In September 1934, average levels of dissolved oxygen in the stretch of Willamette River through the core of Portland Harbor was functionally zero.
Healthy aquatic ecosystems require certain base amounts of oxygen available in the water column—otherwise fish, plants, and other organisms start to die off. If the water gets too low in dissolved oxygen it turns anaerobic and supports only organisms like those that thrive in swamps and produce methane and other odorous gases. This is the kind of stench that this guy was complaining about.
Euro American society succeeded in turning the vibrant and dissolved-oxygen-rich lower Willamette River into a smelly, filthy, functionally anaerobic environment in relatively short time. It only took them about seventy years: from the 1850s, when large-scale White settlement began, until the early 1930s. A river system that had been evolving over millennia to support a diverse array of plants, benthic life, and fish life—including the anadromous salmonids—had been converted into a stinking sewer over the span of just one human lifetime.
The image above illustrates clearly how unhealthy Portland Harbor was to aquatic life during the Willamette River’s annual low-flow period (July-October). I created this map in ArcMap using data from a September 1934 survey conducted by faculty and students from the Oregon State Agricultural College’s Engineering Experiment Station. The data set was a series of eight water quality samples taken during September 1934 to record, among other things, the levels of dissolved oxygen at seven sampling stations in Portland Harbor between Sellwood Bridge in the south and Gillihan’s Landing (Sauvie Island) in the north.
Applying this historical data using modern software tools provides a crystal-clear illustration that the Willamette River was so polluted (as measured in dissolved oxygen) as it reached Portland’s southern boundary that it was already hazardous to fish health. From this point it only got worse until about a half-mile from the Willamette’s confluence with the Columbia, where back-flow from the larger Columbia River counteracted the filthy Willamette River water to raise dissolved oxygen levels to nearly 10 parts per million (ppm).
Dissolved oxygen was not the only measurement of a polluted stream in the 1930s, just as it’s not the only measurement of pollution in the 2010s. However, dissolved oxygen and bacteria counts were the primary measurements of water pollution from the 1910s well into the 1960s. It was only in the 1960s—and particularly beginning in the 1970s—that scientists developed tools and methods to identify and quantify chemical, radiological, and other kinds of pollutants. Since the Superfund legislation in 1980 Americans are increasingly accustomed to hearing about water pollution in relation to PCBs, dioxins, petrochemicals, heavy metals, soil and groundwater contamination, & etc. The map above implies this difference between the pre-1970s and the post-1970s period of water pollution abatement: Advocates in this earlier period were reacting to anaerobic river conditions that killed fish and turned large sections of the river into a frothing, viscous, detritus-laden death-dealing stew of filth; advocates in the current era are reacting to bioaccumulative, persistent substances that are invisible to the unaided eye but that incrementally lead to cancer and a range of other severe, persistent, and often fatal health problems for humans and animals.
I find this map particularly interesting within the discipline of environmental history because I have yet to come across any similar representation or use of historical data in any of the hundreds of primary and secondary sources I have consulted as part of my research. Thus, this map provides a unique and important contribution to the literature of water pollution abatement. Another contribution to the historiography is that the map shows how GIS tools can represent historical water quality data in new and useful ways. It also provides a more tangible approach to representing graphed data in a way that often resonates with more people.
There were a few challenges in bringing these data points into ArcMap. One was that I had to decide how best to represent eight days’ worth of sampling in September 1934 at three depths at seven different locations in Portland Harbor; I decided to average all of these samples into an aggregate number for the month of September. Another challenge was that I had to locate the 1934 sampling stations and then determine geographic coordinates for these locations. With guidance from Steve Jett, my instructor at Portland Community College, I used Google Maps’ “What’s Here?” functionality to determine the latitude and longitude points for locations roughly in the center of the river channel across from the sampling station locations listed in the data source. With these coordinates determined Steve’s colleague Phil Miotto helped me turn the data into a shapefile in ArcMap; Steve then helped me turn these single data points into a raster data set to represent in a graduated manner the severe drop in dissolved oxygen levels as the river wended its way through Portland Harbor.
I consider this map a draft of one of the images I’ll be including in the book. There were two constraints I was working with in creating this map. The first is that I created it in grey scale because I likely won’t be able to include color images in the book. The second is that the book likely will be produced with 6″ x 9″ pages so my map needed to fit within these dimensions while also still having a margin.
Chat me up in the comment thread if you want to learn more about the GIS data and metadata I developed as part of this project.
 I transcribed this data from George W. Gleeson, A Sanitary Survey of the Willamette River from Sellwood Bridge to the Columbia River (Corvallis: Oregon State Agricultural College Engineering Experiment Station, 1936), pp. 20-21.
 Phil Miotto at AIM GIS Consulting, http://www.aimgis.com/index.html.
 Note: I added these constraints to this description June 23, 2013.