Sacramento River Paradox

Globally, overcapitalized regions extract more than their fair share of resources from undercapitalized regions. Water, the most vital resource after air, is no exception. Most residents in these regions go to bed thirsty every night, only to face a new fight for water in the morning. Conversely, most residents in overcapitalized regions need only take minimal notice of where their seemingly inexhaustible water supply comes from (so far).

With the "Sacramento River Paradox" Niemeyer seeks to direct our attention to the awesome Sacramento River, which quenches the thirst of 1/3 of all California residents, grows their food, and washes their bodies.

In a set of three images, Niemeyer visualizes the river's flow and water level throughout the Winter 2023 season. Flow and level are the result of snowmelt, temperature, wind and rainfall, connecting the most remote regions of California with the most densely populated urban centers. The images show this data through novel data visualizations. The left image in the triptych shows an AI forecast, the right image shows actual measurements, and the middle image shows the difference between the two.

These nuanced representations of what we think will be and what actually is direct our attention to the life of the Sacramento River, the vital resource in our midst. Niemeyer intends for the images to serve as a portal of respect and gratitude for what lifelines the Sacramento River throws us every day. Only solid respect for local resources can inform equal respect for similar resources in other regions around the world.

Sometimes we give the least attention to the most important things.

Background

Since 2022, Niemeyer runs a Hydrocolonial Institute with Patrick Owuor and several other founding members. Our goal is to recognize the many flows of the many waters in our world in both scientific, social, techincal and poetic ways. We conduct long-term research projects in several regions, including California. Greg’s intention with the “Sacramento Paradox” is to study how water futures are informed by cultural, social and economic values. Of particular interest is our growing ability to predict water conditions, and therefore water security with AI. For this project, Niemeyer used TimeGPT to generate a forecast for several water condition channels related to the Sacramento River, including snowpack, rainfall, windspeed, temperature, flow and gage. The forecast is trained on decades of historical data which Niemeyer collected from openweather.com, the USGS, NOAA, and the CEDC. All data was processed in Python. Each channel includes 2160 data points for the hours included in the astronomical Winter season for 8 water measurement stations along the Sacramento River, from Mount Shasta to the San Francisco Bay.

While that seems like much data, it still is a sparse sample of the rivers’ complex flow and interaction with every chemical, every organism and every geological feature on its course. To emphasize the sampling, Greg Niemeyer interpolated the sparse data to represent the long and connected flow of the river. The resulting image is 2160 pixels wide and 2700 pixels tall. Each channel produces one black-and-white (L) image, which Greg uses as an alpha mask to control the amount of color that the channel contributes to the final image. Layered on top of each other, the colors add up to a rich representation of some of the seasonal and hydrological dynamics of the Sacramento River.

The forecast looks very different from the measured data. It shows the cyclical dynamics and the expected means somewhat accurately, but does not capture the wild swings of rain and wind. The measured data includes many errors and gaps, and it is a sparse sample. The difference between the forecast and the measurement reflects qualities of both, reflecting a possible future onto a partial past.