09-10, 13:00–13:30 (America/Chicago), Grand C
The STAC render extension offers a flexible schema for managing dynamic tiling rendering metadata. Come discuss how the NASA VEDA (Visualization, Exploration and Data Analysis) open science platform has used the render extension to simplify front end visualization.
Dynamic tiling of raster data offers the benefit of on the fly access and transformation of static cloud native files that can then be visualized on a map. Development Seed's titiler is one such example of a dynamic tiling server capable of creating tiled mosaics from Cloud Optimized GeoTiffs (COGs). Rescaling, applying a colormap or band arithmetic for example are all possible by passing certain metadata as path parameters to a tiling endpoint. Storing these parameters is helpful for front end applications and clients may have custom application logic for dealing with dynamic tiling parameters. When working with a large number of datasets, this makes it cumbersome to manage the dynamic tile rendering metadata with the datasets themselves.
In the STAC ecosystem we've proposed and implemented the render extension that offers a flexible schema for managing dynamic tiling rendering metadata. STAC Collections or Items may include custom rendering metadata that can be used by a client to visualize the dataset where these parameters are needed. The extension was implemented in NASA's VEDA (Visualization, Exploration and Data Analysis) open science platform and allowed the creation of dynamic thumbnails and visualization of STAC Items. It also allowed the offloading of dataset specific metadata from the frontend application code base to the STAC Catalog. The tighter coupling of the render parameters with other dataset metadata via STAC greatly reduces the effort in managing those parameters in front end application logic and minimizes data inconsistencies.
