the grow operation -- used to make the arrest was thrown out, undermining the state's case.

Questions arose about other cases when it turned out this wasn't the first time Young had made a pot bust based on the claim of smelling cultivating marijuana. The Young case brought to the fore questions about the legitimacy of the scent test in determining the presence of growing marijuana. It's still a nebulous issue, as the case of Alaska State Trooper Kyle S. Young shows.

Despite the obvious implications troopers draw from the "study," was the distinct smell of marijuana ever really questioned? The report is full of painfully obvious tidbits like this one:

AST investigators discovered four ounces or more of marijuana in 90.7% of the searches that were preceded by the detection of marijuana odors by one or more investigators. In contrast, four or more ounces of marijuana was discovered in 78.7% of searches that were not preceded by the detection of marijuana odors.

And:

...while the detection of growing/green marijuana odor by investigators is not significantly associated with the discovery of “small” amounts of marijuana (aggregate weights of less than four ounces), it does appear to be significantly associated with the discovery of relatively “large” quantities—that is, four or more ounces and/or 25 more plants—of marijuana. 

So when troopers smelled weed, they found more weed, more often, than when they didn't smell weed.

Considering the odor of marijuana constitutes probable cause for a search warrant, it stands to reason that in cases when troopers were able to use scent to help obtain a search warrant -- 133 of the 200 total cases -- they found marijuana.

The inexact science of smells

The study does not mention how many of those 133 search warrants were issued based solely or in part on the presence of marijuana odor and how many were initiated by some other factor such as heavy use of electricity.

The study also noted that in 25.8 percent of the cases, less than 25 growing plants were found despite one or more troopers' assertion that they smelled green or growing marijuana. The false positive rate for searches that turned up less than four ounces of marijuana was 9.2 percent. "Ideally," the study says, "this proportion should be zero." Between the more than 25 percent of cases where troopers said they smelled marijuana but turned up less than 25 plants -- 31 of the 121 cases -- and the 11 cases that resulted in finding less than four ounces of marijuana, the study indicated that smell was not necessarily a foolproof indicator of the presence of large quantities of marijuana.

Unfortunately, sense of smell is not an exact science, and numerous variables can come into play. Some of those noted in the study include wind, ventilation systems installed at marijuana grows, and the presence of doors or windows, and the troopers' proximity to the marijuana.

The study doesn't note a variety of more nebulous factors -- the individual sense of smell is highly variable. A trooper who was a smoker might smell things differently that one who is not. The day-to-day health of a troopers' sinuses could be an issue.

The study noted that officers did not smell burning marijuana in any of the cases, even though it is a much more pungent indicator of marijuana than most green and growing plants. The study did not note what percentage of troopers reported smelling the marijuana at first contact, or in how many cases did troopers failed to report the scent of marijuana until after the bust had been made. The study did not discuss whether any troopers remembered the smell only after sitting down to write a final report on the search and/or seizure.

The study does, however, strengthen the case for probable cause in cases where an officer does a knock-and-talk at a suspected marijuana cultivator's door and detects the smell of marijuana. The study supports the idea that when a trooper with a good nose smells pot in or near a structure, there is most likely going to be pot in there.

Study inconclusive

But the study fails to quantify whether or not the detection of the smell of growing marijuana is as prevalent as troopers in the field report. Barring a double-blind smell test administered in trooper academy to establish the subjective qualifications of each law enforcement officer when it comes to assuming the role of a drug-sniffing dog, it's difficult to claim that smelling marijuana is an infallible method of detecting drugs.

The authors of the study summarize their conclusions -- in part -- this way: "Did the detection of marijuana odors by AST investigators reliably predict the presence of legally prohibited quantities of marijuana?" and "(t)he answer to this question is both yes and no."

After three years, a lot of research time, and who knows how much money, experts have determined that when a trooper smells pot, there's pot -- except when there's not.

At least no taxpayer dollars were used.

Contact Ben Anderson at ben(at)alaskadispatch.com