MULTIPLICATIVE RANDOM CASCADE HSA METHOD FOR HIGH RESOLUTION RAINFALL FIELD MODELING

抄録

In order to downscale spatial rainfall field from a coarse scale into a finer one, the non-homogenous multiplicative random cascade method is often employed. Currently, this kind of downscaling method is less reliable even though it correctly preserves the statistical description and a long term average spatial pattern. It fails reproducing the spatial patterns in repeated trials; and there is a higher chance of having random outputs. These drawbacks are needed to overcome for applying the downscaled rainfall products in further analysis of hydrological and meteorological studies. In this study, a new method, named as Multiplicative Random Cascade Hierarchical and Statistical Adjustment (HSA) method, is introduced and tested to downscale 1.25 degree GAME Re-analysis data into 10-minute spatial resolution. The obtained results are highly improved, quite robust and reliable than the previous method.