Study on lognormal DSD models for the rain rate estimation applicable for GPM-DPR observation

The gamma distribution is used for the DSD model most frequently. Two parameter gamma model using DFR (Dual Frequency Ratio) is proposed for the DSD estimation algorithm of GPM-DPR (Seto et al. 2013). However, natural DSDs are highly variable. Lognormal models are known to fit the natural DSD as well as the gamma models. In this study a new method is proposed to obtain two parameters of lognormal DSD model using PCA (Principal Component Analysis) which will be applied for GPM-DPR observation. Some merits and demerits of this method will be discussed.