Rain Stations Research Articles

Zonation of Flood-Prone Areas Based on Remote Sensing Data and Hydrodynamic Models

Many watersheds in Manokwari are classified as flood-prone areas with a frequency of more than 1 flood per year. Limited rain stations and climate stations make it difficult to the zoning of flood-prone areas. This study aims to utilize remote sensing data and hydrodynamic models to zone flood-prone areas in watersheds (DAS). The research was conducted in the Wosi Watershed - Manokwari Regency - West Papua Province. The data used in this research is DEMNAS topographic data, Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) data acquired 1996 – 2020, Sentinel 2 imagery acquired 21 September 2020, and river maps. The peak flow in the Wosi watershed was analyzed using rational methods and flood hazard zoning was analyzed using HEC-RAS. The research showed that the Wosi River was not able to accommodate the peak flow at various return periods, consequently, the Wosi River had the potential for flooding every year. The results of this study are relevant to actual events, therefore remote sensing data and hydrodynamic models can be used to analyze peak flow and flood hazard zoning.Keywords: Climate Hazards Group Infrared Precipitation with Stations, DEMNAS, HEC-RAS, Satellite Imagery, Sentinel 2.

Climate Change Impact on Drought Characteristics in North Lombok Regency

Lombok Island is part of Indonesia which has a high vulnerability to drought. The area that often experienced drought severely on Lombok Island is North Lombok Regency. This study wants to examine the effect of climate change on drought characteristics in this region using the standardized precipitation index method using four rain stations in the North Lombok region rainfall recording data of 20 years (1999-2018). The analyses were taken into two groups of a 10-year time scale, 1999-2008 and 2009-2018, to overview both of the 10-years drought characteristics. Drought index values ​​were observed on a period of 1-month (SPI1), 3-months (SPI3), and 6-months (SPI6) to get the map of drought duration, drought magnitude, drought intensity, and relative frequency. The results show that the drought characteristics in Lombok showed a decrease significantly in all measurement parameters used. The SPI3 results have shown the drought duration was decreasing in the last 10 years by 87%, from 2,43 to 0.03 months. The strength of the drought decreased by 88%, from -1.03 to 0.12 (severely dry to normal). Then the drought intensity became lower up to 87% from -0.22 to -0.03 monthly, and the relative frequency of drought events decreased by 83%. Signs of the decline were further followed by the same decreasing trends in the 6-month drought index, SPI6. The decrease in the drought parameter index illustrates that climate change impact has reduced the risk of drought disaster on the island of Lombok in the future

Rainfall Trends Analysis in context of Climate Change in Imam Turki Bin Abdullah Royal Nature Reserve (ITBA) – Saudi Arabia

Rainfall trend analysis, on different spatial and temporal scales, has been used during the past century as the indicate of global climate change by the scientific community. According to some recent studies, the Nature Reserve (ITBA) has been suffering a rainfall decrease, especially in the second half of the studied period (1993-2018). The aim of the present study is to analyze the rainfall trends using the time series of 52 years in five rain stations (Linah, Rafha, Ha’il, Dawmat Al Jandal and Hafr Al Batin) surrounding the Nature Reserve of (ITBA) that extended on area of about 91500 km2 in the Northeast of Saudi Arabia. The trends analysis of annual rainfall, rainy days, shows that the fluctuations or variations in climatic parameters is a recurring phenomena in the studied stations. Inter-annual variability of the rainfall and the actual mean of daily rainfall are characterized by the high coefficients of variation, varying from 0.73 (Linah) to 0.83 (Ha’il) and from 0.53 (Linah) to 1.09 (Rafha), respectively. The variance test of Fmax-ratio method indicates the homogeneous annual rainfall distribution at Linah, Rafha and Ha’il, the homogenous rainy days distribution at only Ha’il and the homogeneous distribution of the actual mean of the daily rainfall at only Linah. The trends of annual rainfall, rainy days and actual mean of the daily rainfall obtained by using T-student test are significant in Ha’il and Dawmat Al Jandal, in all stations and in only Hafr Al Batin, respectively.

