Comparing kriging estimators using weather station data and local greenhouse sensors

Abstract

In order to minimize the impacts of climate change on various crops, farmers must learn to monitor environmental conditions accurately and effectively, especially for plants that are par-ticularly sensitive to the weather. On-site sensors and weather stations are two common methods for collecting data and observing weather conditions. Although sensors are capable of collecting accurate weather information on-site, they can be costly and time-consuming to install and main-tain. An alternative is to use the online weather stations, which are usually government-owned and free to the public; however, their accuracy is questionable because they are frequently located far from the farmers’ greenhouses. Therefore, we compared the accuracy of kriging estimators using the weather station data (collected by the Central Weather Bureau) to local sensors located in the greenhouse. The spatio-temporal kriging method was used to interpolate temperature data. The real value at the central point of the greenhouse was used for comparison. According to our results, the accuracy of the weather station estimator was slightly lower than that of the local sensor esti-mator. Farmers can obtain accurate estimators of environmental data by using on-site sensors; however, if they are unavailable, using a nearby weather station estimator is also acceptable.