本計畫屬於一整合型計畫之子計畫。計畫內容包含基礎科學研究，以及為達成此科學研究目標所需的雷達升級建置計畫。今分述如下: (1) 都卜勒雷達風場與熱力場反演問題的發展在雷達氣象界已有數十年的歷史。1990 年代由於變分技術3D-VAR、4D-VAR 方法的引入，又有了蓬勃的進展。本計畫預計在未來對此問題做如下探討: (a) 將申請人已自行開發之移動坐標反演法由水平改為垂直排列，或推廣至完整三維空間，藉此研究其反演三維風場尤其是垂直流場結構之效果。發展可以保守徑向風的移動坐標。 (b) 發展適用於地形上之熱動力反演方法。 (c) 發展4D-VAR 技術，以便由都卜勒徑向風場與回波場反演三維笛卡爾風場、熱力場、降水參數。其中降水參數的反演正確與否，必須由雙偏極化雷達的觀測資料予以定性與定量的驗證或調整。 (d) 研究同化雙偏極化都卜勒雷達資料對模式預報的影響。 (2) 目前國際氣象雷達科技的最新發展趨勢已經由傳統雷達，都卜勒雷達進展到雙偏極化雷達。而許多研究也證實此類雷達可探測雨雲中水粒子大小、形狀、相位等雲物理特性，因此比都卜勒雷達可得到更多對雲物理之瞭解，包括冰晶、雪花、冰雹、小雨滴或大雨滴在雲內三維空間之分佈，降水粒子之粒徑分佈、定量降雨的估計等。中央大學大氣科學系雷達氣象實驗室研究團隊，已於2002 年成功地將中央氣象局高雄站的傳統雷達，搬遷至校園內，並將其升級成都卜勒雷達。有鑑於國內目前尚無雙偏極化雷達，因此透過與國內外包括中山科學研究院、中央氣象局、美國 NCAR( Dr. Jeff Keeler)、NSSL (Drs. Doviak and Zrnic)的專家多次商討後，認為可利用合理的費用，逐年將中大都卜勒氣象雷達再升級為雙偏極化雷達。如能順利完成後，不僅能提供都卜勒雷達之回波強度、徑向風場外，可另外提供 Zdr 、 Kdp 等雙偏極化雷達參數，所得之雲物理特性可與模式結果相互驗證，並做為資料反演或同化之用，也可提供給做衛星反演算法之參考(如: GPM)。未來吾人將可以使用十分先進的研究用偏極化雷達長期觀測台灣北部的降水系統，做出較深入之研究，同時在改裝升級過程中，將結合 NCAR 、中央大學大氣系與電機系、中山科學院、中央氣象局之科學家及工程師共同參與，提升國內氣象雷達整體技術的紮根與進步。 This proposal is part of an integrated project. It contains a scientific goal for basic research, and a hardware construction plan for the fulfillment of this goal. The following is a general description of this proposal. (1) The problems of Doppler wind retrieval and thermodynamic retrieval have been drawing the radar meteorology community』s attention for decades. Due to the introduction of variational analysis technology in the 90』s, the development in solving these problems regains new momentum. In this project, we plan to study the following issues: (a) Expending the moving frame of reference algorithm, a single Doppler wind retrieval method developed by the applicant during the past several years, to three dimensional, or re-aligning the framework from horizontally to vertically oriented. By doing so, one can examine the performance of this method in recovering the three-dimensional wind, especially the vertical flow structure. We will also develop the optimal reference frame on which the Doppler radial winds can be considered a conserved quantity. (b) Developing a thermodynamic retrieval scheme which can be used over a domain with complex terrain. (c) Developing the 4D-VAR method so that one can use the Doppler radial wind observations to infer the three-dimensional Cartesian winds, the thermodynamic field, as well as the microphysics parameters. The accuracy of the retrieved microphysics parameters needs detailed, both qualitative and quantitative, verifications so that the 4D-VAR adjoint scheme can be modified accordingly. This can be accomplished by using the dual-polarization Doppler radar measurements. (d) Study the impact on the model forecast from assimilating dual-polarization Doppler radar measurements. (2) The latest meteorological radar technology in the world has migrated from traditional and Doppler radar to dual-polarization radar. It has been proven that a dual-polarization radar is capable of detecting the three-dimensional distribution of the shape, size and phase of hydrometers. In other words, the drop size distribution function can be better evaluated. As a result, the so-called quantitative precipitation estimation (QPE) can be obtained with a higher accuracy. 研究期間：9308 ~ 9407
