Shifting from traditional landslide occurrence modeling to scenario estimation with a "glass-box" machine learning.

Extreme rainfall events represent one of the main triggers of landslides. As climate change continues to reshape global weather patterns, the frequency and intensity of such events are increasing, amplifying landslide occurrences and associated threats to communities. In this contribution, we analyze relationships between landslide occurrence and extreme rainfall events by using a "glass-box" machine learning model, namely Explainable Boosting Machine. What sets these models as a "glass-box" technique is their exact intelligibility, offering transparent explanations for their predictions. We leverage these capabilities to model the landslide occurrence induced by an extreme rainfall event in the form of spatial probability (i.e., susceptibility). In doing so, we use the heavy rainfall event in the Misa River Basin (Central Italy) on September 15, 2022. Notably, we introduce a rainfall anomaly among our set of predictors to express the intensity of the event compared to past rainfall patterns. Spatial variable selection and model evaluation through random and spatial routines are incorporated into our protocol. Our findings highlight the critical role of the rainfall anomaly as the most important variable in modeling landslide susceptibility. Furthermore, we leverage the dynamic nature of such a variable to estimate landslide occurrence under different rainfall scenarios.

Perspectives on total phosphorus response in rivers: Examining the influence of rainfall extremes and post-dry rainfall.

Eutrophication is a significant environmental problem caused by nutrient loads from both point and non-point sources. Weather variables, particularly precipitation, affect the concentration of nutrients in water bodies, particularly those from non-point sources, in two contrasting ways. Heavy precipitation causes surface runoff which transports pollutants to rivers and increases nutrient concentration. Conversely, increased river flow can dilute the concentration, lowering it. This study investigates the impact of extreme precipitation, prolonged precipitation, and precipitation after a dry period on the total phosphorus concentration in the Moehne and Erft rivers in Germany, given the projected increase in frequency of extreme precipitation events and long drought periods due to climate change. The study comprises two parts: selecting extreme weather days from 2001 to 2021 and comparing observed Total Phosphorus concentrations with estimated concentrations derived from Generalized Additive Models and linear regression based on the discharge-concentration relationship. Changes in river TP concentration in response to continuous precipitation and precipitation after a dry period were also studied. Our results showed that during wet extreme and post-dry period rainfall events, TP concentration consistently surpassed expected values, underscoring the profound influence of intense rainfall on nutrient mobilization. However, we observed the impact of continuous rainfall to be non-unidirectional. Our work is distinguished by three key innovations: 1) addressing limitations in studying the effects of extreme weather on water quality due to limited temporal resolution, 2) incorporating both linear and non-linear modeling approaches for discharge-concentration relationships, and 3) performing a comprehensive analysis of temporal and spatial patterns of Total Phosphorus concentrations in response to varying rainfall patterns.

A dataset to measure global climate physical risk.

Extreme climate events have become more frequent and have had serious impacts on the global community. Consequently, the risk associated with climate change has gained increasing attention and has been considered as a new source of risk factors. To understand the socio-economic impacts of this new risk, systematically measuring risk around the world is critical for researchers and policymakers. Building on daily observations from meteorological stations, a Climate Physical Risk Index (CPRI) dataset is constructed for 170 countries, paying special attention to four extreme climate events: extreme low temperature (LTD), extreme high temperature (HTD), extreme rainfall (ERD), and extreme drought (EDD). A comprehensive index of climate physical risk for each country has also been constructed, covering the period from 1993 to 2023. The dataset will be updated regularly. Subnational indices or more detailed regional indices are available upon request.

Extreme rainfall erosivity: Research advances and future perspectives.

Extreme rainfall erosivity, the capacity of intense rainfall to induce soil erosion, is vital for anticipating future impacts on soil conservation. Despite extensive research, significant differences persist in terms of understanding influencing mechanisms, potential impacts, estimation models and future trends of extreme rainfall erosivity. Quantitatively describing extreme rainfall erosivity remains a key issue in existing research. In this study, we comprehensively reviewed the literature to assess the relationships between extreme rainfall characteristics and rainfall erosivity, between extreme rainfall erosivity and soil erosion, estimation models and trend prediction. The aim was to summarize previous related research and achievements, providing a better understanding of the generation, impacts and future trends of extreme rainfall erosivity. Future research directions should include identifying the thresholds of extreme rainfall events, increasing research attention on tropical cyclones in terms of rainfall erosivity, considering on the impact of extreme rainfall erosivity on soil erosion, and improving rainfall erosivity estimation and simulation prediction methods. This study could contribute to adapting to global climate change and aiding in formulating soil erosion prevention and environmental protection recommendations.

