Putting history on the [counter]map: sculpting the Loire average monthly flow data from 1964 to 2022

From the previous datasculpture “The Flood Necklace” as a first counter-map proposition, and after discussing with geologist and stratigraphist Zoltan Sylvester about his 3D river meandering model through time, emerged the question of how to give, in one glance, a physical and visual representation of the complete history of one’s river.

Extending the reflection started from flood data

My first attempt focused on the Loire’s floods, as data were easy to find and because it gives an interesting insight into the major events that recently impacted human societies and local communities, and for which we have records. But as Zoltan Sylvester stated, it is a common bias in the field of stratigraphy to focus more on the biggest events, as they left quite visible and interesting sediments behind, than it is to study centuries of regular, mundane water flows, that are also trickier to read with precision.

I wanted to reflect on that bias and think about what a full history of the Loire river could look like by focusing on regular monthly water heights instead of major flooding events. The two sculptures together would carry two different sets of data but complement each other as counter-map propositions carrying knowledge about water heights’ history.

I found monthly water height data from January 1964 to December 2021 on the French public platform HydroPortail. Of course, that is not comparable to the geological timeframe, but as a human geographer, it gives me enough insights to be relevant at the scale of human societies. It also carries interesting information about the age of our own data records and about the way institutions have recorded, kept and open those data to the general public in recent years.

A clear “dryer” season in July and August is visible from the data.

For this datasculpture project, it seemed more difficult to work with manual tools and clay than I did for the Flood Necklace. I got help from my father, now a retired engineer, who wrote a python script to convert the data into 3D shapes in Fusion 360. From there, the files were converted to GCode to be readable by a CNC machine.

If you are interested in the detailed process or want to replicate the steps, all the files are available on GitHub (comments and explanations are in french – I can provide a translation in English upon request).

Building of the first prototype

We built one first prototype from the files, using the CNC machine on Medium wood, at a small scale. The bottom disk is acting as a map legend, for the reader to find their way around the time of the year while looking at the data (this might change in future versions). The top value is 1200 M3/s for the few years represented here, so only one data point goes above the 1000 limit visible on the disk.

Bigger values are concentrated in the winter months (December to February for most years), while in the summer the flow is very low. It even stopped completely in some sections of the river during the 2023 summer (data here represented go from 2000 to 2019).

The line in the middle of the shapes is due to technical constraints, to allow the CNC machine to drill the center hole, to stack the disks which was the initial idea for this project.

Some years are missing in the prototype for now, as a few data points were missing from the Hydroportail data. We are reflecting on how to proceed: either not represent the year at all (but sacrificing 11 valid data points) or extrapolate the missing value from previous and following years, knowing it won’t reflect reality.

Looking at the sculpture from above, the river appears clearly as having two seasons: the high flows between January and June, and the lower flows from June to December

What’s next? A second prototype including all years from 1964 on a bigger scale is in the making… I will update this post with more picture as soon as it is ready.


OpenEdition suggests that you cite this post as follows:
Anne-Laure Fréant (April 2, 2023). Putting history on the [counter]map: sculpting the Loire average monthly flow data from 1964 to 2022. Datartefacts. Retrieved July 19, 2024 from https://doi.org/10.58079/nh5b


Search OpenEdition Search

You will be redirected to OpenEdition Search