Our New 3D Canopy Height Model

It is common knowledge that earth observation data is gathered at the speed of light. But did you know that data is now being published at the same rate too? The paradox of choice has (finally) hit earth observation (EO).

For any given remote sensing topic, you can find an increasing number of active and passive systems at various wavelengths. Not only is the availability of data growing rapidly, but the number of methods to handle each data source has also exploded, leading to increased complexity in EO applications. As each data source and the resulting derivatives are typically of sufficient quality, decision making and the harmonization of data become the new challenge. While an individual map might reflect a beautiful product of applied science, sets of maps generally turn into disjointed clumps rather than bouquets.

At TransparenC, we are constantly experimenting ways to optimize data harmonization. Most recently, we developed a prototype to leverage multiple datasets to increase the accuracy of forest mapping.

Following a set of key criteria, we have aggregated and visualized compelling data sources to achieve immersion. Our main criteria is reproducibility, which means we can apply this method globally regardless of region and timestamp. We aim to further develop this approach to include multi-temporal instances and introduce intuitive confidence metrics.

Feedback is always welcome!