Search

Accurate flood forecasts provide a critical time for authorities and the public to enact flood response measures and initiate evacuations. Long short-term memory (LSTM) is widely used in flood forecasting to ensure sufficient response time. The lag length of rainfall (LR), the number of hidden layers (NL), and the number of neurons (NN) are key parameters of the LSTM model. However, flood forecasting research seldom explores their optimization via the Genetic Algorithm (GA). This study introduces a novel LSTM-GA model, which integrates LSTM with GA to optimize the LR, NL, and NN for flash flood forecasting. The case study pertains to the water level forecasting of the Wu River in Taiwan. To assess the improvement brought by the proposed model, a standard LSTM model was utilized as a benchmark. This model accurately forecasted floods in the next 1 to 6 h, achieving a Nash–Sutcliffe efficiency coefficient (NSE) score ranging from 0.896 to 0.906. It also exhibited strong flood peak forecast performance. The integration of GA enhanced the LSTM's forecasting accuracy, with NSE scores rising to between 0.917 and 0.931. Notably, a shorter forecast lead time augmented the degree of improvement. In the LSTM model, LR was set as the river's concentration time, and NL represented the water storage function of the watershed. For short lead time forecasting, surface runoff was the dominant factor, leading to smaller optimized values for LR and NL. Conversely, long lead time forecasting needed to consider the impact of subsurface and groundwater runoff, resulting in larger optimized values for LR and NL. In conclusion, the parameters optimized through GA consider the watershed's characteristics. [ABSTRACT FROM AUTHOR]

Late-onset asthma (LOA) differs from early-onset asthma (EOA) in terms of prognosis and the treatment response because it has a much worse prognosis and a poorer response to standard asthma treatment. This study sought to investigate the characteristics and clinical outcomes of asthma patients with phenotypes distinguished by age at onset and atopy status. We prospectively recruited patients with asthma who were registered in a pay-for-performance program operated by Taiwan's National Health Insurance Administration (NHIA). These patients received regular outpatient treatment for at least 1 year at every outpatient clinic visit since 2019. Baseline characteristics and clinical outcomes were compared between patients with LOA (≥40 years) and those with EOA (<40 years). Of the consecutive 101 patients with asthma, 21 patients (20.7%) had EOA and 80 (79.3%) had LOA. In the 12-month period, patients with EOA had higher declines in forced expiratory volume in one second (FEV1; −2.1 ± 8.4 vs. 6.8 ± 13.1, % of predicted value, p = 0.037) and forced vital capacity (FVC; −4.6 ± 12.0 vs. 6.1 ± 13.6, % of predicted value, p = 0.023) than patients with LOA. Patients with nonatopic EOA had a significantly higher exacerbation rate at 12 months than patients with nonatopic LOA (50% vs. 11.8%, p = 0.012). Identification of different phenotypes of asthma is important in clinical practice because treatment responses may differ. [ABSTRACT FROM AUTHOR]

This study aimed to assess the effective spatial characteristics of input features by using physics-informed, machine learning (ML)-based inundation forecasting models. To achieve this aim, inundation depth data were simulated using a numerical hydrodynamic model to obtain training and testing data for these ML-based models. Effective spatial information was identified using a back-propagation neural network, an adaptive neuro-fuzzy inference system, support vector machine, and a hybrid model combining support vector machine and a multi-objective genetic algorithm. The conventional average rainfall determined using the Thiessen polygon method, raingauge observations, radar-based rainfall data, and typhoon characteristics were used as the inputs of the aforementioned ML models. These models were applied in inundation forecasting for Yilan County, Taiwan, and the hybrid model had the best forecasting performance. The results show that the hybrid model with crucial features and appropriate lag lengths gave the best performance. [ABSTRACT FROM AUTHOR]

