All Possible Rivers
Federico Pérez Villoro
All Possible Rivers consists of a computational model that aims to anticipate the future geomorphology of the Rio Grande / Río Bravo and, with it, of the U.S.- Mexico border. Taking into account the physics that participate in the large-scale evolution of the river, the program generates speculative landscapes by simulating complex interactions between water and land in specific places and across geological temporalities.
Concept
Using topographic scans from the US Geological Survey the simulator explores the ever-changing course of the river as a border technology. A pipeline of several computational steps calculates the capacities of water to erode land and of land to accumulate sediments. As the river’s waters push towards the inside of its meanders, the outer bends smooth out, generating an intensifying feedback loop that results in geometric transformations. The shape of the simulated rivers evolve over time using a dense grid of computational cells with locally-defined rules. The volume of water in each cell changes in response to the gradients of land and water surrounding it — carrying materials into or away from it. The flows of these materials affect such gradients where water determines where sediments concentrate while land affects water behaviors. The model was implemented as a user-friendly web application using Next.js and React, an open-source JavaScript library. It is hosted for free on Vercel.
The simulator arises along with the recognition of the Rio Grande / Río Bravo as a border technology. It is the result of a technical hypothesis projected into the future from an essay we wrote where we analyzed the constant movement of the river from its cartographic history — starting with the Treaty of Guadalupe Hidalgo when the body of water began to be industrialized as a territorial boundary. It studies the instruments with which we relate to space, but also how technologies constrain movement within spaces. The project is part of a larger series of explorations around the imposition of technical operativity upon living ecosystems and on the notion of landscape as a conceptual technology of domination.
Process
The simulator puts the past ahead of us as a permanent prototype of itself: a computational program designed to approximate the future geometry of the river as a discursive resource that visualizes the paradoxical condition of a moving border. We designed the simulator with expectation of use, as a tool to be transformed by others for concrete research purposes. Ultimately, the goal is to test its potential functionality as a device-to-think-with in tandem with journalists and with grass roots efforts organized around the politics of migration and against the many violences that are experienced along the border.
The simulator is a robust computational model developed with scientific groundings. Yet it is also an extremely low resolution image of the river when you consider the high dimensionality of the hydrosocial cycle. During the process it became critical for us to recognize such an impossibility to synthesize such a complex phenomena like the river, mathematically. The ever-evolving river escapes any attempt to be represented — to be mapped. The simulator, thus, is yet one more imprecise effort, an approximation yet a dynamic program that allows us to relate to the river, and with it, to the border as a permanently evolving network rather than a static continuous line.
Lessons
It is difficult to escape the trap of representation. We think of our model as a tool to relate to the river beyond the temporal and spatial scales within the immediacy of human perception. We wanted to move away from the capturing impulse that underlies most cartography efforts and allow the river to be expressive in its own terms. Yet our simulator reduces the river’s complexity to the tool’s technical limits, mirroring the qualities of mapping instruments and, with it, of technologies of domination. For instance, in our expectations of seeing the landscapes evolve in certain ways and across geological temporaries, we didn’t find visualizing frameworks that would satisfy meaningful analysis. We decided to not provide such informational keys and embrace more open-ended interpretations. The further away the simulation moves into time, the less accurate the images are. This is a provocative paradox: the more simulated cycles of information the program processes, the further away from realizability the outcomes are.
