Conference on Intelligent Data Understanding 2010 (CIDU)

DISTRIBUTED ANOMALY DETECTION USING SATELLITE DATA FROM
MULTIPLE MODALITIES

KANISHKA BHADURI, KAMALIKA DAS, AND PETR VOTAVA

Abstract. There has been a tremendous increase in the volume of Earth Science data over the
last decade from modern satellites, in-situ sensors and different climate models. All these datasets
need to be co-analyzed for finding interesting patterns or for searching for extremes or outliers.
Information extraction from such rich data sources using advanced data mining methodologies is
a challenging task not only due to the massive volume of data, but also because these datasets
ate physically stored at different geographical locations. Moving these petabytes of data over the
network to a single location may waste a lot of bandwidth, and can take days to finish. To solve this
problem, in this paper, we present a novel algorithm which can identify outliers in the global data
without moving all the data to one location. The algorithm is highly accurate (close to 99%) and
requires centralizing less than 5% of the entire dataset. We demonstrate the performance of the
algorithm using data obtained from the NASA MODerate-resolution Imaging Spectroradiometer
(MODIS) satellite images.