LAI-NOS: An automatic network observation system for leaf area index based on hemispherical photography

The leaf area index (LAI) is an important indicator reflecting the growth status of vegetation and is widely used in agriculture, ecology, climate change, and other fields. The shortcomings of the currently available methods for manually measuring LAI include labor-intensive, low sampling frequency, and asynchronous data collection. Focusing on these issues, a LAI sensor based on hemispherical photogrammetry and an automatic network observation system (LAI-NOS) for LAI were developed, which consists of four parts: LAI sensor, sensor node, sink node, and online data management system. The LAI sensor measures LAI values based on hemispherical photography. The sensor node is responsible for controlling the sensor and obtaining the data measured by the LAI sensor. The sink node is responsible for local networking and communication with the remote server. Data storage, data management, data display, and sampling frequency are managed by the online data management system. Comparative studies with LAI-2200C and satellite products were also conducted in this study. The comparative study with LAI-2200C showed that the LAI measurements of different vegetation types from both sources were highly significantly correlated whether based on Pearson regression or Passing & Bablok regression. A preliminary study comparing LAI-NOS measurements with Sentinel-2 inversion LAI and MODIS LAI products (MOD15A2H) showed (1) all LAI-NOS nodes measurements agreed very well with Sentinel-2 inversion LAI in the experimental period (average R2=0.94, RMSE=0.41); (2) the possible overestimate of Sentinel-2 inversion LAI was found in the middle stage of wheat (jointing-anthesis); (3) MOD15A2H and LAI-NOS measurements showed similar crop growth trends in long-term observations.