Figure 6 - uploaded by Sam Hatfield
Content may be subject to copyright.
7: The continuous ranked probability score (CRPS) of the 850 hPa tropical wind fields (averaged from the zonal and meridional winds) for various experimental setups verified with respect to analysis at TCo399 resolution. The CRPS was averaged over 12 forecasts and there were 11 ensemble members for each forecast.
Source publication
The skill of weather forecasts has improved dramatically over the past 30 years. This improvement has depended to a large degree on developments in supercomputing, which have allowed models to increase in complexity and resolution with minimal technical effort. However, the nature of supercomputing is undergoing a significant change, with the adven...
Similar publications
Non-predictive or inaccurate weather forecasting can severely impact the community of users such as farmers. Numerical weather prediction models run in major weather forecasting centers with several supercomputers to solve simultaneous complex nonlinear mathematical equations. Such models provide the medium-range weather forecasts, i.e., every 6 h...
Citations
... The key to 4D-Var's success is in using the tangent linear and adjoint models to propagate in time the background error estimates (the errors of the prior information) which are only provided at the start of the assimilation window. This is different from ensemble methods, like EnKF 57 or 4DEnVar 58 , which con- 53 Wave model 54 struct the 4-dimensional prior forecast covariance using the ensemble of non-linear forecast trajectories. The latter are more prone to sampling errors and require spatial localisation in the case of EnKF or both spatial and temporal localisation in the case of 4DEnVar/4D-EnKF, unless they employ large ensembles, for example, at lower resolution to generate sufficient sampling. ...
... However, it may be possible to accelerate expensive model components or computing kernels. [53] could, for example, show that the TensorCore accelerator on Volta GPUs of NVIDIA could potentially be used to calculate the Legendre transformations of IFS at half precision with no significant impact on forecast results. A reduction of numerical precision below single precision may also be possible for some of the parametrization schemes [94]. ...
The efficiency of the forecasting system on future high-performance computing and data handling systems is considered one of the key challenges for implementing ECMWF’s ambitious strategy. This was already recognised by ECMWF in 2013, and has led to the foundation of the Scalability Programme. The programme aims to address this challenge as a concerted action between the Centre and its Member States, but also draws in the computational science expertise available throughout Europe. This technical memorandum provides an overview of the status of the programme, highlights achievements from the first five years ranging from observational data pre-processing, data assimilation, forecast model design and output data post-processing, and defines the roadmap for the next five years towards a sustainable system that can operate on the expected range of hardware and software technologies. This point in time is crucial because the programme will have a strong focus on implementation and operational benefit in the next period.
