New method predicts rainfall more accurately, improving hydrological models.
Researchers used different methods to create rainfall maps from data collected at 19 rain gauges in the Middle Yarra River catchment, Australia. They found that kriging methods, especially ordinary kriging with a spherical variogram model, provided the most accurate estimates for rainfall across all months. A new method called kriging with genetic programming also performed well, giving similar results to ordinary kriging with the spherical variogram model. This study shows that these kriging-based methods are better than the inverse distance weighting method for estimating and mapping rainfall in the area.