Climate change amplified the 2009 extreme landslide event in Austria.

Landslides are an important natural hazard in mountainous regions. Given the triggering and preconditioning by meteorological conditions, it is known that landslide risk may change in a warming climate, but whether climate change has already affected individual landslide events is still an open question, partly owing to landslide data limitations and methodological challenges in climate impact attribution. Here, we demonstrate the substantial influence of anthropogenic climate change on a severe event in the southeastern Alpine forelands with some estimated 952 individual landslides in June 2009. Our study is based on conditional event attribution complemented by an assessment of changes in atmospheric circulation. Using this approach, we simulate the meteorological event under observed and a range of counterfactual conditions of no climate change and explicitly predict the landslide occurrence probability for these conditions. We find that up to 10%, i.e., 95 landslides, can be attributed to climate change.

The effects of climate change-induced flooding on harvest failure in Burkina Faso: case study.

Background: Climate change leads to more frequent and severe extreme weather events including floods, heatwaves, heavy rainfalls, and droughts. In contrast to the majority of research on weather extremes in sub-Saharan Africa, which focus primarily on how a lack of rainfall causes droughts, this paper aims to elucidate the effect of flooding on harvest failure in rural Burkina Faso. Methods: We conducted a case study in north-western Nouna, Burkina Faso, between August and December 2021 covering a study population of n = 180 participants. The study comprised four components: (i) interviews with farmers (n = 180) on whether any of their fields had been inundated and if so, on harvest loss on these fields; (ii) determining the feasibility of using Sentinel-2 satellite images to validate study participants reports of floods; (iii) characterizing short-term weather including frequency and duration, of extreme rainfall events within the study area, as well as comparing cumulative rainfall (long-term) over the past 50 years; and (v), estimating both the food energy and economic loss of harvest failure due to flooding. Results: 49% of interviewed farmers (n = 88) reported that floods had damaged at least one of their fields. Some fields (n = 13, 7%) had no harvest due to flooding, while some farmers (n = 14, 8%) had lost part of their harvest. Images from the Sentinel-2-Satellite indicated that reported and remotely observed flooding were consistent. According to time series of data from the local weather station, there has been an increase irregular rainfall distribution and at the same time of cumulative annual rainfall in Nouna. Furthermore, a first illustrative calculation allowed us to estimate the amount of energy lost when one hectare of a common crop is flooded. Conclusion: This case study demonstrated that flood-related harvest failures leading to crop losses in sub-Saharan Africa, exemplified by Burkina Faso, are likely to be substantial. This study serves as a proof-of-principle for flooding effects on food security. This could provide more detail for agricultural adaptation and mitigation strategies. Inundation-vulnerable fields need alternative and novel management practices, which may only be effectively implemented if agricultural institutions and national policy-making bodies receive evidence of flooding e.g., from remote sensing.

State-space prior distribution for parameter of nonhomogeneous Poisson spatiotemporal models.

This article proposes a new class of nonhomogeneous Poisson spatiotemporal model. In this approach, we use a state-space model-based prior distribution to handle the scale and shape parameters of the Weibull intensity function. The proposed prior distribution enables the inclusion of changes in the behavior of the intensity function over time. In defining the spatial correlation function of the model, we include anisotropy via spatial deformation. We estimate the model parameters from a Bayesian perspective, employ the Markov chain Monte Carlo approach, and validate this estimation procedure through a simulation exercise. Finally, the extreme rainfall in the southern semiarid region in northeastern Brazil is analyzed using the R10mm index. The proposed model showed better fit and prediction ability than did other nonhomogeneous Poisson spatiotemporal models available in the literature. This improvement in performance is mainly due to the flexibility of the intensity function that is achieved by allowing the incorporation, in time, of the climatic characteristics of this region.

Exploring the spatial pattern of house collapse rates caused by extreme rainfall in central China: The role of natural and social factors.

