Processing and filtering of leaf area index time series assessed by in-situ wireless sensor networks

Abstract

A precise and up-to-date situational awareness of crop conditions is important for precision farming. The temporally continuous monitoring of relevant crop parameters has recently been shown to assist in a large number of applications. In this context, the leaf area index (LAI) is a key parameter. However, continuous LAI monitoring using traditional assessment methods is hardly possible and very expensive. For this reason, low-cost sensors based on Wireless Sensor Network (WSN) technology have been developed and interconnected to agricultural in situ sensor networks that seem promising for LAI assessment. In this paper, an approach for the processing and filtering of distributed in situ sensor data for a credible LAI estimation is proposed. This approach is developed based on a long-term WSN deployment in experimental plots with different wheat cultivars (Triticum aestivum L.) and water regimes. Non-negligible environmental impacts on radiation-based LAI assessment are also taken into account. A comparative analysis with a conventional LAI instrument shows that WSNs with adequately processed data gathered by low-cost sensors have the potential to produce credible LAI trajectories with high temporal resolution, that fit the dynamic crop growth process. Moreover, they are also shown to be able to detect yield-limiting trends and even to differentiate between individual wheat cultivars. Hence, those WSNs enable new applications and can greatly support modern crop management, cultivation, and plant breeding.