Agriculture Information Service Built on Interoperable Sensor Network

Abstract:

We are developing an agricultural information service platform called FieldTouch, that is built on an interoperable geospatial data infrastructure. More than 300 farmers in Hokkaido, Japan, are participating on this development and they are utilizing our services for optimizing their daily agricultural practices. FieldTouch integrates multi-scale sensor data for field monitoring, provides functionality for recording agricultural practices, then supports farmers in decision making. The platform has two layers. The base layer is interoperable geospatial data infrastructure for sensor networks, that provides updated sensor data to upper application layer via OGC( Open Geospatial Consortium) standard API. We are updating RapidEye satellite images every two weeks, and field sensor data from 50 nodes. Agro-Weather data generated by NIAES using national weather observation network; AMeDAS is also a data source at daily basis. We use "cloudSense" sensor backend service as the infrastructure that serves meta-data and data to the application layer via standard web service; SOS (Sensor Observation Service). SOS provides unified procedure to applications when they access different type of sensor platform, which brought us a great flexibility and automation in operating the system. We developed a crop simulation application in FieldTouch so that farmers are able to see the impact of different planting data on yield. Cultivar data of local wheat variety for DSSAT crop model was calibrated by data assimilation. We generate 100 of weather scenario by a weather generator to present the obtain distribution of yield in different planting options. We update the weather scenario day by day in order to reduce the uncertain of the prediction towards harvesting. Visualizing expected yield as distributions help farmers to understand the risks by weather variability as well as to optimize their works and investment.