The collapse of houses represents a prominent hazard associated with floods, mudslides, and other disastrous events resulting from extreme rainfall. Nevertheless, previous research in this area has been insufficiently dedicated to comprehending the factors that specifically contribute to house collapse triggered by extreme rainfall. This study endeavors to address this knowledge gap by proposing a hypothesis that the occurrence of house collapse, induced by extreme rainfall, demonstrates spatial heterogeneity and is subject to the interactive impacts of various factors. In the study, we investigate the relationship between house collapse rates and natural and social factors in the provinces of Henan, Shanxi, and Shaanxi provinces in 2021. These provinces are representative of flood-prone areas in central China. Spatial scan statistics and GeoDetector model were used to analyze spatial hotspot areas of house collapse rates and determinant power of natural and social factors on the spatial heterogeneity of house collapse rates, respectively. Our analysis reveals that the spatial hotspot areas predominantly concentrated in regions characterized by high rainfall, including areas along riverbanks and low-lying regions. Multiple factors contribute to the variations in house collapse rates. Among these factors, precipitation (q = 0.32) is the most significant, followed by the ratio of brick-concrete houses (q = 0.24), per capita GDP (q = 0.13), elevation (q = 0.13) and other factors. Notably, the interaction of precipitation and slope explains 63 % of the damage pattern, making it the strongest causal factor. The results substantiate our initial hypothesis and underscore the fact that the pattern of damage does not solely rely on a singular factor but rather on the interaction of multiple factors. These findings hold significance in advancing the formulation of more precise strategies aimed at bolstering safety measures and safeguarding properties within regions susceptible to flooding.

Mass tourism urban destinations and climate change in small islands: resilience to extreme rainfall in the Canary Islands.

The Canary Islands are one of the main destinations for mass tourism in the European context, characterized by the absence of seasonality in tourist activity. Moreover, the level of activity increases during the winters, coinciding with a greater probability of extreme rainfall events, whose danger seems to be increasing as a result of climate change. Owing to its pronounced orography, the southern coast of the island of Gran Canaria houses several tourist settlements built along ravines and steeply sloping terrain. This scenario presents considerable risk because of spatial probability of landslide occurrence. The case of San Agustín, especially, serves to test the model of tourist urbanization along the hillside, demonstrating its high fragility in the face of extreme rainfall events. Especially owing to its importance in providing assistance in emergency situations, its vulnerability has been analyzed with regard to accessibility, which is entirely dependent on road mobility. The growth model of San Agustín serves as an example of mass tourism in small islands, allowing urban planners and designers to assess corrective measures based on managing its existing road infrastructure and open spaces right from the planning stage.

Phytoplankton functional responses induced by extreme hydrological events in a tropical reservoir.

Climate change is affecting the global hydrological cycle, causing drastic changes in precipitation patterns. Extreme climatic events are becoming more frequent and intense than in the past, leading to water-level fluctuations and affecting aquatic ecosystems. Semiarid regions are very susceptible to changing climate. We analyzed a 10 years dataset from a tropical semiarid reservoir during extreme hydrological events (heavy rains and prolonged drought), and evaluated phytoplankton functional responses to environmental conditions. We found, as hypothesized, that phytoplankton functional structure change in a temporal scale due to water-volume fluctuation induced by the rainfall pattern. Depth and inorganic material acted as environmental filters selecting phytoplankton groups. High water level seems to improve water quality and low water level worsen it. Colonial and filamentous cyanobacteria dominate the wet period; however, it may have a critical threshold during severe periods of drought, which will lead to dominance of groups well adapted to low light conditions and with mixotrophic metabolism. Phytoplankton functional approaches can simplify phytoplankton identification and reflect better the environmental conditions than the taxonomic approach. Therefore, these approaches can help to understand the shifts in aquatic ecosystems under extreme hydrological events and predict functional response of phytoplankton being an important tool to water management and conservation. Supplementary Information: The online version contains supplementary material available at 10.1007/s10750-023-05241-3.

Multicriteria analysis for flood risk map development: a hierarchical method applied to Brazilian cities.

Floods have caused socio-economic and environmental damage globally and, thus, require research. Several factors influence flooding events, such as extreme rainfall, physical characteristics, and local anthropogenic factors; therefore, such factors are essential for mapping flood risk areas and enabling measures that mitigate the damage they cause. This study aimed to map and analyze regions susceptible to flood risk in three different study areas belonging to the same Atlantic Forest biome, in which flood disasters are recurrent. Due to the presence of numerous factors, a multicriteria analysis using the Analytical Hierarchical Process was conducted. First, a geospatial database was composed of layers of elevation, slope, drainage distance, soil drainage, soil hydrological group, precipitation, relief, and land use and cover. Flood risk maps for the study area were then generated, and patterns in the study areas were verified, with the greatest influence being exerted by intense precipitation on consecutive days, elevation at the edges of the channel with low altimetric variation and a flat combination, densely built areas close to the banks of the main river, and an expressive water mass in the main watercourse. The results demonstrate that these characteristics together can indicate the occurrence of flooding events.