Annual flow trend analyses in the Tuntang watershed caused by Kedungsepur national strategic area development

Tuntang watershed is one of the national priority watersheds in the Jratunseluna river basin that has a big potential for water availability. The development of a national strategic area in the Tuntang watershed causes changes in land cover in the Tuntang watershed which can result in reduced water catchment areas. The purpose of this study was to simulate the effect of land change on annual discharge trends in the Tuntang watershed and analyze the water balance projection based on land change. This study uses land cover data for the Tuntang watershed in 2019 and land cover for the Kedungsepur National Strategic Area (KSN) development plan. Daily rainfall data for 11 years from 2010 to 2020 was obtained through the Global Precipitation Measurement (GPM) satellite and validated using data from 7 rain stations located in the Tuntang watershed. Meanwhile, evapotranspiration data was obtained through the Global Land Data Assimilation System. In addition, Digital Elevation Model (DEM), flow accumulation, and direction flow data obtained through HydroSHEDS are used to provide information on the surface shape of the Tuntang watershed. Then the data that has been obtained is used as input for the simulation of the spatial hydrological model to determine the annual discharge in the Tuntang watershed using the Rainfall-runoff-inundation (RRI) model. The result of this simulation is an annual discharge for 11 years. Furthermore, a comparison is made between the flow rate of land cover map input in 2019 with land cover input for Kedungsepur KSN development. Based on the result, it is known that there has been a change in the area of the water absorption region in the development of the Kedungsepur KSN, this change has resulted in an increased extreme discharge trend and reduced annual water flow in the Tuntang watershed region.

Characteristics of Extreme Rainfall Events in North Sumatra

Extreme rain is an event that has an impact on various sectors of life. Therefore, this study aims to determine the spatial and temporal characteristics of extreme rain events in North Sumatra Province in the time period between 1991-2020. In accordance with the WMO recommended rain index, the frequency and intensity of rain were calculated using 50 rain stations spread across North Sumatra where the data had passed the quality control test for empty data. Spatial patterns were analyzed by mapping the climatological mean of the indices then for trend patterns were tested using the Mann-Kendal non-parametric test. The results showed that extreme rainfall events with low frequency and intensity occurred on the east coast while high frequency and intensity occurred in mountainous areas. Based on the temporal trend test, it shows that in general an insignificant trend dominates in this area. Significant and consistently increasing trends are only found at several points, namely in the districts of Deli Serdang (5 station), Batu Bara (1 station), Humbang Hasundutan (1 station), Langkat (2 station), Labuhan Batu Utara (1 station), Medan (1 station), Pematang Siantar (1 station) and Serdang Bedagai (1 station). Meanwhile, a significant and consistently decreasing trend was found in Asahan Regency (1 station), Simalungun (2 station), North Tapanuli (1 station), and North Labuhan Batu (1 station).

Rainfall Infiltration Impact on Road Embankment Stability

This study evaluates the stability of a widened road embankment slope by considering the effect of rain infiltration, which is often ignored in practical design. Rainfall infiltrates the soil and alters pore water pressure, reducing its shear strength and potentially causing slope failure. The Phi Index method is used to determine rain infiltration values, using rainfall data from four nearby rain stations over 10 years. The changes in pore water pressure resulting from rain infiltration are incorporated into the slope stability analysis by the limit equilibrium concept using the Spencer method. Back analysis is used to determine the soil shear strength parameters since the CPT test results provide a correlation with values within a certain range. Soil parameters resulting in a safety factor value close to 1.25 (critical condition) were considered representative field conditions. The analysis shows a decrease in the safety factor from 1.50 to 1.31 (12%). These findings emphasize the importance of accounting for rain infiltration in road embankment stability analyses, particularly in areas with high rainfall where the safety factor may fall below the minimum required by Indonesian geotechnical code. When an infiltration effect analysis is not conducted, the safety factor should be increased at least 12% from normal conditions.

