Unmanned aerial vehicle (UAV), open source software and piloting in REDD site

By Kathmandu University

15 October 2015
Chitwan Crown projection area image captured by the UAV Hexacopter UAV-Geomatics Project

Project at a glance 

Forest biomass is the sum of the above ground living organic material contained in trees, which is expressed as dry weight per unit area. Forest carbon trading has gained attention in recent years, for which practical methods for estimating forest biomass as well as carbon stocks are vital. The current trend in estimating above ground biomass and carbon stock is to use high-resolution satellite remote sensing images, like Geo-Eye, Worldview-2, and Quickbird. However, the shadow content present in such images and high solar angle cause major problems for precise carbon estimation. Potential cloud cover in these images limits their intended application and the use of such traditional photogrammetry is also costly.

This study used low-cost consumer cameras on unmanned aerial vehicles (UAVs) to estimate forest biomass and carbon stock. 

How does it work?

Kathmandu University developed hardware and software for geo-referenced images captured by a UAV. Images of forests are acquired using a UAV with an autonomous flight plan. A VIS-NIR camera is used to acquire forest canopy information. These images are then geo-referenced automatically using tie points and ground control points (GCPs) to create an image mosaic. An accuracy assessment of the image mosaic is then performed against the GCP measured using differential global positioning system (DGPS). Image segmentation and classification is carried out to extract crown projection area (CPA) from the image mosaic and digital surface model (DSM) created using images captured by the UAV. The CPA, DSM, and other field data about tree species, height, and diameter at breast height collected from the field are then put into allometric equations to estimate forest biomass and carbon stock. The estimated biomass obtained from the model is validated. 

Workflow system of UAV app for estimating forest biomass and carbon stock

Benefits of the application/tool

The Current trend in estimating above ground biomass and carbon stock is to utilize high resolution satellite remote sensing images like Geo-Eye, Worldview-2 and Quickbird. However, shadow content present in such images and high solar angle are major hindrances for precise carbon estimation. Potential cloud cover in these images limit their intended application and use of such traditional photogrammetry is also costly.

The UAV can acquire higher spatial and temporal resolution images than those available from satellite remote-sensing sources. As images can be acquired from UAVs at the time when the Sun is at its zenith, this reduces the problem of shadow content observed in high resolution satellite images. Similarly, the low flying height of the UAV (typically from a few metres above the ground to 300 metres) eliminates the problem of cloud cover, which is a major hindrance in satellite images for biomass estimation. It can easily provide 1–5cm resolution images with low-cost consumer cameras. 

The application uses:

  • Ultra-high spatial resolution images
  • Very high temporal resolution data
  • Cloud-free images with no or limited shadow effect
  • VIS-NIR sensor for calculating NDVI
  • UAV with autonomous flight plan and automatic take-off and landing

Cost implication for maintaining the tool

The UAV is a low-cost option compared to traditional photogrammetry. The only cost required is transportation and human resources for operation. The system itself does not have any additional costs as it operates on a rechargeable battery.

Researchers placing Ground Control Points

Upscaling the system in the future

This system is currently being used by Department of Agriculture, Department of Archaeology, Department of Forest, Department of Survey, and Department of Urban Development and Building Construction to estimate forest biomass and carbon stock. It is also being used by research institutes and I/NGOs for the similar purpose. In the future, its uses can be extended to other fields such as 3D mapping of archaeological sites, cadastre, post-disaster response planning, Urban mapping, planning and monitoring… etc. In addition, the project is planning to explore the integration of new and emerging sensors like LiDAR for 3D mapping.

Key partners

This application will find its use in many government and non-government agencies, currently, we see following agencies using the system:  

  • Department of Agriculture
  • Department of Archaeology
  • Department of Forest
  • Department of Survey
  • Department of Urban Development & Building Construction
  • I/NGOs