It's 2:00 on Friday, the temperature is right around freezing, and fog is just starting to gather. Eric Wirkman, Teen Feed Coordinator at the University Street Ministry, has just offered me an extra scarf that he brought along just in case. Wirkman, a One Night Count veteran, is the leader of the counting team I will be following around for the night. One Night Count is an annual attempt to make an organized tally of homeless people living outside. It is led by the Seattle/King County Coalition on Homelessness (SKCCH) and numerous other nonprofits and government entities around the county.

Between 2 and 5 a.m., the counters tally 2,826 unsheltered homeless people across King County—a two percent increase over last year's count. Strangely, the 2009 count for the city of Seattle was exactly the same as last year's: 1,976. Both numbers will almost certainly be used by both supporters and opponents of the Metropolitan King County Council's 10-Year Plan to End Homelessness. Most agree, however, that the count probably underestimates the number of homeless people in King County, because it's impossible to survey the entire county, because people tend to hide, and because counters don't include abandoned buildings in the count.

The most interesting change from last year's count was the sharp rise in the number of unsheltered homeless people living in south King County. Federal Way saw an increase from 90 to 116, and Kent's numbers rose from 65 to 193, a nearly 300% increase. Overall, there were 68 percent more unsheltered homeless people living in South King County this year.

Earlier in the day, volunteers conducted a pre-count to give them a sense of which areas to focus on during the count itself. The count always takes place in the winter, because this gives the best sense of how many homeless people live outside in King County year-round (as opposed to those who move on to shelter or warmer climates in the winter).

There are numerous obstacles to conducting an accurate count, Wirkman says. Unsheltered homeless people who are caring for children may go to great lengths to hide in order to avoid Child Protective Services. The city of Seattle also has broad powers to sweep up homeless encampments and confiscate personal property. Fear of reprisal from the city has driven many homeless people further into hiding or, conversely, into encampments like Nickelsville, which is aimed at making people more aware of homelessness. And some predators target the homeless—another reason to go into hiding. Wirkman says some have even taken to living in trees, because "people don't look up."

The volunteers operated cautiously to avoid accidentally counting a homeless person twice. However, they acknowledge that the process is far from scientific. At one point, while crossing into Ravenna Park, we hear a man coughing in the undergrowth. We never actually see the man, but several other volunteer teams also report hearing him. For the rest of the night, Wirkman and the other team leader debate the validity of counting a cough and triple-check to make sure that the man wasn't counted twice.

The disembodied cough turns out to be the nearest we come, in three hours, to encountering an actual homeless person. Walking the trails in the ravine by Ravenna/Cowen Park, Wirkman and I do find a soiled foam pad that has someone has obviously slept on, but Wirkman doesn't add it to the count because there's no obvious structure nearby. Structures, which include tents or more permanent buildings, are counted as two people. However, even though the greenbelt we were searching had a large homeless population as recently as half a decade ago, we find no structures. Even with the earlier pre-count and Wirkman's veteran status, our nocturnal search takes a wrong turn and my guide, frustrated, notes that we are back on "the same trail we were just fucking on." SKCCH spokesman Joshua Okrent tells me count organizers reserved Ravenna Park, The Jungle, and other greenbelts throughout the county for "younger, more agile volunteers." That's me.

Which brings us to the aforementioned sex. Two volunteers, assigned to the U-Village area, reported seeing appeared to be a human-shaped mound. The volunteers were initially unsure whether to count the mound as a single person or two. The sounds of two people copulating answered that question. "The people we found," one of the volunteers surmised, "were doing it."