Home > Geospatial Application Papers > Agriculture & Soil > Crop Pattern

Overview | Crop Production | Crop Pattern | Crop Yield | Irrigation | Soil Management | Precision Farming | Relevant Products | Relevant Links




Identification and characterisation of grape orchard using IRS 1D LISS III data

Ashvir Singh, N. S. Mehta and S. Panigrahy
Agro-Ecology and Management Division, Agricultural Resources Group
Space Applications Centre, ISRO, Ahmedabad - 380 015



Application of remote sensing in the field of horticultural crops, particularly in orchard crops is relatively a new endeavour. However, it has a tremendous potential in this field, as unlike field crops. Horticulture development has emerged as one of the major thrust areas in agriculture in view of the domestic requirements as well as export potentials of horticultural produce. There is no systematic survey of orchard crops for area, production, productivity and orchard status in the country, which act as the major impediment in the development of horticulture scenario. Grape is one of the most important fruit crop of India and grown for domestic and export purposes. The area and production of grape in India are about 34010 ha and 75000 tones per annum, respectively. Keeping in view the importance of grape industry in India, the present study was taken to study temporal spectral signature of grape orchards and the feasibility of delineating the grape orchard area. Niphad taluk of Nasik district in Maharastra was selected as study area, because grape is predominant crop of this area.

IRS 1D LISS III data acquired on December 26, 1998, January 2 and February 14, 1999 were used. Ground truth data were collected on orchard locations, condition and phenological stages. Unsupervised ISODATA clustering classification was evaluated using various combinations of three date data. Ground truth data were used to identify the clusters belonging to grape classes. Results showed that, in general the classification accuracy was poor using single date data, as the probability of signature overlap with the field crops was high. Three date data has resulted in higher accuracy. Two date data acquired during December and January also gave higher accuracy among the two date data combinations. It can be mentioned that December and January coincide with the peak vegetation period of the grape orchards. Grape signatures were more prominent in the month of December, because of more vegetative cover. Hence, the December data was more appropriate in comparison to February and January data for delineation and discrimination of grape orchards from other vegetation. The temporal profile of the grape orchards was studied using Normalised Difference Vegetation Index (NDVI). It was found to be highest(>0.4) in December, and lowest in February, which matches well with the crop phenology.