Hope Lies in Data Amid Mounting Water Crisis

The current state of humanity’s relationship with water is gut-wrenching.

Over two billion people live where the water supply cannot fully meet human needs, we’re depleting more than half of the world’s 37 largest aquifers, and by 2050, four billion people will live in water-stressed regions. With about 70% of freshwater withdrawals already used for agriculture,  freshwater availability in coastal areas will also decrease substantially as rising sea levels extend the salinization of groundwater.

Amid climate change, the documented impact of Earth’s shifting water cycle on economies and human health is severe.

In the American Southwest, Colorado River stakeholders, immersed in the worst drought in 1,200 years, have recently reached an agreement — after much wrangling — to conserve a massive amount of water over the next three years. But how the stakeholders will do it remains largely undefined. What’s more, some drought-prone states in the US don’t require utilities to audit water loss or report key kinds of its data. Everywhere, leakage rates in aging water infrastructure and municipal pipes are troubling: In Europe, the average water loss is 26% and, in the US, the figure ranges widely, from about 14% to as much as 60%. Some regions in other parts of the world lose as much as 70%.

In the face of these formidable challenges, accurately measuring water sources, distribution and consumption, and intelligently leveraging that data, are critical to sustainability. For almost every kind of water crisis we face, improved data management may sound dry, but it offers real hope.

Centralizing Decades of Data To Improve Water Management

States most at risk from the Colorado River’s low volume are under tremendous pressure to allocate water for human, agricultural, and industrial use in the most efficient manner possible. Water resources can’t be impactfully managed unless new measurements and existing historical datasets from groundwater, surface water, and recharge databases can be understood and accurately contextualized.

Centralizing data from the full array of distributed water resources with a modern data catalog — a Swiss Army Knife of data management tools — makes it usable. Intelligent data platforms automate a lot of processes that were done manually in the past and make it quick and easy to centralize, compare and analyze data. A system in one part of a state can compare itself to others of varying size and location to examine patterns and conservation measures the other systems are using.

By deploying intelligent data discovery, classification and visualization, machine learning can recognize relationships of the different structured data types like geospatial data, for example. Teams can quickly understand the complete lineage of datasets and level of reliability, allowing data scientists to focus on advanced analysis instead of basic data preparation.

Analyzing Sensor Data for Sustainable Water Use in Food Production

In agriculture, artificial intelligence models and advanced analytics from sensor data and various IoT devices are a game-changer, resulting in reduced water consumption and more uniform irrigation to better disease control, minimized nitrate runoff, and accurate prediction of future water needs and usage.

In Idaho, onion and hops farmers have implemented data-driven systems to make sure the right amount of water gets to each and every plant. To reach this level of detail, moisture sensor data from different depths in the soil, weather station temperature data and evapotranspiration data are being used to create detailed irrigation plans. Previously driving 400 miles a day to check monitoring equipment, farmers now can rely on hardware and software that issue alarms if there’s a change in irrigation pipe water pressure — meaning farmers can quickly zero in on corrective action to save time and resources.

Reducing Water Loss in Aging Infrastructure

Decaying municipal water and sewer infrastructure is resulting in significant leakage and safety hazards across the globe. Oliena, Italy, for example, was experiencing high water loss and intermittent supply. By prioritizing air and water pressure measurement and the prevention of pressure variations, they were able to create a more sustainable water network.

Today, municipalities have a litany of options. Small robots can be guided through pipes remotely to proactively find and analyze cracks and fissures. Gathering data on existing breaks, corrosion and wear patterns, the robots can be used to diagnose the extent of current problems and, with AI models, forecast future ones. Fiber optic cables can even be run through old pipes, with multi-sensing capabilities for detecting water level, conductivity, pH, gas content, and other variables.

A key approach with water infrastructure is to create a hypothesis and model of the specific problem. Then, use intelligent platforms to analyze comprehensive data from historical data sources and equipment like IoT sensors tuned for water infrastructure tasks to confirm or modify the hypothesis. Then act.      

Humanity can’t afford to stay idle as rapid changes on Earth, water scarcity, deferred maintenance and delayed problem-solving put populations at risk. It’s imperative that the public and private sectors come together to share best data management practices and normalize models that can help conserve the world’s most precious resource.