Flash floods are water-related disasters that cause damage to properties, buildings, and infrastructures in the flow path. Flash floods often occur within a short period of time following intense rainfall in the high, mountainous area of northern Thailand. Therefore, the purpose of this study is to generate a flash flood susceptibility map using watershed morphometric parameters and hydrological approaches. In this study, the Pai River basin, located in Mae Hong Son in northern Thailand, is divided into 86 subwatersheds, and 23 morphometric parameters of the watershed are extracted from the digital elevation model (DEM). In addition, the soil conservation service curve number (SCS-CN) model is used to estimate the precipitation excess, and Snyder’s synthetic unit hydrograph method is used to estimate the time to peak and time of concentration. With respect to the rainfall dataset, in this study, we combined CHIRPS data (as satellite gridded precipitation data) with rainfall data measured within the study area for the runoff analysis. According to the analysis results, 25 out of 86 subwatersheds are classified as highly susceptible areas to flash floods. The similarities in the morphometric parameters representing watersheds in highly flash flood-susceptible areas indicate that this categorization included areas with high relief, high relief ratios, high ruggedness ratios, high stream frequencies, high texture ratios, high annual runoff, high peak discharge, low elongation ratios, and low lemniscates ratios.

Many landslides triggered by intense rainfall have occurred in mountainous areas in Thailand, causing major economic losses and infrastructure damage. Extreme daily rainfall is a significant trigger for hillslope instability. Increases in extreme daily rainfall intensity due to climate change may be one of the key factors responsible for the increased landslides. Thus, in this context, changes in the intensity of extreme daily rainfall in Chiang Mai Province in North Thailand and their effects on hillslope stability are analyzed. Extreme rainfall is modeled using a generalized extreme value distribution and estimated for various return periods. A numerical analysis of seepage and an infinite slope stability model are combined to understand the hillslope response under extreme rainfall conditions. The analysis period is divided into two periods of 34 years: 1952 to 1985 and 1986 to 2019. According to the analysis results, the distribution of extreme daily rainfall changes in terms of location. The average annual daily maximum rainfall increased by approximately 11.13%. The maximum decrease in the safety factor is approximately 4.5%; therefore, these changes in extreme daily rainfall should be considered in future landslide prevention policies.

A landslide can generate large amounts of debris in the form of boulders, cobbles, soil, and wood. The woody debris produced by a landslide flows into a downstream river or village; it can form obstructions in the stream and destroy houses. In this study, we aimed to develop a procedure for estimating woody debris recruitment into streams following a landslide. Understanding the volume of woody debris can help predict and prevent hazards from this debris. The proposed procedure combines a shallow landslide model, tree density data, and observational data following landslide occurrence. The study site is a sub-watershed of the Omoto River watershed in the town of Iwaizumi in Iwate prefecture in Japan; this town was affected by Typhoon Lionrock in 2016. Typhoon Lionrock delivered over 200 mm of rainfall in 24 h and induced many landslides. Based on field surveys, we found that approximately 524 m³ of woody debris jammed the narrow section under a railway bridge (including voids) and approximately 178 m³ of woody debris to formed a dam in the stream channel of the target watershed (including voids). Using the proposed protocol, we estimate that woody debris recruitment to the stream was approximately 638 m³.