In the mid-1890s, some leading Japanese meteorologists were involved in a short controversy on Bai-u (or Tsuyu in Japanese, rainy season), one of the annual meteorological phenomena in Japan. The controversy was triggered by two meteorologists, one emphasized the low atmospheric pressure as the main factor of bai-u, and the other focused instead on the pacific monsoon. Other participants in the controversy also emphasized the influence of either low pressure or monsoon as the main factor of rainy season. But all of them failed to provide convincing explanations for their respective positions, and the controversy quieted down in 1897 without any consensus. It was mainly because their analytical perspectives were limited to the Japanese islands, and because they could not make use of weather data from neighboring countries and regions of Japan. After the controversy, too, Japanese meteorologists continued to investigate the Bai-u mechanism, but their geographical perspective became broader as the frontier of the empire was expanding during 1890s and 1900s. Okada Takematsu's achievement in 1910 contained the result of all data and researches during the period when Japan experienced two big wars and acquired new colonies. We can see in his paper the broadened analytical perspective by incorporating meteorological data from the expanding frontiers of the Japanese empire, including Taiwan, Sakhalin (Karafuto), Korea, and Manchuria. This paper examines how the Bai-u controversy at the turn of the twentieth century provided the Japanese meteorologists with a unique opportunity to analyze meteorological phenomena in a larger perspective beyond the Japanese islands, and how they in the process tried to define the characteristics of Japanese climate. The expansion of imperial observation network enabled the Japanese meteorologists to realize that the Bai-u was not a local meteorological phenomenon but that operated in a much larger geographical scale. As a result, when one local weather phenomenon came to be regarded as a global phenomenon, the broadened meteorological view made the Japanese islands localized in the empire and thereafter it was applied to analysis of the Japanese climate.