Climate changes affect to The Water avaibility in Sausu Watershed Catchment Area (Case Study)

Climate change in Indonesia is marked by changes in temperature, rainfall patterns, sea level and climate variability. The climate condition of the Sausu watershed is a natural process that occurs in a rainfall patterns, magnitude and changes in air temperature. This study was conducted to identify the dangers of climate change to the water sector in the Sausu watershed. The analytical method used is Modified Penman evapotranspiration, water balance analysis model F.J. Mock, Makesens, Cumulative Distribution Frequency (CDF) statistical analysis and hazard weighting. This study uses secondary data, namely rainfall sourced from the Sausu rain station in 2002-2021 and climate data (climatology) sourced from BMKG in 2002-2021. Results from this study is that climate change occurs in the Sausu watershed which is characterized by a positive Z value (an increasing trend) and negative (a decreasing trend). A positive Z value (increasing trend) occurs in June, and a negative Z value (declining trend) occurs in December. The danger of decreasing water availability (PKA) occurs in the period 2052-2061 with a low level of danger. Flood hazard occurs in the period 2052-2061 with a high level of danger. Drought hazards occur in the period 2042-2051 and the period 2052-2061 with a low level of danger.

Generalized Space Time Autoregressive Integrated Moving Average (GSTARIMA) dalam Peramalan Data Curah Hujan di Kota Makassar

Modeling of rainfall data using time series data involving location elements has not been widely carried out. One model that involves elements of time and location is Space Time Autoregressive (STAR). The development of the STAR model which assumes that each location has heterogeneous characteristics is the Generalized Space Time Autoregressive Integrated Moving Average (GSTARIMA) model. The purpose of this research is to get the best GSTARIMA model and forecast rainfall data in Makassar City based on the best GSTARIMA model. This model incorporates time and geographic dependencies with different parameters for each location. The data used is Makassar city's monthly rainfall data at the Bawil IV/Panaikang, Biring Romang/Panakkukang and Stammar Paotere rain stations from January 2017 to September 2021. Autoregressive (AR) and Moving Average (MA) orders were identified using the Space Time Autocorrelation plot. Function (STACF) and Space Time Partial Autocorrelation Function (STPACF). The spatial order used in this study is spatial order 1 with an inverse distance weighting matrix and normalized cross-correlation. Parameters were estimated using the Generalized Least Squares (GLS) method. The best model for predicting rainfall in the city of Makassar is the GSTARIMA (1,0,0) (1,1,0)12 model using an inverse distance weighting matrix with the smallest average Root Mean Square Error (RMSE) of 132.9661.

ANALISIS CLUSTER CURAH HUJAN TAHUNAN DI INDONESIA

Water is critical for long-term socioeconomic growth. Rainfall is one of the water sources.. Many factors contribute to disparities of rainfall patterns in Indonesia. The study aims to detect and map the implication of clustering annual rainfall patterns in each province in Indonesia using the annual provincial rainfall data 2011 to 2015. K Means Clustering using an unsupervised learning technique was applied for rainfall events by employing rainfall measurement data from one of Indonesia's rain stations in each province. The clustering results indicate that the two provinces in cluster I have very low water availability, cluster II has moderate rainfall with rainy days under 200 days, i.e. wet months of seven months, and cluster III has a very wide variation in maximum and minimal rainfall. The results demonstrate the trend of locations vulnerable to water availability in the provinces of Central and West Sulawesi. Rainfall should be managed as a potential water source in face of the extreme dry season, particularly in three provinces in cluster III: South Sumatra, Papua, and Maluku. Rainfall clustering in Indonesia is expected to be used as a dataframe, as well as a preliminary analysis to reinforce local rainfall, as well as basis for managing and exploiting rainfall, as well as planning mitigation and adaptation to climate change in Indonesia.