Impacts of Eucalyptus plantation on soil and water losses in a typical small watershed in mountainous area of southern China.

Unreasonable exploitation of artificial forest causes severe soil erosion in the mountainous areas of sou-thern China. The spatial-temporal variations of soil erosion in typical small watershed with artificial forest has signifi-cant implications for artificial forest exploitation and sustainable development of mountainous ecological environment. In this study, we used revised universal soil loss equation (RUSLE) and geographic information system (GIS) to evaluate the spatial and temporal variations of soil erosion and its key drivers of Dadingshan watershed in mountainous area of western Guangdong. The results showed that the erosion modulus was 1948.1 t·km-2·a-1 (belonging to light erosion) in the Dadingshan watershed. However, the spatial variation of soil erosion was substantial, with variation coefficient of 5.12. The maximal soil erosion modulus was 191127 t·km-2·a-1. Slight erosion (<500 t·km-2·a-1) accounted for 80.6% of the total watershed area. The moderate erosion and above (>2500 t·km-2·a-1) were mainly distributed in young Eucalyptus forest area with less than 30% of the vegetation coverage, which contributed nearly 75.7% of total soil erosion. During 2014-2019, the interannual variations of mean erosion of Dadingshan catchment was modest, but the spatial variation of soil erosion was large. Vegetation cover, slope, and rainfall were key drivers of such variation. The destruction of natural vegetation resulted by plantation exploitation was the primary cause of soil erosion in afforestation areas. Soil erosion significantly increased with the increases of slope gradient in the young forest area, which was aggravated by extreme rainfall. However, soil erosion gradually decreased with the increases of the age of Eucalypt plantation. Therefore, the hot spot of soil erosion was young forest areas of Eucalypt plantation with slope >25°, and the key period for soil erosion control was the first 2-3 years after Eucalyptus planting. We suggested that reasonable afforestation measures should be used in area with >25° slopes, and that the destruction of natural vegetation should be avoided on hillslope with >35° slope gradient. The road construction standards and forest management should be further improved to address the challenge of extreme rainfalls.

Evaluating soil loss under land use management and extreme rainfall.

Topsoil loss is a widespread environmental concern causing adverse impacts on natural and human systems. Severe weather accompanied with human activities can exacerbate this issue degrading soil health and consequently accelerating global and regional food insecurity. Erosion impairs soil physical and chemical properties such as infiltration rate, water holding capacity, loss of nutrients including soil carbon and nitrogen. Although, temporal properties of a rainfall event have meaningful implications, spatial heterogeneity of a rainfall contributes substantially and cannot be overlooked. Therefore, in this study, we investigated soil loss using weather radar NEXRAD data. We developed extreme rainfall (ER) scenarios and land use practices (nomgt, S0, S1, S2, and S3) and evaluated the watershed response. We found that grazing can manifold soil loss, and if accompanied with extreme rainfalls, soil loss accelerates impacting different subbasins each time. Our results suggest that spatial heterogeneity of ERs can be more significant in individual extreme rainfalls, however, over a year, soil moisture and type of the management practices (grazing and farming) could contribute more to topsoil loss. We classified watershed subbasins into different classes of soil loss severity to determine the soil loss hotspots. Soil loss can go as high as 350 (ton/ha/yr) under the ERs. Land use practices can increase erosion by 3600%. Slight increase in rainfall concentration (S1) can put vulnerable subbasins in extremely severe class (>150 ton/ha/yr). Under moderate increase in the rainfall concentration (S2) more subbasins fall into extremely severe category yielding approximately 200 ton/ha/yr. Under high increase in rainfall concentration (S3) almost all the subbasins fall into the extremely severe class yielding >200 ton/ha/yr. We found that in vulnerable subbasins, up to 10% increase in (Concentration Ratio Index) CI can increase annual soil loss up to 75%. Single ER can generate up to 35% of annual soil loss. Under one ER event soil loss hotspot subbasins can lose up to 160 ton/ha/day. 32% and 80% increase in rainfall amount for an ER event can increase soil loss by 94% and 285% respectively. The results, also, reveal that grazing and farming can be responsible for up 50% of soil loss. Our findings indicate the importance of site-specific managements to mitigate soil loss and all the consequences. Our study can help in better soil loss management implementation. Insights of our study may also help in water quality control and flood mitigation planning efforts.

