CALCULATING VEGETATION INDEX BASED ON THE UNIVERSAL PATTERN DECOMPOSITION METHOD (VIUPD) USING LANDSAT 8 Xiaojun She a,b , Lifu Zhang a , Muhammad Hasan Ali Baig a,b ,Yao Li a,b a Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing, China b University of Chinese Academy of Sciences, Beijing, China ABSTRACT This study introduced the vegetation index based on the universal pattern decomposition method (VIUPD) and then applied on a new sensor—Landsat 8 Operational Land Imager (OLI). VIUPD is a valuable sensor-independent spectral analysis method. Each pixel is described as the linear mixture of standard spectral patterns for water, vegetation, soil and supplementary patterns included when necessary. In the present paper, processing procedure about the data acquisition, radiometric calibration and atmospheric correction have been elaborated. The normalized reflectance (P) of four standard samples resampled to OLI has been listed. For validation of the results, Normalized Difference Vegetation Index (NDVI) and VIUPD have been calculated for comparison. The results showed that VIUPD is more sensitive to the vegetation amount change even in the high vegetation coverage, while the NDVI is more rapidly saturated in high vegetation cover area. In addition, VIUPD is more sensitive to the soil background than NDVI. Index Terms— VIUPD, UPDM, Landsat, Operational Land Imager, sensor independent. 1. INTRODUCTION Identifying the state of vegetation over a wild range and long-term is important in regional and global modeling, ecological monitoring and climate change detection[1]. Remote sensing data obtained using satellite instruments, offer the potential of measuring vegetation status across a wide area of spatial and temporal scales. Vegetation indices are the simplest and most efficiently way to evaluate the information of vegetation from remote sensing data. The traditional vegetation index such as Normalized Difference Vegetation Index (NDVI) was put forward to use the usual satellite red and near-infrared bands. In another word, they relies on the sensors so it will has limits when two or more sensors fusion or compared. VIUPD is a sensor independent vegetation index based on the Universal Pattern Decomposition Method (UPDM) which can be applied on both multispectral data and hyperspectral data. Pattern decomposition coefficients for each pixel contain almost all the sensor-derived information. More importantly, the VIUPD is independent of sensors[2]. Many researchers demonstrated the applications of VIUPD by using different sensors, such as TM, ETM+, MODIS and hyperspectral sensors like Hyperion and CHRIS[2, 3]. Landsat program has been dedicated to sustaining data continuity for nearly four decades[4, 5] and accumulate rich high spatial resolution and multi-spectrum remote sensing images[6]. The continuous Landsat data source provides long-term observation for global change, especially on land cover change detection and vegetation phenology [6-9]. Lately the launching of Landsat 8 injected new blood into this big and ancient family to make it continually serve for researchers to earth observation. Since the Landsat 8 is a new sensor, the independent vegetation index—VIUPD has not been used on the OLI onboard the Landsat 8. The objective of this paper is to demonstrate how to get the VIUPD parameters for Landsat 8 data. 2. METHOD VIUPD is based on the universal pattern decomposition method (UPDM)[10]. UPDM has improved the pattern decomposition method (PDM) to make the parameters independent on sensors. The principle of UPDM is based on that each pixel measured by the sensor can be decomposed into three standard spectral elements which are the water, vegetation and soil as the following equation (1) described. 4 4 () () () () () w w v v s s Ri C Pi C Pi C Pi C Pi         (1) Where R(i) is the reflectance of band i for each pixel. C w , C v , C s are the decomposition coefficients, in other words, they stand for the abundance for each element. P w , P v , P s are the normalized reflectance of standard samples which are the water, vegetation, soil and supplementary yellow leaf. Three components can cover about 95.5% information of spectral reflectance. Adding on the fourth supplement component will reduce the error. Each P can be got as follows: 4734 978-1-4799-5775-0/14/$31.00 ©2014 IEEE IGARSS 2014