An avanced framework for semantic querying of the dynamic world dataset

What is the 'Dynamic World'?

This is a LULC dataset provided by Google and the World Resources Institute. The Dynamic World dataset offers near real-time global Sentinel-2 land use/land cover (LULC) mapping, generated using a Fully Convolutional Neural Network (FCNN).

Objective

The main objective of this project is to showcase how semantic querying can enhance the analysis of the Dynamic World dataset beyond simple mode-based temporal reduction. By employing semantic querying techniques, users can perform more complex spatial and temporal analysis, unlocking new insights and possibilities for utilizing the dataset.

Methodology

The implementation utilizes the semantique Python package for semantic querying by defining the layout, mapping, and query recipe for the semantic querying process.

Concept

Preliminary results

Future Directions

• Integration of semantic querying framework directly into Google Earth Engine for seamless analysis.
• Testing the dataset and semantic querying approach in various applications to explore its full potential.

References

[1] C. F. Brown et al., “Dynamic World, Near real-time global 10 m land use land cover mapping,” Sci Data, vol. 9, no. 1, p.251, 2022, doi: 10.1038/s41597-022-01307-4.

[2] M. Sudmanns, H. Augustin, L. van der Meer, A. Baraldi, and D. Tiede, “The Austrian Semantic EO Data Cube Infrastructure,” Remote Sensing, vol. 13, no. 23, p. 4807, 2021, doi: 10.3390/rs13234807.