Impact of extreme rainfall on non-point source nitrogen loss in coastal basins of Laizhou Bay, China.

Extreme rainfalls often lead to large amounts of nitrogen (N) loss from river basins. However, the composition and spatial variation of N loss caused by extreme events and the effects of control measures are not well understood. To shed light into this question, the Soil and Water Assessment Tool (SWAT) was used to evaluate the spatiotemporal characteristics of organic and inorganic nitrogen (ON and IN) losses in the coastal basins of Laizhou Bay during typhoons Rumbia and Lekima. The effects of best management practices on controlling N loss were also explored during such extreme rainfall events. Results showed that extreme rainfall promoted transport of ON more than IN. The mass of ON and IN transported by the two typhoons exceeded 57 % and 39 % of the average annual N flux, respectively, and the loads were positively correlated with streamflow. During the two typhoons, the loss of ON was mainly concentrated in areas with steep slopes (θ > 15°) and natural vegetation (forests, grasslands, and shrublands). The IN loss was higher in areas with a 5-10° slope. Furthermore, subsurface flow was the main IN transport mechanism in areas with steep slope (θ > 5°). Simulations showed that implementation of filter strips in areas with slopes exceeding 10° can reduce N loss, with much greater reductions in ON (>36 %) than IN (>0.3 %). This study provides important insights into N loss during extreme events and the key role filter strips can play in trapping them before they reach downstream waterbodies.

Framework for rainfall-triggered landslide-prone critical infrastructure zonation.

Rainfall-induced landslides cause frequent disruptions to critical infrastructure in mountainous countries. Climate change is altering rainfall patterns and localizing extreme rainfall events, increasing the occurrence of landslides. For planning climate-resilient critical infrastructure in landslide-prone regions, it is urgent to understand the changing landslide susceptibility in relation to changing rainfall extremes and spatially overlay them with critical infrastructure to determine risk zones. As such, areas requiring financial reinforcements can be prioritized. In this paper, we develop a framework linking changing rainfall extremes to landslide susceptibility and intensity of critical infrastructure - exemplified on a national scale using Nepal as a case study. First, we define a set of 21 different unique rainfall indices that describe extreme and localized rainfall. Second, we prepare a new annual (2016-2020) inventory of 107,900 landslides in Nepal mapped on PlanetScope satellite imagery. Next, we prepare a landslide susceptibility map by training a random forest model using the collected extreme rainfall indices and landslide locations in combination with spatial data on topography. Fourth, we construct a gridded critical infrastructure spatial density map that quantifies the intensity of infrastructure (i.e., transportation, energy, telecommunication, waste, water, health, and education) at each grid location using OpenStreetMap. The landslide susceptibility map classified Nepal's topography into low (36 %), medium (33 %), and (32 %) high rainfall-triggered landslide susceptibility zones. The landslide susceptibility map had an average area under the receiver characteristic curve value of 0.94. Finally, we overlay the landslide susceptibility map with the critical infrastructure intensity to identify areas needing financial reinforcement. Our framework reasonably mapped critical infrastructure hotspots in Nepal prone to landslides on a 1 km grid. The hotspots are mainly concentrated along major national highways and in provinces 4, 3, and 1, highlighting the need for improved land management practices. These hotspots need spatial prioritization regarding climate-resilient critical infrastructure financing and slope conservation policies. The research data, output maps, and code are publicly released via an ArcGIS WebApp and GitHub repository. The framework is scalable and can be used for developing infrastructure financing strategies for landslide mountain regions and countries.

Extreme rainfall characterisation under climate change and rapid population growth in the city of Niamey, Niger.

