A Landsat MSS time series model and its applications in geological mapping

Abstract

Although temporal factors in remotely sensed data have long been recognized by remote sensing scientists, studies on the separation of temporal and stable factors have been very limited. Applications of the time-variant characteristics of the data are also limited in the environmental monitoring and change detection fields. The merits and applications of the stable factors in the time-variant historical image data or image time series have not been fully recognized. This paper presents a time-variant model for the historical multitemporal multispectral scanner (MSS) data and a technique for processing the image time series for geological mapping. This model has been successfully applied to the surface geological mapping of a region in central eastern Nevada, USA.