If you wanted to know more about cyanobacteria blooms, those nasty bursts of microscopic creatures that can turn a lake into a scum-topped mess, where would you look?
It’s obvious: Census data.
OK, so that’s not exactly obvious. For Ken Johnson of UNH’s Carsey School of Public Policy, this is why it’s so cool.
Johnson has spent his career parsing census data to study demographic changes around the world. I’ve quoted him many times on topics like New Hampshire’s silver tsunami (lots more old people!) and migration patterns (Massholes keep coming!). I will probably quote him many more times in the future. But I never expected to do it concerning organic water pollution.
Johnson says he’s a bit surprised himself. “I know a lot more about satellites and drones and marine biology than I used to,” he said.
Blooms of cyanobacteria, sometimes inaccurately called blue-green algae, are a growing problem in New Hampshire’s ponds and lakes. Warm water and too much nitrogen from runoff or (ewwww) leaky septic systems causes the population to explode, creating a mess than can be dangerous when the bacteria die and release toxins into the water.
Knowing how to control the problem requires knowing when and where it will occur, which in turn requires a detailed knowledge of what the shore is like. There are a skadillion miles (my estimate) of freshwater shoreline, a.k.a. the riparian edge, in New Hampshire, so that’s a non-trivial problem.
“The typical way in natural sciences is to use remote sensing platforms that produce a national land-cover database. It’s for the whole country, updated every few years, classifying land cover into a variety of categories,” Johnson told me. “The problem is when we looked at the lake shores, they all looked heavily forested, but we knew there was actually a lot of housing along the lakes — cottages, camps, homes that don’t show up on the satellite imagery.”
Enter census data, broken down into the smallest unit known as block level.
“Combining the land-cover data from satellites with census data measuring housing, we found a lot of the areas along riparian edge actually have fairly dense housing. The national land-cover data would call this land forest, but actually, they have houses in them,” Johnson said.
That sounds straightforward, but anybody who has tried to join different datasets knows it can be a nightmare. Data based on satellite imagery is about as different as you can get from head-counts collected door-to-door. Merging them to make conclusions took a lot of twisting and turning.
“We did it for 10 full counties — five in Great Lake states, five in New England. It was a lot of work to overlay these,” Johnson said.
The team from UNH and Dartmouth included Johnson, a demographer, plus a forest scientist, an aquatic biologist and a GIS analyst and was funded by the disparate duo of NASA and the U.S. Forest Service. For Johnson, that mix was the best part of the project, letting him doing science with folks in fields that exist in different campus buildings and never overlap.
“Normally demographers aren’t going to talk to remote-sensing experts or water quality experts. But multidisciplinary research can produce findings that are important… and neither would find on its own,” he said.
The work was published in Ecosphere, an open-access research journal, titled “U.S. Census data reveal high-density housing hidden under forested land cover along lake shorelines.” The big conclusion is that we have underestimated how much development occurs on the shores of lakes which are surrounded by woodlands and thus don’t realize the effect of development on water quality.
“We hypothesize that analyses of lake water quality based solely on the [National Land Cover Database imagery] may substantially underestimate the impacts of humans on forested lakes,” is how the report put it.
In theory this should give greater impetus to efforts to control development along lake shorelines and reduce future bacteria blooms, although stopping people from mowing right down to the water or dumping fertilizer all over the place is a slog.
But if you’re swimming this summer in a New Hampshire lake and it’s nice and clean, take a moment and thank the U.S. Census Bureau. If nothing else, you’ll confuse your friends and that’s always fun.
The paper can be seen at esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.70550