Since recent years, the Sahel semiarid region has experienced devastating floods-causing significant losses and damages. The present paper attempts to characterise extreme rainfalls responsible for pluvial floods in the city of Niamey, in Niger, under climate change and rapid population growth. Past damaging rainfall records spanning 1992-2015 were used to estimate the optimal temporal scale and to define a threshold for extreme rainfall. The characteristics of extreme rainfalls were then assessed under stationary and non-stationary conditions using peaks over threshold (POT) with the generalised pareto distribution (GDP). In the non-stationary POT, population data was used as threshold covariate whereas air temperature was used as scale parameter covariate. A suitable temporal scale of 3 h was found, whereas the threshold depth was 28.71 mm under stationary conditions and between 21 and 27 mm for the time dependent threshold. The analysis of the extreme rainfall series revealed no significant trend neither in the magnitude nor in the frequency. The influence of air temperature in the characterization of extreme rainfall were less compared to rapid urbanisation, represented herein by population growth. By 2040, 3-hourly rainfall depths of 20 mm could be considered as extreme rainfall.

Joint Risk Analysis of Extreme Rainfall and High Tide Level Based on Extreme Value Theory in Coastal Area.

Extreme rainfall and high tide levels are the main causal factors of urban flood disasters in coastal areas. As complex interactions between these factors can exacerbate the impact of urban flood disasters in coastal areas, an associated flood risk assessment involves not only the estimation of the extreme values of each variable but also their probability of occurring simultaneously. With a consideration of the Shenzhen River Basin (China), this study used bivariate copula functions to quantitatively evaluate the joint risk of extreme rainfall and a high tide level. The results showed that a significant positive correlation exists between extreme rainfall and the corresponding high tide level, and that if the positive dependency was ignored, the probability of simultaneous extreme events would be underestimated. If a dangerous event is defined as one in which heavy rainfall and high tide level events occur concurrently, the "AND" joint return period based on the annual maxima method should be adopted. If a dangerous event is defined as one in which either only a heavy rainfall or a high tide level event occurs, the "OR" joint return period should be adopted. The results represent a theoretical basis and decision-making support for flood risk management and flood prevention/reduction in coastal areas.

Assessing the mechanism for flood control: a case of plain river network cities under extreme rainfalls.

In recent years, frequent floods hit Chinese cities and caused heavy casualties and property losses, making China faced with severe flood problems. In this study, Nanhai Future City in the IX Flood Control Area of Yancheng City, Jiangsu Province, was selected as the research area to simulate water-level changes under different control schemes meeting extreme rainfalls. MIKE model simulated the inundation with the designed storm of different return periods. The results showed that flooding inside the research area was severe. Higher drainage capacity of the pump stations with more engineering and non-engineering measures can reduce the adverse effect of extreme rainfall. The results provide a reference for planning future infrastructure and flood control decisions for Nanhai Future City and the surrounding areas.

Spatial dependence of floods shaped by extreme rainfall under the influence of urbanization.

Weather system are spatially and temporally dependent, and these dependencies can result in flood events with similar behaviors. While it is well known that urbanization profoundly impacts the flood generation process, much less is known about the impacts of urbanization on the spatial dependence of floods, a major determinant of flood risk severity. To this end, a scheme was proposed to detect the flood dependence variations influenced by urbanization. Based on the scheme, we found that 1) the flood dependence can be weakened with extreme rainfall increasing from a short return period to a long return period; 2) The flood dependencies can be amplified in slightly urbanized regions and mitigated in highly urbanized regions due to intensifying urbanization. In addition, the change characteristics of the flood dependencies influenced by urbanization are first identified from the perspective of network structure. We found that urbanization can significantly affect the network structure (i.e., hub and connectivity) of flood dependence, especially in highly urbanized regions. The catchments with high hub and connectivity are prone to widespread floods and should be given more attention in flood warning and control management work, which can contribute to helping defend against floods in hazard-prone areas.

The impact of different temporal resolution rainfall data on biosphere model of long-term safety assessment in groundwater release scenario for radioactive waste disposal in the tropical monsoon zone.

The impact of different temporal resolutions of rainfall data on the Biosphere assessment for radioactive waste disposal in the tropical monsoon region with concentrated rainfall is evaluated in this research. Two scenarios are considered to verify the effect of release location. A simplified surface water budget model is used to generate the surface water flow rates for the biosphere model, which is implemented using three different temporal averaging intervals to consider the uncertainty caused by short-term impact. Kaohsiung in Taiwan is chosen as an example because of its extreme rainfall distribution. The results show that it is improper to use the annual rainfall data for this case, and it is suggested to consider a 20% margin to cover the underestimation of dose if the monthly rainfall data are used. The results of the biosphere models built with different timesteps show that a non-negligible difference occurs for radionuclides with a low Kd value in the river water release scenario and no difference for the well water release scenario.