We can apply anomaly detection algorithms from the machine learning,
data mining, and statistics communities to scientific data, such as
remote sensing data used in Earth science or telescope data used in
space science. These algorithms find parts of the data ("anomalies")
that are different from the rest of the data. The scientists can then
direct their attention to these parts of the data, which can potentially
lead to scientific discoveries. It can also lead to the discovery of
errors in the data, which can then be corrected. We have applied anomaly
detection to vegetation data, in collaboration with Earth scientists.
Specifically, we used the Leaf Area Index (LAI) and Fraction Absorbed of
Photosynthetically Available Radiation (FPAR) products from the Advanced
Very High Resolution Radiometer (AVHRR) and Moderate Resolution Imaging
Spectroradiometer (MODIS) satellite-borne instruments. We discovered an
error in the data, which we determined was caused by an error in the
software that was used to produce the LAI and FPAR products. The
software error was corrected, and new versions of the data were produced
and made available to the scientific community.

TRL: 4