地震预报的科学哲学与方法论探索——50年回眸与前瞻

地震预报是具高度复杂性的世界科学难题。以我国50余年实践中的问题为导向，以其揭示的自然现象为依托，从科学哲学及方法论视角，回顾经验归纳、还原论动力学、归纳-演绎和复杂动力系统等方法。通过汶川、唐山等震例探讨认识论因素在预测预报中的重要作用。进入地球系统科学新时期，建议以大陆变形复杂动力系统演化及其地震行为、经验与数值预测桥梁的图像（斑图）动力学作为地震预报的自然观和方法论。它们虽不完善，但更贴近大自然的整体性、进化性和非线性本质，又能包容连接多种思路与方法；既具前瞻性，又具可操作性；地震具可预测性，又具预测的不完全确定性；预测→预报→减轻灾害尚存在颇大的可创新空间。

Research into Philosophy of Science and Methodology on Earthquake Prediction: Looking Back over 50 Years and Looking Forward

Earthquake prediction is a worldwide scientific problem with high complexity. Guided by the problems in our country’s practice for more than 50 years, relying on the natural phenomena revealed by it, from the perspective of philosophy of science and methodology, we review the methods of empirical induction, reductionist dynamics, induction-deduction and complex dynamic systems. The important role of epistemological factors in prediction is discussed through the examples of the Wenchuan and Tangshan earthquakes. In the new era of earth system science, it is suggested that the evolution of complex dynamic systems of continental deformation and seismic behavior, and that image (pattern) dynamics of bridges for empirical and numerical prediction be taken as the natural view and methodology of earthquake prediction. Although imperfect, it is closer to nature’s holistic, evolutionary and non-linear nature, and can contain and connect various ideas and methods. It is both forward-looking and operable. Earthquakes are predictable and uncertain. There is still considerable room for innovation in forecasting, prediction and disaster mitigation.