Comparing the Radiometric Behavior of LISS III Data of IRS 1D and IRS P6 for Landscape Ecological Application

Abstract

In the recent time, the usage of multi-date satellite remote sensing data is a commonest practice to monitor and model the landscape dynamics. But the reliability of information obtained from such time series analysis of satellite imagery depends heavily on the coherence in the spectral behavior of the multi-dates satellite data. The inconsistency in the spectral behavior among the multi-date satellite data may be attributed to differences in satellite orbit and sensor characteristics and even the environmental conditions. Indian Remote Sensing satellite (IRS) 1D and IRS P6 have been planned to have sensors of almost identical specifications to cater continued data service and to facilitate temporal analysis for resource planning and management. In the present paper an attempt has been made to compare the data quality of the Linear Imaging Self scanning System III (LISS III) data from IRS 1D and IRS P6 satellite of identical haracteristics and to ascertain their usability for temporal analysis. The adiometric characteristics of the LISS III images of the closest dates from IRS 1D and IRS P-6 satellites were compared band wise. Spectral classes obtained through unsupervised clustering were used to compute several class level and landscape level metrics. Statistical analysis of radiometric and landscape characteristics revealed a significant difference in the investigated dataset and their usability for temporal analysis is subjected to proper calibration of radiometric behavior of the two data sources.

Keywords: Indian remote sensing satellite, landscape analysis, radiometric behavior, spectral cluster