Characterization of spatio-temporal variation in vegetation phenology of regional forest landscapes using remote sensing and GIS

Object-oriented image classification, allows the integration of a broad spectrum of different
object features, such as spectral values, shape and texture. Nine vegetation and land cover
classes were identified in selected regional landscape of Similipal in Odisha. It has about
70% of area under forest cover. Spatio temporal analysis of normalized difference vegetation
index (NDVI) data from satellite measurements can be used to characterize and quantify
changes in vegetation phenology. In this study Sentinel 2 based NDVI measurements for
2018 were used for the four distinct regional landscapes in India (Biligirirangan hills,
Chandoli, Shettihalli and Similipal). The study used the software package CropPhenology
which computes 15 phenological metrics. The derived phonological metrics can be used to
characterize growth conditions, including new leaf emergence and leaf offset. The land
cover maps were used a reference to account the phenological variations across selected
vegetation classes. It was found that all the four study areas reaches its maximum NDVI
value at different time of months mostly between Mid July to October. Further, species level
phenological characterization was carried out for Biligirirangan Hills and Similipal
landscapes based on field plot data. It was found that the Terminalia alata-Pterocarpus
marsupium has the maximum length of growing period of 11 months and the month with
maximum NDVI in October in Biligirirangan whereas in Similipal, Michelia champacaSyzygium
cumini

showed maximum length of growing period of 11 months and reaches its
maximum NDVI value in the month of September. The spatial data generated in the study
would be useful for vegetation-climate modelling and linking with influence of abiotic and
biotic factors on vegetation dynamics.

MPhil EI 2018-2019 INT Dr. R Jaishankar