A Review: Design and Build Damage Detection Equipment on Sensors and Power Supply Automatic Rain Gauge (ARG) With Long Short Term Memory Integrated

The Meteorology, Climatology, and Geophysics Agency (BMKG) technicians have successfully developed automatic rain measuring devices. This tool is called Automatic Rain Gauge - BMKG (ARG-BMKG). The existence of this instrument can replace conventional rain measuring observation systems or public rain stations in Indonesia. ARG – BMKG consists of a tipping bucket sensor, solar panels, GPRS modem, dry battery, and data logger. This repeated operation causes sensor measurement errors due to damage to the sensor due to the sensor voltage supply not meeting specifications, resulting in inaccurate data sent. Predicting sensor damage can be done with predictive maintenance. The results of field tests in previous studies showed that the system could operate properly where the device could measure the voltage of each sensor and send data to the database A sensor damage prediction model was designed and implemented using long-sort term memory (LSTM) by generating root mean square error (rmse). The system can provide damage prediction information on the sensor, and the power supply is displayed through the website properly

Rainfall Mapping in Medan Selayang Sub-District using Geographic Information System

Indonesia is a tropical country which has two seasons, namely the rainy season and the dry season. Rainfall that occurs will vary in each area depending on the altitude of the area, climate, season, and other factors. Medan Selayang Subdistrict is the Subdistrict with the closest distance to the rain station. Where the closer the location of the study area to the rain station, the more accurate the rainfall data obtained will be. One way to make it easier to record and process data is to create a rainfall map by utilizing spatial analysis through a Geographic Information System (GIS). The data in this study are rainfall data and geographic coordinate data obtained from the Meteorology, Climatology and Geophysics Agency in the Medan City area. After data collection, data processing was carried out using the QGIS 3.10 application. with the Inverse Distance Weighted (IDW) interpolation method. From the results of processing rainfall maps in the Medan Selayang sub-district using Inverse Distance Weighted (IDW) interpolation, the rainfall values (rainfall) ranged from 3029.74 mm/year – 3164.20 mm/year.

FREQUENCY ANALYSIS OF ANNUAL AND MAXIMUM DAILY RAINFALL IN HA’IL AREA USING PROBABILITY DISTRIBUTION MODELS

This study uses the Frequency analysis of the maximum daily and annual rainfall recorded during the period 1976-2020 at eight rain-measuring stations in the Ha’il region: Jubbah (106H-196), Baqa’a (103H-193), Simirah (105H-793), Uqlah Ibn Jibrin (208H-508) and Faydat Ibn Suwaylim (108H-198), Al Ha’it (111H-797), Al Ghazalah (215H-812) and Ha’il (101H-191), currently supervised by the Ministry of Environment, Water and Agriculture (MEWA). These stations were selected due to their locations in the Hail region, which is in great urban development causing an increase in demand for water resources to meet the growing population in urban centers. This paper compares four probability distributions (Normal, Log Normal, Gumbel and Exponential) using the rainfall data recorded in eight rain stations located in Ha’il area. Sets of data cover a period of about 45 years. Therefore, this study aims to estimate the maximum daily and annual rainfall by applying 4-four probability distributions, which are the Normal, the Log-Normal, the Exponential, and the Extreme Values-type 1, using the (Chow) model estimation. Results showed that Gumbel (EV 1) distribution has the least critical values in the tests. Hence, it is considered the best fit distribution for annual rainfall in Simirah and Baq’a. Furthermore, based on the results, it is evident that the exponential distribution is the best in Al Ha’it, Faydat Ibn Suwaylim, Al Ghazalah, Ha’il, and Jubbah. Whilst, the log normal distribution is the best in Al Uqlat. The results also showed that Gumbel (EV 1) distribution has the least critical values in the tests. Thus, it is regarded as the best-fit distribution for the maximum daily rainfall in Simirah and Ha’il, and the normal distribution is considered the best for the rest of the rain stations.

