The Water Mark: Visualizing Data
Recently I was thinking about Water Quality Indices (WQI) which are used to broadly compare different water sources on a standard metric and a color chart. However accounting for changes in water quality is a fiendishly difficult task. Chemical composition is much too complex to be represented on a single scale. Also, on a WQI the arithmetic of additive calculation is insensitive to the effect a single bad parameter value, so a critical rating on one variable may be masked when aggregated with “good” ratings of other variables.
I decided to turn the water quality index on its head and explore the opposite concept. What if we were to chart every available attribute of every measurement, in a lurid representation of everything all at once? I was curious. What ways can communicate hopefully accurate & complicated data about water in a way that people can understand to take effective action?
Data, in my water?
Environmental data can quickly become overwhelmingly complex. There are over 20 variables to consider in water composition alone. Some are usually inconsequential, but sometimes a single one at an inadequate level can render the water unusable (or elicit lawsuits).
We’re talking about the presence and interactions of temperature, conductivity, pH, BOD5, turbidity, total solids, phosphates, fecal Coliforms, pesticides, sodic compounds, cations (Ca, Mg, Na, K), anions (sulfates SO4, HCO3, Cl, CO3, nitrates N-NO3) and the motely crew of B, Fe, Mn, Cu, Zn, Mb, As, Ni, etc.
Throw in information about soil composition (biotic and abiotic), climate, human interactions, intended use and geography, and you have an ocean of data to sort through which may also need to be georeferenced or on a timescale.
People need immediately extractable insight from scientifically complicated data to make smart decisions.
Dynamically distilling the information depends on who will consume it. Once data is round up and ground up one needs to condense and communicate it in different representations that elucidate precise insight to specific audiences. Is it government, the public, engineers, emergency responders or humanitarian organizations? All this without discarding information that is secondary to the problem at hand but is crucial to the documentation and must be easily accessible à la three-click-rule.
Compare these two water quality lab reports from 1998 and 2017. We see increased capabilities of Excel spreadsheets but still the same paradigms on the way this information should be reported.
We know the chemical composition. So what? The importance of each factor depends on the intended use. Some labs take into account intended use and issue “report cards” on a color scale, but surely more could be done.
Take Sodium for example. A concern in agriculture is that berries are highly sensitive to sodium. Meanwhile corn could care less about it, up to a point, after which soil may oversalinate and become useless for growing food. The issue highlights the intimate relationship between soil and water and the importance of thought-out irrigation methods. It’s a major talking point in the UN Convention to Combat Desertification.
Proper visualization can help close the cognitive loop between decisionmakers and their information. It is the difference between correct critical decisions and life as usual.
A story about water
Let’s create tools that use the “broad bandwidth of the human sensory system in steering and interpreting complex processes involving voluminous datasets across diverse scientific disciplines.” (Hagen)
David Warner, famous for trading “Beer for Data” in Afghanistan, developed a tool for spatial visualization of data, named after its original use to study social behavior of ants. ANTz is an open source “3D data visualization software for the mind”. It’s an experience of spatial reasoning where one can see rings of red, purple and blue hover and rotate over Google maps.
Imagine if you could visualize a 3D map the purpose, composition and availability of each body of water in a particular geography. Where the meaning of information is not only on a color scale but also varies in shape, animation, speed, size, and placement. Where a non-expert could learn what it’s all saying to make informed judgements.
Imagine if integrating frequent water quality analysis with multispectral & high resolution satellite imagery, on-ground sensor data, meteorological predictions, expected demand and real-time flow-rates.
What if we could intuitively feel the pulse of the planet as if rivers and aquifers were its arteries?
Digging into the data
The field of Hydroinformatics combines water data analysis with visualization tools to translate the complicated into the simple. It develops tools that set benchmarks to assess relative risks of specific uses, evaluate the success of management strategies, and assist decision-makers to best allocate resources.
In a couple lines of code one can automate interpretations that dig deep into data and merge the different pieces of information to create a narrative around something that would otherwise take experts and hours to understand. This could be done, for example, based on WHO water and sanitation guidelines, fundamental knowledge crop requirements, risk assessment, and general engineering considerations that already inform decision-makers.
What does the future look like?
Over the next posts I will explore different tools that might help us think more clearly about water.
Color-coded tables and infographics still have their place, but we can increasingly use interactive 3D spaces and statistical analysis to help sort through overflowing spreadsheets. (David’s current work involves incorporating sound cues to VR and AR data visualization.)
Let’s use an integrated approach to make better decisions about the sustainable use, protection, improvement all water.