Climatic trends in most parts of the world show a significant increase in rainfall amount, intensity and its frequency. Similarly, these trends are likely to continue in future as well. The major catastrophe caused by these rainfall trends comes as flooding, which is getting harder to predict, and mainly over the mountainous regions. Modelling these extreme rainfall events is crucial, which needs better forecasting skills and more understanding of existing Numerical Weather Prediction (NWP) modelling setup. Although, recent developments in NWP models increase the capabilities to simulate rainfall more precisely but the strengths and weaknesses of model need to be evaluated based on climatic conditions, terrain characteristics -- which include landuse, topography, new physical schemes and static datasets. Therefore, we have conducted comprehensive sensitivity and qualitative analysis with a numerous model setup, physical schemes and terrain datasets. / We contemplate various physical parameterizations and updated terrain datasets to simulate the rainfall over the complex topography using WRFV3-ARW modelling system. Additionally, the impact of topography and landuse on rainfall are discussed in detail along with a several combinations of newly available land surface, planetary boundary layer (PBL), cumulus, and cloud microphysics (MP) schemes. As a case study, we select the north region of Pakistan, which includes Khyber-Pakhtunkhwa (KPK) province and part of Hindukush-Karakoram-Himalaya (HKH), and this region have diversified the landscape and complex topographic features. For the sake of better understanding and comparative discussions, we study three extreme rainfall events; two of them occurred during monsoon period (i.e., July), while one in post monsoon period (i.e., September). / WRF-ARW 3.5.1 model is tuned and tested with GFS0.5 and CFSR/CFSv2 as forcing and lateral boundary conditions with a number of parameterization schemes. Similarly, to minimize the errors induced by terrain features, we apply wind correction and drag parameterizations. Furthermore, 3-arc-second hydrologically corrected SRTM digital elevation model (DEM), modified MODIS IGBP 30-arc-second, MODIS 15-arc-second and GLCNMO2008 landuse datasets were also integrated to WRF along with default datasets in WRF modelling system. / We verify the simulated rainfall by using observed, the Tropical Rainfall Measuring Mission (TRMM) and the Climate Prediction Centre morphing method (CMORPH) rainfall datasets. The GIS-based verification technique, called fisher-net is also introduced which is more compatible and flexible with other tools as well. / 隨著全球氣候變化的加劇，強暴雨極端天氣事件呈現突發、多發、併發的特點，其頻次、強度、持續時間、籠罩範圍近年來均呈現急速上升的趨勢。如何構建高時效性、高精度和高可用的極端天氣事件模擬工具，已經成為災害應急管理與回應等領域迫切需要解決的關鍵科學問題。作為極端天氣事件類比的核心，數值氣象模型對極端暴雨事件的模擬能力日益完善；然而，到目前為止，氣象模型的可靠性和有效性評價仍是其推廣應用的關鍵，特別是如何顧及土地利用類型、地形、新型物理機制和多源靜態資料庫前提下的模型可靠性評價仍面臨很多挑戰。因此，本文提出了顧及不同物理機制、地形特徵的模型構建、模型敏感性評價和定量分析方法。 / 首先，本文在顧及不同的物理參數和地形特徵的基礎上，利用WRFV3-ARW 建模系統實現了對複雜地形特徵下降雨過程的類比與分析。在此基礎上，本文充分考慮和利用新的土地下墊面、行星邊界層、積雲以及雲微物理機制，以詳細分析了地形和土地利用類型對降雨影響。實驗選用具有複雜地形結構和特徵的巴基斯坦北部的Khyber-Pakhtunkhwa (KPK)省和部分Hindukush-Karakoram-Himalaya (HKH) 區域；為了獲取更充分的分析結果，本位對該實驗區域內三次極端降雨事件進行了模擬和分析，包括季風期（例如七月）的兩次降雨事件和季風期過後（例如九月）的一次降雨事件。 / 其次，本文利用GFS0.5 和 CFSR/CFSv2 作為強迫和側邊界條件，設置多參數方案對WRF-ARW 3.5.1 模型進行了優化和測試。與此同時，為了降低由地形特徵導致的類比誤差，本文引入了風向糾正參數和風阻參數。除此之外，本文充分利用了水文糾正過後的3 弧秒精度的SRTM DEM 資料、30 弧秒精度的MODIS IGBP 資料、15 弧秒精度的MODIS 資料、GLCNMO2008 格式的土地利用資料、以及WRF 建模系統的預設資料，支撐WRF 的建模過程。 / 最後，為了驗證本文實驗結果的可靠性，本文利用TRMM 獲取的實測降雨量資料以及TRMM 提供的降雨資料庫驗證，基於GIS 的漁網驗證法，對上述模擬結果進行了詳細的分析 / Sultan, Shahzad. / Thesis Ph.D. Chinese University of Hong Kong 2015. / Includes bibliographical references (leaves 111-125). / Abstracts also in Chinese. / Title from PDF title page (viewed on 09, September, 2016). / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only.