Abstract—In agri-business, to keep track of the potential harvest is an important issue to production managers. For this purpose, field observation is necessary for collecting the information of crop growth and climate changes. In Taiwan, the central weather bureau (CWB) is the national authority that provides weather reports of national scale, which is estimated from weather stations established by Taiwan government. The estimated weather of cities or counties in Taiwan can be found on the official website of CWB. To our knowledge, however, there is no official tool for farmers to track and to record the climate data of his/her farmland. In this paper, we utilize the open data of 29 weather stations published by CWB, developing a farmland climate tracking tool for farmers. In this system, we implement four different algorithms for climate estimation, which are the nearest neighbor (NN), the inverse distance weighting (IDW), the Kriging with partial least square regression (KPLS), and the IDW with altitude filter (IDWAF). We compare the performance of these four algorithms by the leave-one-out cross validation on the 29 weather stations with observed climates as open data. The considered types of climate are temperature, humidity, insolation, and wind speed. According to the experimental results, the temperature and  insolation are of lower errors in each algorithm, which indicates that field temperature and insolation could be tracked with open data. In addition, IDWAF achieves better accuracies than IDW. Therefore, altitude filtering is a valuable approach to be combined with inverse distance.  
 
Index Terms—climate tracking, farmland, open data

Cite: Chin-Shun Hsu, Yung-Hsing Peng, and Po-Chuang Huang, "Climate Tracking for Farmland using Open Data from the Central Weather Bureau in Taiwan," Vol. 6, No. 4, pp. 227-232, November, 2015. doi: 10.12720/jait.6.4.227-232