3 rd ISDE DIGITAL EARTH SUMMIT 12-14 June, 2010, Nessebar, Bulgaria EVALUATING WITHIN-FIELD RICE GROWTH VARIABILITY USING QUICKBIRD AND IKONOS IMAGES IN NORTHEAST CHINA M.L. Gnyp a,b , Y. Yao b , Y. Miao b , K. Yu a , S. Huang b , E. Dornauf a , C. Hütt a , V.I.S. Lenz-Wiedemann a,b , R. Laudien a,b , R. Jiang b , X. Chen b , and G. Bareth a,b a Institute of Geography, University of Cologne, 50923 Cologne, Germany b International Center for Agro-Informatics and Sustainable Development (ICASD), College of Resources and Environmental Sciences, China Agricultural University, 100193 Beijing, China Measuring within-field variability is essential for precision farming, and remote sensing is an important tool to obtain the needed information. The objective of this study was to evaluate rice (Oryza sativa L., irrigated lowland rice) growth variability in Qixing farm considering different fields, management practice and cultivars using high resolution optical satellite image data. Based on the spectral dataset of 2007 and 2008, vegetation indices were calculated to quantify the rice growth variability. The vegetation indices were statistically evaluated and were joined with the spatial field data. The results of the spatial analyses show that growth variability was influenced by field conditions (Coefficient of Variation from 1.3 % to 19.8 %) and the farmer's management, mainly dominated by cultivar differences. Keywords: within-field variability, multispectral, remote sensing, vegetation indices, rice growth, Northeast China Martin Leon, Gnyp, Institute of Geography (GIS & RS group), University of Cologne Albertus-Magnus-Platz, 50923 Cologne, Germany Phone/Fax: ++49-(0)221-470-6551/-1638, mgnyp1@uni-koeln.de Ass. Prof., Yuxin, Miao, PhD Department of Plant Nutrition College of Resources and Environmental Sciences, China Agricultural University, 100193 Beijing, China Phone/Fax: ++86-10-62732865, ymiao@cau.edu.cn 1. INTRODUCTION Measuring within-field growth variability is very important for precision farming. Farmers and scientists generally recognize that soil properties and crop growth are not uniform within a field (Yamagishi et al., 2003). In addition to that, fields show a temporal and spatial variability in crop growth. Washmon et al. (2002) e.g. reports a temporal within- field variability for wheat canopy based on a nine years study in Oklahoma. The within-field variability is caused by many parameters such as soil, nutrition, water availability, slope, local microclimate, and farmer's management (amount and type of fertilizer input, choice of cultivars, etc.). These accumulated factors vary from year to year in the same field and influence the growth and finally the yield. Due to the spatial variability, a uniform nitrogen (N) application without consideration of variable crop growth and N nutrition status in a rice field may result in over-fertilization in some locations (Nguyen et al., 2006). If the causes of this spatial variability can be identified, then site-specific management practices can reduce costs e.g. for fertilizer, increase yield and reduce negative environmental impacts, e.g. caused by over-fertilization (Roel & Plant, 2004). Usually, the within-field variability is quantified by agronomic parameters (Cavero et al., 2001; Robertson et al., 2008) such as biomass, plant height, plant and soil nitrogen and vegetation indices (VIs) (Encloda et al., 2004; Zarco-Tejada et al., 2005), calculated from satellite or airborne images or field spectrometer data. The present generation of high spatial resolution data from sensors such as Ikonos and Quickbird offers the most advanced spatial, temporal and radiometric resolution to meet the requirements of precision farming. By applying methods of geostatistics combined with