عنوان مقاله [English]
نویسندگان [English]چکیده [English]
The present research was carried out with the aim of site-specific nitrogen fertilizer management, through ASTER imagery. A 23 ha corn field in Pakdasht town in southern Tehran Province constituted the research area. Plant sampling was carried out simultaneous with the passing by of the satellite sensor over, throughout the farm. A total of 53 pixels were selected through systematic Randomized sampling method. Nitrogen content was determined through Kjeldahl method. Geometric correction was performed through RMS 0.2 pixels. To predict corn canopy nitrogen content, NDVI, MSAVI2, MCARI2 and MTVI2 indices were investigated. Results revealed that MTVI2 presented the highest correlation with a coefficient of R2=0.87. SAM approach as a supervised classification technique was performed to set apart different nitrogen levels. The overall accuracy was observed as 97.53% with a Kappa coefficient of 0.9669. Results of classification showed that, there were three nitrogen levels distinct on the farm namely: high nitrogen level (2.5-3% nitrogen content), medium (2-2.5% nitrogen content) vs. and low nitrogen level (1-2% nitrogen content). This indicates that 15.8%, 25.2% and 59% of the farm area had received high, medium vs. low nitrogen levels respectively. Based on high nitrogen variations observed in the experimental area, precision management of nitrogen fertilizer application is deemed economical and necessary.