LiDAR DTM Application for Flood Inundation Modeling and Visualization (Case Study: Batu City, East Java)

LiDAR DTM has been widely used for simulating flood inundation from river overflows. This flood inundation simulation needs to be done as an alternative step in the flood control system and flood disaster mitigation to determine the estimated area affected by the flood. Flood simulation is a form of unsteady flow that is simulated with the HEC-RAS program that allows the user to perform one-dimensional steady flow, one and two-dimensional unsteady flow calculations, sediment transport/mobile bed computations, and water temperature/water quality modeling, and visualizes flooded areas and flood-affected areas using a geographic information system approach. Hydrological data is obtained from daily rainfall processing at the time of station rain. Hydrograph analysis was performed using the SCS CN model on the HEC-HMS software to estimate the flow rate and peak discharge. Watershed area, curve number, Impervious value, Initial ion, and time lag are parameters that affect hydrograph analysis. From the results of the hydrograph analysis, the peak discharge of the model is 119.992 m2/s. Meanwhile, hydraulic analysis using HEC-RAS resulted in an inundation of 364 m2 which inundated three villages, namely Bulukerto, Bumiaji and Sidomulyo villages. The depth of the inundated area ranged from 0-15 m with an average depth of 4.534 m. The Sidomulyo area affected by the flood inundation scenario is 4,924 meters with an area of 0,02 ha. Bumiaji village was flooded with an area of 0,007 ha with a scenario of inundation depth of 4,049 m. Bulukerto village which was affected by the flood has an area of 20,972,832 m2 with a scenario of inundation depth reaching 4,343 m. Then some land cover such as residential land that is inundated by this flood simulation is 0,002 ha. Buildings from residential areas that were flooded reached 209 buildings. Then the area of agricultural land affected is 0,02 ha. In addition, there are also green open spaces affected by this simulation with an inundation area of 0,003 ha.

Analysis of Satellite Rain Data UsageOn The Rationalization Activities of The Rain Post Network (Case Study: Rationalization of The Jelai Watershed Rain Post Network)

An alternative solution to the availability of inadequate rain data as hydrological data input is with the help of Global Precipitation Measurement (GPM) satellite rainfall data using remote sensing technology (satellite). The purpose of this study was to find correlations and corrections of data and validate GPM satellite data with rainfall data at rain stations and observation data in the Jelai watershed. The corrected GP M rain data validation results in Nash-Sutcliffe Efficiency (NSE), Root Mean Squared Error (RMSE), Correlation Coefficient (R), and Relative Error (KR). The validation results resulted in NSE values of 0.33, RMSE 48.54, Correlation Coefficient (R) of 0.75, and Relative Error of 0.19 for 2019 and yielded NSE values of -0.14, RMSE 100.24, Correlation Coefficient (R) of -0.36, and Relative Error of 0.23 for 2020. The overall analysis shows that GPM data can be used as an alternative to rain data if in a watershed there is a small number of rain posts that do not meet the WMO criteria. As a suggestion for further research, it is necessary to calibrate and validate by distinguishing between rain data in wet years and dry years

Spatial Climate Forecasting for Climatology Disaster Mitigation

Climate is the average weather conditions over a relatively long time. Climate change is a problem for living things, such as the agricultural sector. It affects the availability of community food reserves, public health, and socio-economics. The impact of climate change causes dramatic changes in agricultural patterns and shifts in planting time. Water availability for agricultural needs is complicated to predict, so the failure rate of agricultural production is getting bigger. Failure is due to a mismatch between the availability of water and the applied cropping pattern. This study aims to analyze and predict the spatial distribution of the Schmidt-Ferguson climate classification on Lombok Island from 2018-2035. This study generated rainfall data using the Thomas Fiering method from 18 rain stations on Lombok Island for 24 years (1994-2017). Determine the dry, moist, and wet months based on the Schmidt-Ferguson classification. From the analysis results, the most significant Q value is at the Geres Daya, Sapit, and Sopak rain stations with the classification

