Machine learning produces more accurate rainfall predictions
Researchers from the Institute of Industrial Science, The University of Tokyo, develop a novel method for more accurate prediction of local rainfall patterns
Researchers from the Institute of Industrial Science, The University of Tokyo, developed a method for producing more accurate estimates of local precipitation. A bias correction method using machine learning, which recognizes the complex relationships influencing local precipitation, was used to estimate hourly precipitation frequencies. The method could accurately predict spatial precipitation systems at scales of 2500-40000 km2, which determined local hourly precipitation frequencies. The improved estimates can be used for more precise water management.