Refining environmental satellite data using a statistical approach

Abstract

The proposed approach in this article applies an efficient and novel statistical technique to accurately describe radiometric data measured by Advanced Very High Resolution Radiometers (AVHRR) onboard the National Oceanic and Atmospheric Administration’s (NOAA) Polar Orbiting Environmental Satellites (POES). The corrected data set will then be applied to improve the strength of NOAA Global Vegetation Index (GVI) data set for the 1982- 2003 period produced from AVHRR. The GVI is used extensively for studying and monitoring land surface, atmosphere and recently for analyzing climate and environmental changes. The POES AVHRR data, though useful, cannot be directly used in climate change studies because of the orbital drift in the NOAA satellites over the lifetime of the satellites. This orbital drift causes inaccuracies in AVHRR data sets for some satellites. The main goal is achieved by implementing a statistical technique that uses an Empirical Distribution Function (EDF) to produce error free long-term time-series for GVI data sets. This technique permits the representation of any global ecosystem from desert to tropical forest and to correct deviations in satellite data that are due to orbital drifts and AVHRR sensor degradations. The primary focus of this research is to generate error free satellite data by applying the EDF technique for climatological research.