دانلود ترجمه مقاله تجزیه و تحلیل عدم قطعیت و حساسیت تخمین خسارت ناشی از سیلاب ساحلی – مجله Doaj

فهرست مطالب:

۱ مقدمه
۲منطقه مورد مطالعه
۳ داده ها و مدل ها
۴ اجرای الگوریتم مونته کارلو
۴ ۱ تجزیه تحلیل های عدم قطعیت و حساسیت
۴ ۲ دامنه عدم قطعیت در پارامترهای ورودی
۴ ۲ ۱ سطح آب
۴ ۲ ۲ رشد خط طغیان
۴ ۲ ۳ براورد خسارت
۵ نتایج و بحث
۶ نتیجه گیری

بخشی از ترجمه:

در این مطالعه، آنالیزهای عدم قطعیت و حساسیت روی براوردهای خسارت سیلاب مربوط به طوفان های ساحلی در غرب هلند انجام شدند. برای تسهیل این موضوع، یک رویکر د جدید، برای محاسبه عمق طغیان که در شبیه سازی های مونته کارلو واقع می شود ایجاد و به طور موفقیت امیزی اجرا شد.مشابه با مدل سریع طغیان ایجاد شده توسط وارد و همکاران ۲۰۱۱، هدف این رویکرد جدید مدل سازی اجرای روش های هیدرودینامیک دو بعدی برای پاسخ به سوالاتی است که مستلزم طیف وسیعی از شبیه سازی های مدل هستند.رویکرد کارامد مدل سازی محاسباتی جدید اعماق طغیان سیل را به طور کارامدی براورد کرده و امکان انجام انالیزهای حساسیت و عدم قطعیت را با استفاده از شبیه سازی های مونته کارلو در اختیار می گذارد. مجموع انالیزهای عدم قطعیت و حساسیت اطلاعات بیشتری در خوص پارامترهایی در اختیار می گذارند که برای عدم قطعیت در خروجی زنجیره مدل سازی بسیار مهم هستند.
ما نشان میدهیم که در رابطه با براورد های خسارت سیل مربوط به بارش های ساحلی در غرب هلند عدم قطعیت قابل توجهی وجود دارد. براورد های کران بالا و پایین با حدود اطمینان ۹۵% ۴ برابر کوچکر و بزرگتر از میانه هستند.موثرترین پارامتر دخیل در این عدم قطعیت، عدم قطعیت مربوط به شکل منحنی های خسارت-سیلاب است. به طور کلی، سهم عدم قطعیت در پارامترهای مربوط به محاسبه خسارت معادل سهم پارامترهای مربوط به حجم آب ورودی است(ارتفاع بارش، زمان بارش و مصاالح به کار رفته در ساخت سیل گیر). این نشان می دهد که توجه به ارزیابی خطر سیل و پژوهش های مربوطه بایستی بین این دو به تعادل برسند.

بخشی از مقاله انگلیسی:

Introduction Flood management in Europe has traditionally mainly been concerned with measures designed to prevent flood events. More recently, a paradigm shift can be observed in flood risk management, moving towards an approach not only focussed on flood prevention (i.e. the hazard), but also on measures that reduce economical damage and casualties (DKKV, 2004; Tunstall et al., 2004; Vis et al., 2003). The interest in a risk-based approach has been amplified by severe flooding events in recent decades, as well as the observation that exposure to flooding has increased dramatically through population growth and urban development in vulnerable areas (Munich Re, 2005; Nicholls et al., 2008; De Moel et al., 2011). The prospect of future flood risk is continuing to increase because of these socio-economic drivers combined with the prospect of climate change (e.g. Te Linde et al., 2010; Milly et al., 2002) added further to this interest in risk-based flood management. When focusing on flood consequences to support riskbased flood management, different models have been designed to assess direct economic damages to, for instance, buildings and infrastructure (Penning-Rowsell et al., 2010; Thieken et al., 2008; Kok et al., 2005). However, research shows there is quite some uncertainty in such flood damage simulations, which is rooted in the uncertainty of the underlying data and model assumptions (Merz et al., 2004; Merz and Thieken, 2009; Apel et al., 2008; De Moel and Aerts, 2011). These sources of uncertainty can relate to both epistemic uncertainty (incomplete knowledge), or aleatory uncertainty (natural variability) (see e.g. Apel et al., 2004). Research suggests to improve our understanding of the uncertainties in flood damage modelling, in order to make the Published by Copernicus Publications on behalf of the European Geosciences Union. 1046 H. de Moel et al.: Uncertainty and sensitivity analysis of coastal flood damage estimates results defensible and reliable (Saltelli et al., 1998), build public confidence in the output (Brugnach et al., 2007), communicate implications of limitations in data and scale (Hall and Solomatine, 2008), and ultimately provide the best possible basis for effective decision making (Ascough et al., 2008; Downton et al., 2005). Moreover, studying uncertainties forces researchers to scrutinize their own models (Merz et al., 2008), improving understanding of the system and the interpretation of the results. Uncertainties can be addressed in several ways, for instance by using scenarios to create a range of possible outcomes. This is often done in studies concerning assessments of future situations which are inherently uncertain (Hall et al., 2005a; Aerts et al., 2008; Klijn et al., 2007; Bouwer et al., 2010), but can also be done to explore current uncertainties using, for instance, different inundation scenarios (Klijn et al., 2007; RWS-DWW, 2005; Van der Most and Wehrung, 2005). There are various studies that do not use scenarios but rather look at uncertainty in various components of a damage assessment. Some studies looked at uncertainties in a specific component, like the direct damage calculation part of the assessment (Merz et al., 2004; Egorova et al., 2008) or the estimation of inundation depths (e.g. Hall et al., 2005b; Noack and Yoruk, 2008). More recently, studies have focused on the combined uncertainty in various components of flood damage assessments. Merz and Thieken (2009), for instance, used different flood frequency curves, inundation models and damage models to estimate uncertainty bounds around flood risk curves. A similar approach was used by De Moel and Aerts (2011) who explored uncertainties in damage models, inundation depth and land use. A more thorough analysis of uncertainty would require a statistical analysis of uncertainty in individual parameters, which are used as input for the various models and how they propagate through the entire modelling chain (Heuvelink, 1998). This is usually done using a Monte Carlo (MC) framework in which individual parameter values are (quasi-)randomly sampled and then evaluated in the model(s). This requires a large number (hundreds) of model evaluations. In flood damage assessments this is mainly hampered by the estimation of inundation depths as the models used for this (2-dimensional hydrodynamic flow models) are computationally very demanding (Gouldby and Kingston, 2007; Apel et al., 2008). The study by Apel et al. (2008) is a good example of applying MC to estimate uncertainty in flood risk modelling. To overcome the computational burden of 2-dimensional hydrodynamic models, they ran such models a priori for various dike breach scenarios to create lookup tables for inundation characteristics, which were subsequently used in a MC uncertainty analysis. They varied many different parameters in their study, though most were associated with the hazard part, and not so much the damage estimation. The damage estimation was performed after the MC analysis using three different depth-damage curves for buildings. In addition to analysing the uncertainty in the output of a model or modelling chain, calculating a large number of model evaluations also makes it possible to perform sensitivity analyses. Sensitivity analyses enable distinguishing which uncertainty sources impact the output the most. Such insights allow identifying parameters, which are justified to keep constant, and can be used to direct research efforts to address parameters that heavily impact the output in order to make a more robust assessments.

دانلود رایگان مقاله انگلیسی

خرید ترجمه فارسی مقاله با فرمت ورد

خرید نسخه پاورپوینت این مقاله جهت ارائه