ANALYSIS OF SEASONAL AND ANNUAL VARIATIONS OF EXTREME TEMPERATURES AND PRECIPITATION IN ETHIOPIA

One of the most serious challenges in the world is climate change. This study examines variations in seasonal and yearly fluctuations in extreme temperatures and precipitation in Ethiopia from 1987 to 2020 using 36 meteorological stations located in various climatic zones. The outcomes of the study demonstrated increased trends in the average annual maximum and minimum temperatures, with values of 0.094∘C and 0.052∘C, respectively. Despite falling trends in the summer, most stations experienced higher winter temperatures than summer. Furthermore, the analysis revealed a trend toward an increase in the number of rain stations over the summer season, while a trend toward a decrease in the amount of precipitation for a number of stations in the winter season; otherwise, large changes expected in the dega (cold), woina dega (cool), and kola (warm) climatic zones during the summer season did not occur. Precipitation activity has been near normal in most dega (cold), woina dega (cool), and kola (warm) climatic zones over the preceding 33 years. However, precipitation in the berha (hot) climatic zone was lower than average, with fluctuating levels noted throughout the year. The average annual precipitation has been decreasing at a rate of 0.146[Formula: see text]mm per year over the past 33 years, indicating that only the winter season has shown a downward (decreasing) trend, while the other three seasons, such as summer, spring, and autumn, have shown growing trends. Furthermore, Ethiopia’s average annual rainfall continues to be on a downward trend (decreasing).

ANALYSIS OF INFILTRATION WELLS AS PRO-CONSERVATION DRAINAGE (Case Study of Kalongan, Maguwoharjo, Depok, Sleman, Yogyakarta)

Indonesia has two seasons, namely the dry season and the rainy season where the rainfall in the two seasons has a very large difference. Rainfall in the rainy season will be a source of groundwater reserves, because in the dry season there tends to be no rain. Efforts to increase groundwater reserves are carried out, among others, to use environmentally friendly drainage, namely by using infiltration wells and water will be infiltrated so as to reduce runoff.This research will begin by mapping the land covered by buildings in Kalongan and then calculating the area of land covered by the buildings. Rainfall data used is data at the Maguwoharjo rain station which is close to Kalongan with a span of 10 years. The number of infiltration wells currently available is 8 and by analyzing the needs of infiltration wells it will be known how many infiltration wells are needed so that as much water can be infiltrated to reduce runoff and increase groundwater reserves.The results of the frequency analysis for rainfall with a maximum return period of 5 years as a basis for planning infiltration wells. This large runoff during the rainy season can be reduced by having sufficient infiltration wells, so that rainwater entering the infiltration well will become a source of groundwater reserves. The total requirement for infiltration wells is 464 pieces with a diameter of 80 cm and a depth of 3 m. There are 8 infiltration wells so it is necessary to make 456 infiltration wells to reduce runoff.Keyword: groundwater, infiltration, runoff

Identification and classification of the wet Red Sea Trough over Israel

AbstractThe Red‐Sea Trough (RST) extends toward the Levant from the south and is mostly associated with dry conditions. This study analysed the rarer type, rain‐producing RSTs, referred to as ‘wet RSTs’ (WRST), using data from 1958 to 2019. A WRST archive was established, based on rain data from a study region covering the arid part of Israel, the region where this system is most active. The study also aimed to find an atmospheric proxy for rainfall, to replace precipitation estimates obtained from climate models. A WRST day was defined as a day when the RST was identified and rainfall of ≥0.1 mm was observed by at least two of the 39 rain stations used. The annual occurrence of WRSTs is 16.4 days, ~16% of all RSTs, peaking in the winter. The relative contribution of WRSTs to the annual rainfall ranges from 16% in the northern part of the study region to 50% in the south. The tropical air intrusion was quantified through trajectory analysis and the results revealed that in most WRST days, 72%, the fraction of air originating from the tropics was less than 10% and only in 3.7% of the days did it exceed 50%. Attempts to extract an atmospheric proxy which identifies a WRST did not yield significant separation between dry and wet RSTs, because of significant differences among several WRST types. Therefore, identification of future changes in WRSTs will still require the use of rainfall data offered by climate models. We automatically classified the WRST days into four types. Two types are associated with 850 hPa negative temperature anomalies and the other two with higher‐than‐normal temperatures, resembling ‘tropical‐like’ RSTs, which are known as ‘active RSTs’ and tropical plumes. A predictive equation was extracted to quantify the tropical air intrusion and replace the use of trajectory analysis in future studies.

