Analog methods (AMs) have long been used for precipitation prediction
and climate studies. However, they rely on manual selections of
parameters, such as the predictor variables and analogy criterion.
Previous work showed the potential of genetic algorithms (GAs) to
optimize most parameters of AMs. This research goes one step further and
investigates the potential of GAs for automating the selection of the
input variables and the analogy criteria (distance metric between two
data fields) in AMs. Our study focuses on daily precipitation prediction
in central Europe, specifically Switzerland, as a representative case.
Comparative analysis against established reference methods demonstrates
the superiority of the GA-optimized AM in terms of predictive accuracy.
The selected input variables exhibit strong associations with key
meteorological processes that influence precipitation generation.
Further, we identify a new analogy criterion inspired by the
Teweles-Wobus criterion, but applied directly to grid values, which
consistently performs better than other Euclidean distances. It shows
potential for further exploration regarding its unique characteristics.
In contrast to conventional stepwise selection approaches, the
GA-optimized AM displays a preference for a flatter structure,
characterized by a single level of analogy and an increased number of
variables. Although the GA optimization process is computationally
intensive, we highlight the use of GPU-based computations to
significantly reduce computation time. Overall, our study demonstrates
the successful application of GAs in automating input variable selection
for AMs, with potential implications for application in diverse
locations and data exploration for predicting alternative predictands.