China 
Africa 
Explore chapters and articles related to this topic
Modeling and Validation Challenges for Complex Systems
Published in Larry B. Rainey, Mo Jamshidi, Engineering Emergence, 2018
Ensemble forecasting. The core idea of ensemble forecasting is to execute multiple runs of a model, each of which was initialized with slightly different initial states, and then develop a prediction based on the multiple results.9 The differences in the inputs are intended to reflect the uncertainty in the knowledge of the initial state. The multiple results may be aggregated or averaged, and the variation and divergences between them analyzed; the details of aggregation and analysis depend on the application, but statistical methods are often employed. In some forms and contexts this is a familiar idea; modelers using a discrete event simulation to study a queueing system often conduct multiple trials, each beginning with a different random number seed (Law, 2015). In the case of weather models, different values for the initial conditions of the atmosphere may be used, with the differences generated based on the noise or uncertainty in the observations upon which the input is based (Smith, 2007). The uncertainty of the forecast may be estimated based the variation in the different forecasts generated.
Uncertainty
Published in Andrew Cook, Damián Rivas, Complexity Science in Air Traffic Management, 2016
Finally, a source of uncertainty that affects all scales is the modelling of weather. Given that deterministic weather forecasts are not accurate, probabilistic models are necessary. One current trend is to use ensemble prediction forecasts, which attempt to characterise and quantify the inherent prediction uncertainty based on ensemble modelling; see, for instance, Hacker et al., 2003. Ensemble forecasting is a prediction technique that aims to generate a representative sample of the possible future states of the atmosphere. An ensemble forecast is a collection of typically 10 to 50 weather forecasts which may be obtained in different ways based on time-lagged, multi-model, and/or multi-initial conditions approaches (Arribas et al., 2005; Lu et al., 2007). Different national meteorological offices provide ensemble prediction forecasts, which can be used to model weather uncertainty, be it at the flight, traffic, or network scale.
A Brief review of flood forecasting techniques and their applications
Published in International Journal of River Basin Management, 2018
Sharad Kumar Jain, Pankaj Mani, Sanjay K. Jain, Pavithra Prakash, Vijay P. Singh, Desiree Tullos, Sanjay Kumar, S. P. Agarwal, A. P. Dimri
The concept of ensemble forecasting originated in the atmospheric community to overcome the limitations associated with the deterministic models. In ensemble prediction systems (EPS), a set of possible future states of the variable are provided through small changes in the initial conditions, different representations of the physical processes, and changes in parameterization schemes and solution schemes. Rather than providing a single deterministic forecast, the EPS offers an ensemble prediction of hydrological variables, such as streamflow or river level, allowing the identification of the most likely scenario. An EPS, in a way, consists of the propagation of uncertainties through the forecasting system. Notwithstanding the other uncertainties, prediction of rainfall is often the dominant source of uncertainty in FF.
Selecting components in a probabilistic hydrological forecasting chain: the benefits of an integrated evaluation
Published in LHB, 2021
Joseph Bellier, Guillaume Bontron, Isabella Zin
Historically, the meteorological community has paved the way for probabilistic forecasting with ensemble forecasting, which consists in running a numerical weather prediction (NWP) model multiple times with slightly perturbed atmosphere initial conditions, and sometimes different modelling assumptions (Bauer et al., 2015). More generally, we refer in this paper to the ensemble approach as any method that, for a given physical process to be modelled, constructs an ensemble of simulations whose dispersion represents the uncertainty of that process. The introduction of the ensemble approach into hydrological forecasting has most often consisted in propagating meteorological ensemble forecasts, issued by a given operational centre, into a hydrological model (see Cloke & Pappenberger, 2009 for a review). But other practices exist, such as using “poor man’s ensembles”, which refers to picking multiple deterministic forecasts from independent meteorological centres (Ebert, 2001), or “grand ensembles”, which merge together ensembles issued by different meteorological centres (He et al., 2009; Pappenberger et al., 2008). The objective is to benefit from the differences between the NWP models (e.g. in resolution, physical parametrisation, assimilation or perturbation methods) to widen the range of possible meteorological scenarios. Yet a meteorological ensemble forcing only exhibits the uncertainty associated with the atmospheric processes. To encompass uncertainties linked to hydrological processes and to their modelling, a common practice consists in using multiple hydrological models instead of a single one, an approach referred to as “multi-model”. Most often, several models with different structures and modelling assumptions are selected (Thiboult et al., 2016; Velázquez et al., 2011), although one could also consider different parameter sets for the same model (e.g. the generalized likelihood uncertainty estimation [GLUE] method, Beven & Freer, 2001). One may remark that the multi-model approach follows the same principle as the “poor man’s ensemble” (i.e. different deterministic simulations are gathered to form an ensemble), although it has never been named as such, the use of physically perturbed members in hydrological modelling being still in the early stages.