Search Results

You are looking at 141 - 150 of 5,371 items for :

  • Forecasting techniques x
  • Monthly Weather Review x
  • Refine by Access: Content accessible to me x
Clear All
Eric Vanden-Eijnden and Jonathan Weare

1. Introduction The assimilation of noisy observations into a model to improve its predictive capabilities is a recurring challenge in many applications. Examples include weather prediction and forecasting, robot tracking, stochastic volatility estimation, image analysis, etc. (see Doucet et al. 2001 ). In these applications and many others one is interested in predicting how the system evolves in time given a model for its dynamics and sequentially available, incomplete observations of its

Full access
Dominik Jacques and Isztar Zawadzki

1. Introduction Because of their high spatial and temporal resolution, radar observations have great potential for improving atmospheric analyses and the ensuing forecasts. Despite 30 years ( Lilly 1990 ; Sun et al. 1991 ) of ongoing research, our skills in forecasting mesoscale convection have remained modest. Over continental scales, radar data assimilation was shown to improve forecasts for periods not exceeding 6–8 h ( Berenguer et al. 2012 ; Stratman et al. 2013 ). Over regional scales

Full access
Ja-Rin Park, Hyeong-Bin Cheong, and Hyun-Gyu Kang

1. Introduction In numerical models and data analysis, filtering or smoothing is often used to smooth out unwanted, noiselike, small-scale variations from discrete data ( Wallington 1962 ; Shapiro 1970 ; Orszag 1971 ; Cocke 1998 ; Cocke and LaRow 2000 ; Gelb and Gleeson 2001 ; Denis et al. 2002 ; Cheong et al. 2002 , 2004 ; Feser and von Storch 2005 ; Cheong 2006 ). The numerical technique of filtering is also frequently used for decomposing the discrete grid data into two different

Full access
Kwangjae Sung, Hyo-Jong Song, and In-Hyuk Kwon

i , χ ⁡ ( υ ) , k − 1 | k − 1 i ]   for   i = 0 , … , 2 L a , where χ ⁡ ( x ) , k | k − 1 i is the prior or forecast sigma point state vector at time k , χ ⁡ ( w ) , k − 1 | k − 1 i and χ ⁡ ( υ ) , k − 1 | k − 1 i are sigma point vectors that correspond to the model error and observation error, respectively, and z k i represents the transformed sigma point in observation space. Using the transformed sigma points obtained from Eqs. (12) and (13) , the predicted state estimate

Free access
Paul D. Williams

numerical analysis, this technique falls into the category of implicit–explicit (IMEX) methods. The implicitly treated terms are crucial for high-frequency gravity wave oscillations, which are often of secondary importance and, which, if treated explicitly, would violate the Courant–Friedrichs–Lewy (CFL) stability condition unless impractically short time steps were taken. Semi-implicit numerical schemes are used widely in practical applications, because they suffer from neither the computational

Full access
Timothy DelSole and Xia Feng

using techniques based on analysis of variance. Certain subtle and unverified assumptions that are required for the Shukla–Gutzler method to work, such as the fact that certain quantities follow the chi-square distribution and that the time mean is independent of its residuals, even for autocorrelated data, were clarified. Moreover, Monte Carlo experiments demonstrate that these assumptions are adequate even for autocorrelated data, provided the effective time scale T 0 is known. We derive the

Full access
Xinrong Wu, Wei Li, Guijun Han, Lianxin Zhang, Caixia Shao, Chunjian Sun, and Lili Xuan

) have been developed to address the first issue, a localization technique is usually used to ease the second problem. The original localization scheme (e.g., Houtekamer and Mitchell 1998 ; Anderson 2001 ; Hamill et al. 2001 ; Szunyogh et al. 2008 ) was the fixed localization approach, which is usually realized by a Schur product (an element-by-element multiplication) of the ensemble-evaluated error covariance with an analytic localization operator. All fixed localization models need to determine

Full access
Paul D. Williams

changing the time-step size changes the Kelvin wave speed and the convective and stratiform precipitation. Furthermore, in a simplified atmosphere general circulation model, Amezcua (2012) reports that the sensitivity of the skill of medium-range weather forecasts to the time-stepping method is about the same as it is to the physics parameterizations. A popular time-stepping method is the second-order centered-difference scheme, which is affectionately known as the leapfrog scheme (e.g., Haltiner

Full access
Glen S. Romine, Craig S. Schwartz, Ryan D. Torn, and Morris L. Weisman

model weather prediction with 1–3 days of lead time. They used a variety of techniques to identify source regions of initial condition uncertainty that had the potential to lead to rapid forecast error growth and were suitable for targeted sampling. After observation collection, impact studies (e.g., Baker and Daley 2000 ; Langland and Baker 2004 ) assess changes in initial condition uncertainty and forecast error owing to the assimilation of particular observation sets (e.g., Ancell and Hakim

Full access
Michail Diamantakis and Linus Magnusson

1. Introduction The semi-implicit, semi-Lagrangian (SISL) technique is a well-established method for solving the governing equations of global atmospheric NWP models. It is a very efficient technique as it offers unconditional stability without loss of accuracy (small dispersion errors) permitting integrations with long time steps ( Staniforth and Côté 1991 ). The ECMWF Integrated Forecast System (IFS) model employs a two-time-level SISL scheme ( Temperton et al. 2001 ) combined with a spectral

Full access