This week, a major snowstorm stretching from Texas to Maine is forecasted to affect millions of people. Many people throughout the U.S. now face the same questions. Do I have time to run to the store for more groceries before the heaviest snow is going to fall? Is school going to be delayed or canceled? Will it be snowing at the start of the football game?
Spatiotemporal data is data that changes across both geographic location and time. The radar map is a classic rendering of this type of data, tracking latitude and longitude coordinates with forecast timestamps to track the movement, intensity, and type of precipitation. In times like these, we all rely on these maps to help us make decisions.
Forecast of the likelihood of shaded regions receiving over 1 inch of snow in the January 25-26, 2026, storm over time. Source: National Weather Service.
In partnership with James Madison University (JMU), we’re integrating a novice-friendly suite of software into CODAP to explore time-based geospatial datasets (or spatiotemporal data) in our new National Science Foundation-funded Mapping Time project. Chad Dorsey serves as the Mapping Time Principal Investigator with Kate Miller as Co-Principal Investigator at the Concord Consortium, and Bob Kolvoord as Co-Principal Investigator at JMU.
Why students must learn to work with spatiotemporal data
Essentially all problems of consequence in modern society play out in the form of time-evolving scenarios with geospatial data at their core. From tracking Earth’s evolving climate to understanding the spread of disease or examining changes in global commerce and distribution of resources, changes in geospatial data over time occupy center stage. Today’s students must learn to work effectively with spatiotemporal data to draw conclusions and inform decisions.
However, identifying and analyzing time-based trends across geospatial datasets has long flummoxed researchers and teachers. The challenges are both technical and interpretive. Spatiotemporal data is difficult to analyze because of the inherent complexity of data that changes in multiple dimensions.
Effective learning with technology must allow learners to visualize, filter, and search through large, multivariable datasets in order to identify patterns, plot a variety of data types, and correlate across multiple variables. In terms of interpretation, learners often have trouble following their own lines of questions and inquiry, as well as attune to the specific scientific, social, and technical domains that spatiotemporal analysis and understanding demand.
Building on past work
Mapping Time builds on our NSF-funded Data in Space and Time project, which pioneered research into learners’ conceptions of spatiotemporal data through think-alouds with high school and college students and identified affordances of complex spatiotemporal representations.
The Data in Space and Time project designed and prototyped the Space-Time Cube, an interactive 3D interface that introduces time as a z-axis variable, representing temporal events within a three-dimensional cube, situated above or below their corresponding spatial locations.
Spatiotemporal data of tornadoes from 2018-2025 in the Space-Time Cube.
The map above shows all of the tornadoes that have occurred in the past seven years in North America. The dots’ color can be set to represent a variety of attributes about each tornado, including the length or width of the track, property loss in dollars, and much more. Here, the data is colored by intensity of each tornado, on a scale of 1-5.
Looking top down at the data in the Space-Time Cube (left), the visualization may look familiar. Many maps of the U.S. display data plotted on top. Because the Space-Time Cube includes the z-axis of time, it can be rotated and tilted (right), adding a new view of the three-dimensional data. In this view, temporal patterns and insights are revealed, and the data can be explored from multiple angles.
The Mapping Time project is developing a curriculum around the Space-Time Cube that scaffolds students to explore spatiotemporal data, find patterns and connections across datasets and visualizations, and make conclusions about trends in the data. The curriculum will be implemented with high school students enrolled in James Madison University’s Geospatial Semester (GSS). The GSS is a dual-enrollment course offered by JMU and school districts across Virginia in which high school juniors and seniors focus on geospatial technologies and apply those technologies to a range of inquiry-based projects, including a capstone project they design.
GSS teachers are co-designing classroom activities with Mapping Time project staff, aiming to identify both the technology features and approaches that will be most useful for integrating into the existing GSS curriculum and extending the learning objectives to encompass the time-based component of geospatial learning.
Research goals
The Mapping Time project is studying the design of innovative technologies for spatiotemporal data exploration and the affordances and methodologies that best support high school students in that exploration. We hope to advance understanding of the importance of and approaches to exploring and learning from time-based geospatial datasets. That way, we’ll all be better prepared to understand radar maps and other spatiotemporal data in the future.