The Create Space Time Cube By Aggregating Points, Create Space Time Cube From Defined Locations, and Create Space Time Cube From Multidimensional Raster Layer tools structure and summarize datasets into a netCDF data format by creating space-time bins that form a cube-like structure. Each bin in the cube contains a value, or count, of the number of events that occurred at the bin location for the time-step interval specified. Additionally, the cube may contain one or more summary fields or variables with statistics for the attribute fields specified in each bin.
There are a variety of analysis tools in the Space Time Pattern Mining toolbox that you can run against the cube after it is created. The Emerging Hot Spot Analysis tool takes the netCDF cube as input; runs a space-time hot spot analysis (Getis-Ord Gi* statistic); and identifies trends in the aggregated count data, summary fields, or variables, such as new, intensifying, diminishing, and sporadic hot and cold spots. The Local Outlier Analysis tool also takes the netCDF cube as input and runs an interpretation of the Local Moran's I statistic to identify statistically significant clusters and outliers in the context of both space and time. The Time Series Clustering tool identifies the locations in the space-time cube that are most similar and partitions them into distinct clusters in which members of each cluster have similar time series characteristics. There are a variety of forecast methods in the Time Series Forecasting toolset that allow you to forecast and estimate future values and identify temporal outliers as well as a tool to help you evaluate and compare the different forecast models created. The netCDF cube is updated with the results of these analyses in the form of several new variables.
The data and variables stored in the space-time netCDF cube can be visualized in either two or three dimensions using the Visualize Space Time Cube in 2D or Visualize Space Time Cube in 3D tools or by creating a space-time cube layer using the Make Space Time Cube Layer tool. The space-time cube can be visualized and explored in three dimensions in a 3D scene. Visualization of the cube is useful for several reasons:
- It can help you understand the structure of the space-time cube and how the process of aggregation into the cube works.
- It can also offer insights into the results of Emerging Hot Spot Analysis, Local Outlier Analysis, and the other Space Time Pattern Mining analysis tools, providing evidence that can help you understand the results.
- Additionally, visualizing the value of the summary fields, variables, or other display themes can help you understand how confident you can be in subsequent analyses by displaying the spatial pattern of empty bins that had to be estimated or temporal outliers in your analysis.
Visualize a space-time cube layer in 3D
To visualize the space-time cube in 3D, complete the following steps:
- Start ArcGIS AllSource.
- Open a scene. To open a scene so the results of the tool can be rendered in 3D, click the Insert tab, click New Map, and choose a new global or local scene.
- Open the Make Space Time Cube Layer tool, set the Input Space Time Cube parameter, and run the tool. This will add the space-time cube layer to the Contents pane.
- Select the space-time cube layer in the Contents pane. This will open the space-time cube ribbon.
- In the space-time cube ribbon, set the Variable parameter.
- Choose a Theme option for the selected cube variable. The themes that are available in the themes gallery will depend on the analysis tools you have run on the space-time cube. You can use the Describe Cube tool to determine which analysis tools have been run on the space-time cube.
Learn more about visualization display themes for space-time cube layers
- Change the displayed space-time cube by using the controls in the space-time cube ribbon to modify the time and range settings, elevation properties, and more. Use the tools in the ribbon to view time series charts and slice into the space-time cube visualization to view interior bins.
Visualize the space-time cube in 2D
To visualize the space-time cube in 2D, complete the following steps:
- Start ArcGIS AllSource.
- Open the Visualize Space Time Cube in 2D tool.
Note:
You can access the Visualize Space Time Cube in 2D tool from the space-time cube ribbon by selecting Add 2D Layer in the Utilities group of the ribbon. The tool will open with the source space-time cube automatically set as the Input Space Time Cube parameter value.
- Choose a cube variable to explore. After the cube has been created, variables stored in the cube include count, as well as any summary fields or variables that you chose to aggregate when creating the cube.
- Choose a display theme to display the selected cube variable. Examples of some display themes are as follows:
- Locations with data—Allows you to see all of the locations that contain data for the selected variable, and the Trends option shows where values have been increasing or decreasing over time (the results of the Mann-Kendall statistic run on the selected cube variable for each location). The Locations with data option is not available for space-time cubes using defined locations (or when aggregating points to defined locations).
- Hot and cold spot trends—Shows where hot and cold spot z-scores are increasing or decreasing over time (the results of the Mann-Kendall statistic run on the z-scores of the space-time hot spot analysis for the selected cube variable), and Emerging Hot Spot Analysis results re-creates the results you saw when you ran the Emerging Hot Spot Analysis tool. Both Hot and cold spot trends and Emerging Hot Spot Analysis results are only available when Emerging Hot Spot Analysis has been run on the selected cube variable.
- Local Outlier Results—Re-creates the results you saw when you ran the Local Outlier Analysis tool for the selected cube variable.
- Number of estimated bins—Shows how many bins were estimated at each unique location, allowing you to see whether there is a spatial pattern of places with missing values. If entire sections of the map have high numbers of estimated bins, those areas may be best left out of the analysis. Locations excluded from analysis shows those places that had data but had empty bins that could not be filled because they did not meet the criteria for estimation. Both Number of estimated bins and Locations excluded from analysis are only available for summary fields.
- Choose where to save the output features and click Run.