Quantitative Precipitation Estimation (QPE) Rainfall from Meteorology Radar over Chi Basin

This study of the Quantitative Estimation Precipitation (QEP) of rainfall, detected by two Meteorology Radars over Chi Basin, North-east Thailand, used data from the Thai Meteorological Department (TMD). The rainfall data from 129 rain gauge stations in the Chi Basin area, covering a period of two years, was also used. The study methodology consists of: firstly, deriving the QPE between radar and rainfall based on meteorological observations using the Marshall Palmer Stratiform, the Summer Deep Convection, and Regression Model and calibrating with rain gauge station data; secondly, Bias Correction using statistical method; thirdly, determining spatial variation using three methods, namely Kriging, Inverse Distance Weight (IDW), and the Minimum Curvature Method. The results of the study demonstrated the accuracy of estimating precipitation using meteorological radar. Estimated precipitation compared against an equivalent of 2 years of rain station measurement had a probability of detection (POD) of 0.927, where a value of 1 indicated perfect agreement, demonstrating the effectiveness of the method used to calibrate the radar data. The bias correction method gave high accuracy compared with measured rainfall. Furthermore, of the spatial estimation of rainfall methods, the Kriging methodology showed the best fit between estimation of rainfall distribution and measured rainfall distribution. Therefore, the results of this study showed that the rainfall estimation, using data from a meteorology radar, has good accuracy and can be useful, especially in areas where it is not possible to install and operate rainfall measurement stations, such as in heavily forested areas and/or in steep terrain. Additionally, good accuracy rainfall data derived from radar data can be integrated with other data used for water management and natural disasters for applications to reduce economic losses, as well as losses of life and property.

Spatial Frequency Analysis by Adopting Regional Analysis with Radar Rainfall in Taiwan

This study proposed a spatially and temporally improving methodology adopting the Regional Frequency Analysis with an L-moments approach to estimate rainfall quantiles from 22,787 grids of radar rainfall in Taiwan for a 24-h duration. Due to limited radar coverage in the eastern region, significant discordant grids were found in the coastal area of the eastern region. A total of 171 grids with Di > 6 were set as discordant grids and removed for further analysis. A K-means cluster analysis using scaled at-site characteristics was used to group the QPESUMS grids in Taiwan into 22 clusters/sub-regions based on their characteristics. Spatially, homogeneous subregions with QPESUMS data produce more detailed homogeneous subregions with clear and continuous boundaries, especially in the mountain range area where the number of rain stations is still very limited. According to the results of z-values and L-moment ratio diagrams, the Wakeby (WAK), Generalized Extreme Value (GEV), and Generalized Pareto (GPA) distributions of rainfall extremes fitted well for the majority of subregions. The Wakeby distribution was the dominant best-fitted distribution, especially in the central and eastern regions. The east of the northern part and southern part of Taiwan had the highest extreme rainfall especially for a 100-year return period with an extreme value of more than 1200 mm/day. Both areas were frequently struck by typhoons. By using grid-based (at-site) as the basis for assessing regional frequency analysis, the results show that the regional approach in determining extreme rainfall is very suitable for large-scale applications and even better for smaller scales such as watershed areas. The spatial investigation was performed by establishing regions of interest in small subregions across the northern part. It showed that regionalization was correct and consistent.