Outcomes and challenges of global high-resolution non-hydrostatic atmospheric simulations using the K computer

Abstract

This article reviews the major outcomes of a 5-year (2011-2016) project using the K computer to perform global numerical atmospheric simulations based on the non-hydrostatic icosahedral atmospheric model (NICAM). The K computer was made available to the public in September 2012 and was used as a primary resource for Japan's Strategic Programs for Innovative Research (SPIRE), an initiative to investigate five strategic research areas; the NICAM project fell under the research area of climate and weather simulation sciences. Combining NICAM with high-performance computing has created new opportunities in three areas of research: (1) higher resolution global simulations that produce more realistic representations of convective systems, (2) multi-member ensemble simulations that are able to perform extended-range forecasts 10-30 days in advance, and (3) multi-decadal simulations for climatology and variability. Before the K computer era, NICAM was used to demonstrate realistic simulations of intra-seasonal oscillations including the Madden-Julian oscillation (MJO), merely as a case study approach. Thanks to the big leap in computational performance of the K computer, we could greatly increase the number of cases of MJO events for numerical simulations, in addition to integrating time and horizontal resolution. We conclude that the high-resolution global non-hydrostatic model, as used in this five-year project, improves the ability to forecast intra-seasonal oscillations and associated tropical cyclogenesis compared with that of the relatively coarser operational models currently in use. The impacts of the sub-kilometer resolution simulation and the multi-decadal simulations using NICAM are also reviewed.

The coupling of deep convection with the resolved flow via the divergence of mass flux in the IFS

“…However, global convection‐permitting models (e.g., Satoh et al . ; Prein et al . ), and regional models (e.g., Marsham et al .…”

“…For similar reasons, other global models (Fowler et al, 2017) also employ some degree of convective stabilization down to 3 km resolutions. However, global convection-permitting models (e.g., Satoh et al, 2017;Prein et al, 2017), and regional models (e.g., Marsham et al, 2013;Holloway et al, 2016) have shown clear improvements in the organization and propagation of mesoscale convective systems (MCSs) compared to models with parametrized convection, likely due to better coupling of the convective motions and microphysics with the large-scale flow (effect of cold pools, wind shear, memory and grid-scale motions being in phase with the condensational heating). These studies also suggested that the largest model sensitivity in the (5 km) resolution range is due to the use (or not) of a deep convection parametrization.…”

Constraining stochastic parametrisation schemes using high‐resolution simulations

“…Instead of requiring the user to produce a pair of low‐ and high‐resolution simulations simultaneously, it makes use of existing high‐resolution simulations. This sets a relatively low bar for carrying out such coarse‐graining studies, and makes use of the wealth of high‐resolution simulations available (Satoh et al ., ; ; Schalkwijk et al ., ; Heinze et al ., ). The coarse‐grained dataset is coupled to a forecast model through the use of a SCM framework.…”

“…Recent years have seen a surge in the production of very high‐resolution atmospheric simulations. The continued increase in computational power has led to an increase in domain size and duration of simulations, with resolutions regularly reaching convection‐permitting, if not convection‐resolving, scales (Holloway et al ., ; Satoh et al ., ; ; Schalkwijk et al ., ; Heinze et al ., ; Stevens et al ., ). The availability of such datasets opens up the option of using high‐resolution simulations as a proxy for the “true atmosphere” and identifying the difference between a low‐resolution forecast model and a high‐resolution simulation as the model error that a stochastic parametrisation seeks to represent (Shutts and Palmer, ; Shutts and Pallares, ).…”

Forcing Single‐Column Models Using High‐Resolution Model Simulations

“…For example, the Cascade project produced high-resolution convection-permitting simulations using the UK Met Office Unified Model (MetUM) at 1.5-km and 4-km resolution over a domain covering most of West Africa, and a domain spanning 15,500 km by 4,500 km over the Indian Ocean, Warm Pool and tropical west Pacific (Holloway et al, 2013;Marsham et al, 2011), while the High-Definition Clouds and Precipitation for advancing Climate Prediction [HD(CP) 2 ] project used the ICOsahedral Nonhydrostatic atmosphere model for a LES over Germany at 156-m resolution for 4 days (Heinze et al, 2017). The increase in production of extensive high-resolution data sets (Heinze et al, 2017;Holloway et al, 2012;Satoh et al, 2014Satoh et al, , 2017Schalkwijk et al, 2015) has largely become possible due to computational advances in recent years.…”