AUTOMATIC CROP PHENOLOGY DERIVATION USING NDVI TIME-SERIES AND ITS DISSEMINATION USING WEBGIS

Hits: 82
Research areas:
Year:
2017
Type of Publication:
Article
Keywords:
NDVI, OCM, Proba-V, Crop, WebGIS
Authors:
Shweta Mishra, Shashikant A. Sharma Markand P. Oza
Abstract:
Information of crop phenology is essential for crop management.Remote Sensing has been found as one of the consistent and reliable ways for crop phenology estimation. Time series data of various vegetation indices derived from medium and high resolution satellite data are widely used for vegetation monitoring on a global and regional level. Normalized Difference Vegetation Index (NDVI) is one of the popular indices to study the vegetation phenology. NDVI time series is associated with vegetation growth cycles. Using temporal analysis of NDVI series, it is possible to estimate cropping intensity, which is the numbers of crops (single, double and triple) per year in a unit area. Although NDVI data sets are pre-processed to minimise noises due to orbital and sensor degradation, still some noise remains in the data sets, primarily due to varying cloud and atmospheric conditions. This requires filtering of the data. For smoothening NDVI series two filtering methods namely, Gaussian filtering and Fourier Transform filtering algorithm based on harmonic analysis, were applied and compared. Using spatio-temporal analysis of smoothed NDVI series, variations in annual vegetation phenology were estimated. Based on crop cycles at particular pixel location, number of crops is estimated at that location.OCM and Proba-V NDVI data series of one agricultural year over Gujarat is used for carrying out this study.It was found that Fourier transform based algorithm works better than the Gaussian filter for minimising noise in OCM and Proba-V NDVI datasets. Automated module is developed in python for NDVI smoothening and crop cycle estimation. Web-GIS is a distributed information system which holds the potential to make geographic information available worldwide in cost effective and an easy way. Thus, Final output showing crop intensity in particular period over study area are derived and presented using Web-GIS technology.
Full text: 157.pdf [Bibtex]
You are here: Home ACRS ACRS Overview Proceedings automatic-crop-phenology-derivation-using-ndvi-time-series-and-its-dissemination-using-webgis