Spatiotemporal Variability of Current and Future Sub‐Daily Rainfall in Japan Using State‐Of‐The‐Art High‐Quality Data Sets

Abstract

AbstractUsing potentially best available rainfall data sets for the entire country of Japan (spatial scales of 1‐ and 20‐km), we analyze the 1–24 hr and city‐scale (1–400 km2) extreme rainfalls for both current (2006–2020) and future periods (2081–2109) at 1.5 K global warming scenario, complementing previous work that focuses on either coarse spatial and temporal scales or other warming scenarios (e.g., RCP and 2 K warming scenarios). A peak‐over‐threshold (POT)‐based approach is applied to compute areal reduction factor (ARF) for subsequently establishing intensity‐duration‐area‐frequency (IDAF) curves. Our results reveal that ARF values generally decrease with increasing area size and increase with longer duration and are affected by multiple underlying physical phenomena. Moreover, we find a greater increase in the rainfall intensities for shorter durations and higher return periods, ranging from 9.4% (1‐hr) to 6.2% (24‐hr), averaged for all return periods and 8.3% (25‐year) to 7.3% (2‐year), averaged for all rainfall durations. Spatially, extreme rainfall intensities are projected to increase by 8.9% in northern Japan (albeit with a less intense rainfall intensity in the current period), which is greater than the rest of the country (6.8%), underscoring the need to focus more on infrastructure designs in northern Japan. The projected IDAF curves further display an increase in the frequency of extreme events at city‐scale, for example, 25‐year extreme rainfall events in 2006–2020 would likely be 5‐10‐year events in 2081–2109. Our results with the state‐of‐the‐art data and implementable approach can be utilized for policymaking to reduce the warming‐induced risks in Japan and beyond.