Jonathan Flowerdew

Ensemble Forecasting of Storm Surges

The overtopping of flood defenses by coastal storm surges constitutes a significant threat to life and property. Like all forecasts, storm surge predictions have an associated uncertainty, but this is not directly predicted by current operational systems. The dominant source of this uncertainty is thought to be uncertainty in the driving atmospheric forecast of conditions at the sea surface, which can vary substantially depending on the meteorological situation. Ensemble prediction is a technique used to assess uncertainty in forecasts of complex nonlinear systems such as weather, where small errors can quickly grow to produce significantly different outcomes. It works by running not one but several forecasts, using slightly different initial conditions, boundary conditions, and/or model physics. These are chosen to sample the range of uncertainty in model inputs and formulation so that the corresponding forecasts will sample the range of possible results that are consistent with those uncertainties. The United Kingdom Met Office has recently developed the Met Office Global and Regional Ensemble Prediction System (MOGREPS), which provides 24 different predictions of meteorological evolution over a North Atlantic and European domain with a 24 km grid length. The aim of the present project is to run a barotropic storm surge prediction for each MOGREPS ensemble member, and thereby estimate the risk of damaging events given the forecast uncertainties which are sampled by the ensemble. The system forecasts 54 hours ahead and runs twice per day. In most situations, the ensemble develops rather little spread, suggesting a fairly predictable situation and a high degree of confidence in the forecast. On some occasions, however, the spread is much larger, suggesting a greater degree of uncertainty. Initial verification results are encouraging, although statistical evaluation suggests the ensemble spread is generally too small.

Calibrating ensemble reliability whilst preserving spatial structure

Ensemble forecasts aim to improve decision-making by predicting a set of possible outcomes. Ideally, these would provide probabilities which are both sharp and reliable. In practice, the models, data assimilation and ensemble perturbation systems are all imperfect, leading to deficiencies in the predicted probabilities. This paper presents an ensemble post-processing scheme which directly targets local reliability, calibrating both climatology and ensemble dispersion in one coherent operation. It makes minimal assumptions about the underlying statistical distributions, aiming to extract as much information as possible from the original dynamic forecasts and support statistically awkward variables such as precipitation. The output is a set of ensemble members preserving the spatial, temporal and inter-variable structure from the raw forecasts, which should be beneficial to downstream applications such as hydrological models. The calibration is tested on three leading 15-d ensemble systems, and their aggregation into a simple multimodel ensemble. Results are presented for 12 h, 1° scale over Europe for a range of surface variables, including precipitation. The scheme is very effective at removing unreliability from the raw forecasts, whilst generally preserving or improving statistical resolution. In most cases, these benefits extend to the rarest events at each location within the 2-yr verification period. The reliability and resolution are generally equivalent or superior to those achieved using a Local Quantile-Quantile Transform, an established calibration method which generalises bias correction. The value of preserving spatial structure is demonstrated by the fact that 3×3 averages derived from grid-scale precipitation calibration perform almost as well as direct calibration at 3×3 scale, and much better than a similar test neglecting the spatial relationships. Some remaining issues are discussed regarding the finite size of the output ensemble, variables such as sea-level pressure which are very reliable to start with, and the best way to handle derived variables such as dewpoint depression.

Use of field measurements to improve probabilistic wave overtopping forecasts

The paper investigates the practicalities and potential value of coupling offshore to nearshorewave models, and nearshore-wave and surge models to coastline overtopping models and floodrisk indicators. The available coupled models include Ensemble surge models, offshore wave models, hydrodynamic wave transformation models and empirical wave overtopping formulae. The uncertainty associated with an individual model output, and the propagation of this incertainty forward through the modelling chain is handled through Monte Carlo simulation from either discrete or continuous probability distributions, based on the information available on each of the variables involved. The paper will concentrate on the collection and analysis of measured overtopping discharges at a site in North-west England. At the same site inshore wave measurements will be recorded for selected high tide events. The results from these field measurements will be used to calibrate and refine the Monte Carlo simulations.