Canopy closure estimates with GreenOrbs

Abstract

Motivated by the needs of precise forest inventory and real-time surveillance for ecosystem management, in this paper we present GreenOrbs [2], a wireless sensor network system and its application for canopy closure estimates. Both the hardware and software designs of GreenOrbs are tailored for sensing in wild environments without human supervision, including a firm weatherproof enclosure of sensor motes and a light-weight mechanism for node state monitoring and data collection. By incorporating a pre-deployment training process as well as a distributed calibration method, the estimates of canopy closure stay accurate and consistent against uncertain sensory data and dynamic environments. We have implemented a prototype system of GreenOrbs and carried out multiple rounds of deployments. The evaluation results demonstrate that GreenOrbs outperforms the conventional approaches for canopy closure estimates. Some early experiences are reported in this paper.