為了應對對精確且實時的環境監測和預測的日益需求，本文介紹了一個基於NVIDIA Omniverse構建的系統，用於創建地球的數位孿生。這個數位孿生系統旨在提供一個全面且動態的地球大氣狀況表示，以便更好地理解和管理環境現象。我們的方法整合了衛星雲圖、降雨圖和雷達數據作為輸入，並使用數值方法來計算雲的物理性質和預測未來的降雨區域。結果展示了在Omniverse中呈現的高保真度可視化效果，突顯了系統在精確建模複雜大氣過程方面的能力。此外，我們利用Web Socket技術將這些結果顯示在網頁平台上，使智能手機和筆記本電腦等低端設備也能夠訪問，確保了更廣泛的可及性和可用性。這個系統不僅增強了我們監測和預測天氣模式的能力，還為災害準備和應對、城市規劃和氣候研究提供了寶貴的工具。;In response to the increasing need for accurate and real-time environmental monitoring and prediction, this paper presents a system built on NVIDIA Omniverse to create a digital twin of the Earth. The digital twin aims to provide a comprehensive and dynamic representation of the Earth′s atmospheric conditions, facilitating better understanding and management of environmental phenomena. Our approach integrates satellite cloud imagery, rainfall maps, and radar data as inputs. We employ numerical methods to compute the physical properties of clouds and predict future rainfall regions. The results showcase the high-fidelity visualizations rendered within Omniverse, highlighting the system′s capability to model complex atmospheric processes accurately. Additionally, we utilize web sockets to display these results on web platforms, enabling access from low-end devices, such as smartphones and laptops, ensuring broader accessibility and usability. This system not only enhances our ability to monitor and predict weather patterns but also provides a valuable tool for disaster preparedness and response, urban planning, and climate research.
