Re: [sdw] [SDW Best Practices Update]: Discuss Tiling in the BP (#1288)

From the document, I took some time to try and answer these questions.
Why we tile data? Tiling in itself is one big best practice for the web. Tiling data geospatially is essential to manage and deliver large datasets on the web. Tiling involves dividing geographical space into smaller, manageable ‘tiles’ or subsets, which can contain raster images, vector data, point clouds, or 3D models. These tiles enable various benefits, such as optimizing data storage and retrieval, supporting parallel processing for data analytics, and enhancing web map accessibility by enabling alternative rendering styles. Tiling provides a versatile and scalable solution for handling geospatial information effectively in web applications.
Where did it come from? Tiling data for web mapping and geospatial standards was initiated and developed by organizations and communities working on geographic information systems (GIS) and web mapping. It evolved as a best practice in the geospatial industry rather than having a single individual or organization credited with its inception.
Here are the specific considerations for tiling summarized:
1. Use of Generic Tile Containers: Adopting tile container formats like COG (Cloud Optimized GeoTIFF) and PMTiles, which support HTTP range requests for efficient access to individual tiles, is recommended for serving various types of geospatial data.
2. Efficient Vector Tile Encoding: Emphasizes the efficiency of vector tile encoding, which quantizes geometry coordinates to a grid at an appropriate resolution, making it suitable for visualization while also achieving good compression.
3. Tiling Point Clouds: Tiling point cloud data is advised to handle large, dense point cloud datasets by setting an upper limit on points per tile and using clustering or filtering algorithms for lower resolution tiles.
4. Use of Efficient Formats: Recommends using efficient formats like LAS/LAZ for point cloud data and Cloud Optimized Point Cloud (COPC) based on LAZ for HTTP range access.
5. Tiling 3D Data: Suggests tiling large-scale 3D environments using standards like OGC 3D Tiles and I3S to efficiently exchange and visualize multi-resolution 3D scenes.
6. Accessing Data via OGC API: Promotes the use of OGC API standards for accessing geospatial data as REST Web APIs, enabling consistent and modular access mechanisms for different types of data.
7. Raw Data Tiles for Accessibility: Proposes using raw data tiles (e.g., vector tiles and gridded coverage tiles) to enhance web map accessibility, as they allow for alternative rendering styles, larger fonts, and improved accessibility devices for users with disabilities.
8. Parallel and On-Demand Processing: Highlights the role of tiling in facilitating parallel and on-demand processing for large geospatial datasets, aiding memory management and distributed processing.
9. Interoperability with DGGS: Discusses the use of Discrete Global Grid Systems (DGGS) and their integration with tiling practices to support spatial queries efficiently.
10. Data Partitioning and Optimization: Recognizes the practice of partitioning geospatial data into tiles as an effective means to optimize data storage, retrieval, and processing.


-- 
GitHub Notification of comment by wilkesg
Please view or discuss this issue at https://github.com/w3c/sdw/issues/1288#issuecomment-1793000319 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Friday, 3 November 2023 19:40:05 UTC