The researchers developed a prediction model that was trained to recognize weather conditions that were likely to cause lightning.

研究人员开发了一种预测模型，并训练这个模型识别可能引发闪电的天气条件。

With GRACE data, the researchers found that they could have predicted the Missouri River floods months before current prediction models.

通过对比，科学家发现，与现有的洪水预警仪器相比，GRACE卫星可以提前数月据预测出洪水的发生。

Today I want to introduce to you three prediction models that have been developed.

今天我想向大家介绍三个已经发展起来的预测模型。

At that point human researchers intervene to test the model's predictions in a real biological system.

这个时候，人类研究人员可以进行干预，在真实的生物系统中检验模型的预测结果。

The first prediction model looks along earthquake fault, those cracks in the Earth's crust, to find what are known as seismic gaps.

第一个预测模型沿着地震断层（地壳中的裂缝）寻找所谓的地震间隙。