An improved methodology for paddy field measurement using LANDSAT ETM+

Farmland measurements and crop yields are two main scientific aspects of precision agriculture. Farmland development at Heilongjiang Province, China has been rapidly expanding since 1950's. Meanwhile, there has been a trend that dry farmlands are converted to paddy field since 1990's. Therefore, there is a demand for accurate updated information on the newly developed and expanded paddy field area. The objective of this research was to develop an operational approach and improve the existing methods for assessing paddy field area that would facilitate remote sensing based algorithms for measuring paddy field areas at county-level. LANDSAT ETM+ offers better spatial resolution over other sensors such as Moderate Resolution Imaging Spectroradiometer (MODIS). Four granules of LANDSAT ETM+ dataset were acquired for the study, and this research describes the methodologies used for processing the data and their functions in developing the classification of paddy field area. The software tools used for this research are ENVI and ArcGIS. The results suggest the methods that can be used for operational application of LANDSAT ETM+ for classification at county-level, and the measurement accuracy was approximately 87% of the validated ground truth data.