Historical records of climate fields are often sparse due to missing measurements, especially before the introduction of large-scale satellite missions. Several statistical and model-based methods have been introduced to fill gaps and reconstruct historical records. Here, we employ a recently introduced deep-learning approach based on Fourier convolutions, trained on numerical climate model output, to reconstruct historical climate fields. Using this approach we are able to realistically reconstruct large and irregular areas of missing data, as well as reconstruct known historical events such as strong El Ni\~no and La Ni\~na with very little given information. Our method outperforms the widely used statistical kriging method as well as other recent machine learning approaches. The model generalizes to higher resolutions than the ones it was trained on and can be used on a variety of climate fields. Moreover, it allows inpainting of masks never seen before during the model training.

Tokyo marked a record high temperature for November on Tuesday, rewriting the previous record set 100 years ago, the Meteorological Agency said. The mercury hit 27.5 degrees Celsius just past noon, topping the November 1923 mark of 27.3 C. This record comes amid an unusually warm autumn and follows an especially hot summer, not only in Tokyo but across Japan. On Monday, the temperature in Tokyo reached 25.1 C, marking a record 142 days of the year topping 25 C. The city of Kumamoto saw a high of 25.8 degrees the same day, the 177th day of summer-like weather this year -- the most since records began in 1890.