A study was conducted to determine whether generic pattern recognition
techniques could be applied to SSM/I
data to detect rain over land. Three pattern recognition techniques
( Minimum Distance Classifier, Backpropagation Network and Discrete
Bayes Classifier) were compared against a science algorithm (WetNet
PIP SSM/I Algorithm) for this study.
Composite Rainfall product from the Global
Hydrology Resource Center (GHRC) was used to obtain an independent
estimate of rainfall over the area of interest (ground truth).
This was used as the ground truth. SSM/I and US rain data over
the southeastern United States for the period January and July of
1995 were compared in this study. The recognition accuracy for the
different algorithms for this study are given in the figure below.
Collaboration with Domain Expert:
Dr. Steve Goodman (NASA/MSFC GHCC)