Forecastability of Meso-γ-scale, Localized, Extreme, Heavy Rainfalls during the Warm Season in Japan using High-resolution Precipitation Nowcasts

Development and verification of a forecast method for localized, meso-γ-scale (2–20 km), extreme, heavy rainfall (MγExHR) is important, because it can cause urban flash flooding and inundation with accompanying damage and potential loss of life. Although previous studies have examined the forecastability of precipitation at a very short range (≤ 1 h) using extrapolation-based nowcasts, they did not specifically focus on MγExHR. In this study, we examine the forecastability of 23 selected events of MγExHR (one-hour rainfall accumulation ≥ 50 mm) that occurred during the warm season of 2014 in Japan using High-Resolution Precipitation Nowcasts (HRPNs) provided by the Japan Meteorological Agency, which is one of the most advanced extrapolation-based nowcasts. The forecastability was examined using both traditional grid-scale verification with categorical verification statistics (e.g. equitable threat score) and neighbourhood verification with Fractions Skill Score. Our findings are summarized as follows:

(i) The forecastability limit for heavy rainfall areas of ≥ 20 mm h^-1 was very short (~10 min) at the grid-scale (Δx = 1 km).

(ii) The forecastability limit for heavy rainfall areas of ≥ 20 mm h^-1 can be extended to ~30 min by tolerating displacement errors of ~10 km.

This result suggests that the numerical weather prediction (NWP) model, whose accuracy is useful after ~30 min, is necessary to seamlessly provide useful forecasts in heavy rainfall areas of ≥ 20 mm h^-1 for MγExHR with ~10 km displacement errors, by blending extrapolation-based nowcast with NWP.