Compound dry-to-wet transitions and intensifying wet-day rainfall in monsoon-exposed Southern Nghe An, Vietnam

Abstract

Monsoon rainfall hazards depend on event intensity and sequence, not annual totals alone. We analyzed 8,766 daily observations from four gauges in southern Nghe An during 1998–2021, together with daily 0.05° Climate Hazards Center InfraRed Precipitation with Stations (CHIRPS) rainfall at Nam Dan. Thirteen indices represented amount, wet-day frequency, intensity, extremes, and persistence. A pre-specified core set was evaluated with Sen slopes, modified Mann–Kendall tests, confidence intervals, and within-index false-discovery-rate control. A transition required at least 10 consecutive days below 1 mm followed within seven days by rainfall at or above the station-specific wet-day 95th percentile. Station-level detections (n = 58) formed 41 same-day clusters; 18 detections occurred in 1998–2009 and 40 in 2010–2021. The later-to-earlier rate ratio was 2.22 (bootstrap 95% interval 1.10–4.90), the year-label permutation p-value was 0.056, and a Poisson model with station effects gave a rate ratio of 2.02 per decade (95% confidence interval 1.29–3.16). These estimates indicate recent clustering within this short record, not a detected monotonic station trend. CHIRPS underestimated Nam Dan total rainfall by 13.2% and the simple daily intensity index by 17.2%; annual total wet-day precipitation correlation was 0.81, but only one of 14 gauge transitions matched a CHIRPS event within three days. Thus, CHIRPS supported annual-scale variability more strongly than daily event timing. After within-index adjustment, wet-day intensity increased at Do Luong and Dua, while wet-spell duration increased at Dua, Nam Dan, and Yen Thuong. The proposed transition metric is a meteorological screening tool requiring local impact calibration.