Generating time series of medium-resolution albedo images by Kalman filtering algorithm

Abstract

Albedo is fundamental to monitoring the energy and mass exchanges between the land surface and the atmosphere from local to regional scales. There is an growing need for high spatio-temporal resolution surface albedo to satisfy the applications at medium-to-fine scales. In this paper, the Kalman filter (KF) algorithm was used to simulate temporal sequence of 30m synthetic albedo images combining Landsat ETM+ imagery and MODIS products. This method considered the change rules of long time series albedo data and the uncertainty information. The predicted time series of medium-resolution albedo not only captured the seasonal variations of MODIS albedo, but also revealed the spatial features in the validation albedo derived from Landsat ETM+ images. It showed the feasibleness of applying different images to produce shortwave albedo data of high spatial temporal resolution.