The Umbrella Conundrum: NETL & Uncertainty Visualization

We may not always be aware of it, but uncertainty is a constant part of our lives. It figures into decisions as major as where to invest our retirement portfolios and as minor as if we should grab an umbrella on our way out the door.

When the weatherman says there’s a chance of rain, when do you decide to bring that umbrella—when there’s a 30 percent chance of a drizzle, or when you’re 70 percent likely to get soaked?

Understanding likelihoods and the range of potential error in a prediction is important in everyday life, but it’s vital in research. Data analysis is an integral part of modern life, providing a foundation of knowledge for decision making in nearly every sector—from economics to energy. However, data is inherently uncertain. No matter how carefully gathered information is or how accurate the testing, no measurement is exact. The credibility of the scientific, technical, management, and policy decisions that are based on the understanding that data needs to include a perception of the accompanying uncertainty and error.

Recognizing this need, researchers at NETL created a technology to help tackle the challenges presented by accurately representing data uncertainty. The Variable Grid Method (VGM) developed by NETL focuses on the visual quantification and representation of spatial or spatial-temporal data trends—graphics that help illustrate information associated with the data analysis of geography, space, and time, like the constant changing of a weekly weather forecast across a map. Spatial data plays a large role in informing decisions related to today’s critical issues, such as energy security, weather forecasting, and human health.

Through the use of computer modeling, VGM allows users to measure the uncertainty associated with their datasets and create a visual representation of both the data itself and its range of error. This technology is not new, but previous tools lacked an important function—the ability to include the role of uncertainty and error, enhancing the accuracy of the data, prediction, or model visualization. By providing information on the unknown, users immediately have a better understanding of the data as a whole, improving the foundation on which to make decisions. After all, when the uncertainty is left out of the equation, it can be pretty difficult to decide if you need to bring an umbrella!

Important decisions can hinge on the provision of uncertainty in data. Spatial data can influence decisions that impact your life every day—decisions that can only be improved by understanding every scrap of information available.

VGM offers an alternative way to visualize knowledge gaps and uncertainty. The modeling tool enables users to represent ranges or categories of uncertainty as variable grid cell sizes.