Accurate hourly real-time flood forecasting is necessary for early flood warning systems, especially during typhoon periods. Artificial intelligence methods have been increasingly used for real-time flood forecasting. This study developed a real-time flood forecasting model by using back-propagation networks (BPNs) with a self-organizing map (SOM) to create ensemble forecasts. Random weights and biases were set for the BPNs to learn the characteristics of a catchment system. An unsupervised SOM network with a classification function was then used to cluster representative BPN weights and biases; clusters of BPNs with high accuracy were selected to act as experts for the ensemble models to forecast flow rates. The model was applied to flood events in the Wu River Basin of Taiwan. Most observed values were within the forecasting intervals of the BPN clusters in the calibration and validation phases, indicating that the models had acceptable accuracy. For the large flood events of typhoons Saola in the calibration phase and Soulik in the validation phase, the mean average error of the ensemble mean model for the cluster A was 143.1 and 327.4 m3/s, respectively; these values were lower than those for the best individual model within the cluster (194.3 and 917.9 m3/s). The ensemble model thus outperformed the individual models and can accurately forecast flood values and intervals. Therefore, the model can be used to accurately forecast floods. [ABSTRACT FROM AUTHOR]

In response to the impacts of extreme precipitation on human or natural systems under climate change, the development of climate risk assessment approach is a crucial task. In this paper, a novel risk assessing approach based on a climate risk assessment framework with copula-based approaches is proposed. Firstly, extreme precipitation indices (EPIs) and their marginal distributions are estimated for historical and future periods. Next, the joint probability distributions of extreme precipitation are constructed by copula methods and tested by goodness-of-fit indices. The future joint probabilities and joint return periods (JRPs) of the EPIs are then evaluated. Finally, change rates of JRPs for future periods are estimated to assess climate risk with the quantitative data of exposure and vulnerability of a protected target. An actual application in Taiwan Island is successfully conducted for climate risk assessment with the impacts of extreme precipitation. The results indicate that most of regions in Taiwan Island might have higher potential climate risk under different scenarios in the future. The future joint probabilities of precipitation extremes might cause the high risk of landslide and flood disasters in the mountainous area, and of inundation in the plain area. In sum, the proposed climate risk assessing approach is expected to be useful for assisting decision makers to draft adaptation strategies and face high risk of the possible occurrence of natural disasters. [ABSTRACT FROM AUTHOR]

INTRODUCTION: Pharmaceutical expenditure (PE) of the National Health Insurance (NHI) programme in Taiwan grew from 62.2 billion Taiwan new dollars ($NT) in 1996 to $NT94.5 billion in 2003.The government has been introducing many strategies to control PE since the inception of NHI including price adjustment based on the prices of international products or existing products (inter-brands comparison), or market price and volume survey; delegation of financial responsibility to regional bureaux; co-payment for outpatient drugs; generic grouping (the reference pricing scheme based on chemical equivalence); a global budget payment system for clinics and hospitals; and reduction in the flat daily payment rate of the drugs for clinics. The aim of this study was to evaluate the impact of these cost containment strategies on the PE of the NHI programme from 1996 to 2003. METHODS: To take the growth and seasonal trends of monthly PE into consideration, Box and Tiao’s time-series event intervention analysis based on the Box-Jenkins auto-regressive integrated moving-average model was applied to evaluate the impact of various cost containment strategies on total and subsector (outpatient, inpatient, clinic and hospital sectors) PE. Monthly data of PE of the NHI programme from 1996 to 2003 (the dependent variables) were obtained from the Bureau of the NHI. Drugs prescribed by dentists and Chinese medical doctors at outpatient departments were excluded. RESULTS: After fitting the patterns of time series and controlling for the calendar effect of the Chinese New Year and the severe acute respiratory syndrome outbreak in 2003, three strategies (generic grouping, delegation of financial responsibility and reduction of the flat payment rate of clinics) were significantly associated with a reduction in PE. However, the hospital global budget strategy offset partial savings from these three strategies. Cumulative savings during the study period were estimated to be $NT25.442 billion ($US0.80 billion). Of all the strategies, generic grouping was the most effective although it had less effect on the clinic subsector. Neither drug co-payment nor price adjustment based on the international or inter-brand price comparison had significant impacts on PE. CONCLUSION: Generic grouping, reduction of the flat payment rate and delegation of financial responsibility were effective in controlling PE. A global budget alone would be unable to control PE without other direct financial incentives. Neither drug co-payment nor brand-specific price adjustment based on prices of international/existing products had a significant impact on PE. [ABSTRACT FROM AUTHOR]