氣候科學：電腦模擬的不確定性

Climate Science: Uncertainties in Computer Simulation

全球暖化既是科學議題也是政治議題。由於全球氣候系統的複雜性，科學家在評估自然因素與人為因素對於全球暖化的影響程度時，依賴的不是傳統意義下的實驗，而是電腦模擬。而由於電腦模擬的特性，模擬結果是否可信的問題，不只是科學家關心的合理科學議題，也提供了以科學之名的政治操作空間。本文評述圍繞在當前氣候科學所依賴的模擬之爭議。本文將從說明模擬與模型、理論、實驗的關係開始，接著回顧氣候模擬的歷史，並以氣候模擬中兩個爭議性的實作為例，指出模擬實作所蘊涵的不確定性。然而，除了不確定性之外，許多的證據也相當程度地印證了氣候模擬的結果。

Global warming is a scientific as well as a political issue. Because of the complexity of global climate systems, scientists rely not upon experiments in the traditional sense but upon computer simulation in the evaluation of the relative impact of natural and anthropogenic factors in climate change. The uncertainties involved in computer simulation make its findings (and their potentially grave implications) vulnerable to the question of political manipulation in the name of science. This article reviews the technical controversies over computer simulation in climate science. It begins by introducing the complex relationships between simulation, model, theory, and experiment, followed by a short review of the history of climate simulation. This article explores the issue of uncertainty by focusing on two controversial practices in climate simulation. Despite the uncertainties, climate simulation is confirmed to a considerable degree by the evidence.