根據過去許多台灣的研究指出，各類波段（包含Ｓ、Ｃ及Ｘ公分波段）的雙偏極化雷達，均已顯示能提升雷達定量降水估計的準確性，同時利用雙偏極化雷達及雨滴譜儀，進行雨滴粒徑分佈的估測，是能提升雷達定量降水估計準確性的主要原因。在二至三年後，台灣將建置由五部Ｃ波段雙偏極化雷達為主的定量降水估計雷達網，此雷達網將搭配氣象局現有的Ｓ波段都卜勒雷達及雙偏極化雷達，針對台灣複雜地形，提供準確的定量降水估計產品，以提供作業及研究單位使用。然而不同波段雷達的硬體設定、周遭環境條件等差異，產生各自不同的非氣象資料干擾，在定量利用雷達資料前，須正確的分辨氣象、非氣象資料。利用自適模糊邏輯法辨識非氣象雷達資料技術，已在106年度科技部自然司防災科技學門專題研究計畫支持下完成開發。此外，因濕天線罩所造成的變動系統偏差及衰減效應，也因不同雷達特性而有所不同，發展自動化的修正技術對於後續應用至關重要。本計劃將依據107 年度科技部自然司防災科技學門專題研究計畫課題重點說明(氣象領域，學門代碼：M1710)之研究課題1-3(臺灣與周邊地區雷 達遙測應用於氣象 防災的技術開發、 驗證與應用)，發展雷達資料處理相關技術。本三年期計畫將分三階段研究如下：一：建立自動化系統性偏差、濕天線罩效應及回波衰減修正技術。二：以中央大學Ｃ波段及五分山Ｓ波段雙偏極化雷達資料，測試第一階段建立的通用雷達品質管制程序,並建立多雷達整合技術。三：測試通用雷達品質管制程序對雷達定量降水產品的影響。
Several previous studies of using various dual-polarimetric radars with different wavelengthes (i.e., S, C and X-band) in Taiwan have shown the pronounced improvement of the accuracy of the quantitative precipitation estimation (QPE). The utilizing of the dual-polarimetric radar and the disdrometer data (i.e., drop size distribution) simultaneously is the main reason. In the coming few years, a new C-band QPE dual-polarimetric radar network will be constructed and cooperated with S-band network. Due to various manufacture configurations and surrounding environments, each radar has distinct error characteristics of non-meteorological signals. The removal of these non-meteorological signals is essential for quantitative applications from dual-polarimetric radar. The self-adaptive fuzzy-logic non-meteorological signals identification algorithm has been developed in previous project. Moreover, the dynamic system bias due to wet-random effect and the rain attenuation effect are crucial for further applications as well. The goals in this three-years project are: (1) developing the unified quality-control (QC) procedures for various radars (e.g., system bias, wet random and attenuation corrections), (2) evaluating the unified quality-control (QC) procedures by applying to NCU CPOL and RCWF SPOL and developing the integration technique of multi-frequency radar data.(3) investigating the performance of radar-based QPE before and after applying unified quality-control (QC) procedures.
