Land use/Land cover changes in Hugli Estuary using Fuzzy C-Mean algorithm

In the present study, land use/land cover changes has been evaluated in the decade between 1989 and 2010 utilizing Landsat TM5 satellite images in the Hugli estuary which stretches across 4817.98 km 2 of Gangetic delta in West Bengal, India, predominantly dominated by mangrove plantation. The study utilizes supervised classification techniques using Fuzzy Cmean classification algorithm. The main advantage of this approach of classification is that no spectral information is lost like in the case of hard partitioning of feature space, thereby, generating a much accurate classified image. With an overall accuracy of 90% for 1989 image and 89% for 2010 image, and kappa coefficient of 0.87 and 0.85 for both images for 120 sample pixels; it can be said that this technique is very satisfactorily classifying the pixels into different features without losing any information. The total land area has been increased from 2007.63 km2 (41.67%) in 1989 to 2159.39 km2 (44.82%) in 2010, at an average rate of 7.23 km2/year. This is mainly due to the increased land area in the category of waste land and forest including mangrove plantation. This increase in land area is attributed to the huge deposition of silt through the silt-laden river networks in the region.