# A high-resolution canopy height model of the Earth

**Authors:** Nico Lang, Walter Jetz, Konrad Schindler, Jan Dirk Wegner  
**arXiv:** [2204.08322](https://arxiv.org/abs/2204.08322)  
**Date:** 2022-04-13  
**Summary page:** https://www.summarizepaper.com/en/paper/2204.08322v1/

## Key points

- Authors developed a high-resolution canopy height model of the Earth
- Model emphasizes crucial role of vegetation height in global carbon cycle and ecosystem functioning
- Need for geospatially explicit and highly resolved information to manage terrestrial ecosystems, mitigate climate change, and prevent biodiversity loss
- Innovative approach involves fusing data from space missions like GEDI and optical satellite images through a probabilistic deep learning model
- Method reduces saturation effect in estimating canopy height and resolves tall canopies with high carbon stocks
- Only 5% of global landmass covered by trees taller than 30 m
- Model provides consistent and uncertainty-informed mapping globally for ongoing monitoring and decision-making processes
- Research implications extend beyond forest conservation to advance climate, carbon, and biodiversity modeling
- Study identifies areas with tall canopies not within protected areas (only 34%), highlighting need to incorporate detailed information into conservation strategies

## Summary

The authors of the groundbreaking study - Nico Lang, Walter Jetz, Konrad Schindler, and Jan Dirk Wegner - have developed a high-resolution canopy height model of the Earth. This model sheds light on the crucial role of vegetation height in the global carbon cycle and ecosystem functioning. The need for geospatially explicit and highly resolved information to manage terrestrial ecosystems, mitigate climate change, and prevent biodiversity loss is emphasized by the researchers. Their innovative approach involves fusing data from dedicated space missions like GEDI and optical satellite images like Sentinel-2 through a probabilistic deep learning model. This allows for accurate retrieval of canopy height from satellite images worldwide. Not only does this method reduce the saturation effect commonly encountered in estimating canopy height, but it also enables the resolution of tall canopies with potentially high carbon stocks. Surprisingly, their map reveals that only 5% of the global landmass is covered by trees taller than 30 m. This highlights the importance of such data for conservation efforts. Furthermore, their model provides consistent and uncertainty-informed mapping globally. This supports ongoing monitoring to detect changes in vegetation height and inform decision-making processes. The implications of this research extend beyond forest conservation as it has the potential to advance climate, carbon, and biodiversity modeling. By identifying areas with tall canopies that are not located within protected areas (only 34% according to their findings), this study underscores the significance of incorporating such detailed information into conservation strategies for preserving biodiversity and mitigating climate change impacts on ecosystems worldwide.