Refinement of Microwave Vegetation Index Using Fourier Analysis for Monitoring Vegetation Dynamics

Knowledge of the vegetation phenological dynamics is crucial to the understanding of the Earth ecosystems and carbon cycles. A dual-frequency dual-polarization microwave vegetation index (MVI) has been developed recently for the Advanced Microwave Scanning Radiometer for the Earth Observing System; however, the noisy behavior of MVI limits its applications in the studies on terrestrial vegetation. In this letter, a method based on Fourier analysis is proposed to refine MVI. By excluding the high-frequency variations, the method helps to recover the trend of vegetation seasonal changes inherently contained in the raw MVI data series. Comparisons between the refined MVI and the normalized difference vegetation index (NDVI) data sets from years 2002 to 2005 show that the correlation between the refined MVI and NDVI is significantly increased for the three study sites. The refined MVI along with the optical vegetation indices provides complementary information on vegetation and can be served as a useful tool to monitor regional and global vegetation dynamics.