3D Modeling for Visual Analyses of Recent and Historical Weather and Water Patterns

In recent decades, multidimensional models became a common visualization technique in many disciplines, especially in statistical analyses, 3D web graphics, and topography map generation. While 3D modelling offers a unique benefit to represent trends and patterns of large data sets, its application in some sectors of meteorology and climatology has remained limited.

This paper aims at extending research in this field by providing examples and delineating the process of utilizing various weather-related data sets to generate 3D models for geospatial and temporal analyses of diverse weather and water patterns. As 3D modelling can take different forms, this research specifically demonstrates meteorological analyses of recent and historical weather and water patterns in Oklahoma and Texas by means of physical, digital, and interactive 3D models. The following examples will be presented:

1. Visualization of radar rain estimations in Oklahoma with a comparisons analysis of year totals of the 2014 drought year and the 2015 record rain year (both physical and digital 3D models will be presented and explained).
2. Visual representation of groundwater levels across Oklahoma and Texas identifying trends with visual ques (live digital and interactive 3D models based on GIS maps will be demonstrated).
3. Comparison of multiple weather sensors to represent correlations over a time span of multiple years in an interactive 3D map for a variety of weather and water variables. This kind of visualization can be broadly utilized in interdisciplinary research. We will present its practical application to our ongoing socio-economic and environmental evaluation of weather and water resources in Oklahoma.

Data sets from the Oklahoma Mesonet, radar rainfall estimates from National Weather Service, and groundwater levels from Oklahoma Water Resources Board and Texas Water Development Board were used to generate various physical, digital, and interactive 3D models. C++ computing language was used to represent the data sets in digital 3D models, while we developed the Ternary Visual Shape Logic method to statistically visualize and analyze groundwater data sets with an interactive 3D model. Also methods will be presented on how the physical 3D models of geospatial rain patterns over time were designed and created.

All three forms of the 3D models presented here (physical, digital, and interactive) allow for geospatial and temporal analyses of weather and water related changes and patterns across the states and across the state boundaries. Due to their practical applicability the 3D models can serve as a useful tool both for educational and outreach purposes as well as for decision-making support. They facilitate a better understanding of otherwise very complex data sets. The 3D perspective further promotes a geospatial meteorological analysis, which is important for better comprehension and comparisons of two or more 3D models from equal time periods. Finally, this research also sets a background for new software that will allow for automatic weather data storage in 3D digital viewing formats, which will then facilitate 3D printing and/or transfer into virtual reality systems.