Tim Palmer

Abstract:

The need for high‐precision calculations with 64‐bit or 32‐bit floating‐point arithmetic for weather and climate models is questioned. Lower‐precision numbers can accelerate simulations and are increasingly supported by modern computing hardware. This paper investigates the potential of 16‐bit arithmetic when applied within a shallow water model that serves as a medium complexity weather or climate application. There are several 16‐bit number formats that can potentially be used (IEEE half precision, BFloat16, posits, integer, and fixed‐point). It is evident that a simple change to 16‐bit arithmetic will not be possible for complex weather and climate applications as it will degrade model results by intolerable rounding errors that cause a stalling of model dynamics or model instabilities. However, if the posit number format is used as an alternative to the standard floating‐point numbers, the model degradation can be significantly reduced. Furthermore, mitigation methods, such as rescaling, reordering, and mixed precision, are available to make model simulations resilient against a precision reduction. If mitigation methods are applied, 16‐bit floating‐point arithmetic can be used successfully within the shallow water model. The results show the potential of 16‐bit formats for at least parts of complex weather and climate models where rounding errors would be entirely masked by initial condition, model, or discretization error.

Abstract:

Recent studies have found evidence of multidecadal variability in northern hemisphere wintertime seasonal forecast skill. Here we assess the robustness of this finding by extending the analysis to analysing a diverse set of ensemble atmospheric model simulations. These simulations differ in either numerical model or type of initialisation and include atmospheric model experiments initialised with reanalysis data and free‐running atmospheric model ensembles. All ensembles are forced with observed SST and seaice boundary conditions. Analysis of large‐scale Northern Hemisphere circulation indicesover the Northern Hemisphere (namely the North Atlantic Oscillation, Pacific North American pattern and the Arctic Oscillation) reveals that in all ensembles there is larger correlation skill in the late century periods than during periods in the mid‐century. Similar multidecadal variability in skill is found in a measure of total skill integrated over the whole of the extratropics. Most of the differences in large‐scale circulation skill between the skillful late period (as well as early period) and the less skillful mid‐century period seem to be due to a reduction in skill over the North Pacific and a disappearance in skill over North America and the North Atlantic. The results are robust across different models and different types of initialisation, indicating that the multidecadal variability in Northern Hemisphere winter skill is a robust feature of 20th century climate variability. Multidecadal variability in skill therefore arises from the evolution of the observed SSTs, likely related to a weakened influence of ENSO on the predictable extratropical circulation signal during the middle of the 20th century, and is evident in the signal‐to‐noise ratio of the different ensembles, particularly the larger ensembles.

Abstract:

We quantify the value of sub‐seasonal forecasts for a real‐world prediction problem: the forecasting of French month‐ahead energy demand. Using surface temperature as a predictor, we construct a trading strategy and assess the financial value of using meteorological forecasts, based on actual energy demand and price data. We show that forecasts with lead times greater than two weeks can have value for this application, both on their own and in conjunction with shorter‐range forecasts, especially during boreal winter. We consider a cost/loss framework based on this example, and show that, while it captures the performance of the short‐range forecasts well, it misses the marginal value present in medium‐range forecasts. We also contrast our assessment of forecast value to that given by traditional skill scores, which we show could be misleading if used in isolation. We emphasise the importance of basing assessment of forecast skill on variables actually used by end‐users.