Subseasonal-to-seasonal (S2S) prediction, especially the prediction of extreme hydroclimate events such as droughts and floods, is not only scientifically challenging, but also has substantial societal impacts. Motivated by preliminary studies, the Global Energy and Water Exchanges (GEWEX)/Global Atmospheric System Study (GASS) has launched a new initiative called “Impact of Initialized Land Surface Temperature and Snowpack on Subseasonal to Seasonal Prediction” (LS4P) as the first international grass-roots effort to introduce spring land surface temperature (LST)/subsurface temperature (SUBT) anomalies over high mountain areas as a crucial factor that can lead to significant improvement in precipitation prediction through the remote effects of land–atmosphere interactions. LS4P focuses on process understanding and predictability, and hence it is different from, and complements, other international projects that focus on the operational S2S prediction. More than 40 groups worldwide have participated in this effort, including 21 Earth system models, 9 regional climate models, and 7 data groups. This paper provides an overview of the history and objectives of LS4P, provides the first-phase experimental protocol (LS4P-I) which focuses on the remote effect of the Tibetan Plateau, discusses the LST/SUBT initialization, and presents the preliminary results. Multi-model ensemble experiments and analyses of observational data have revealed that the hydroclimatic effect of the spring LST on the Tibetan Plateau is not limited to the Yangtze River basin but may have a significant large-scale impact on summer precipitation beyond East Asia and its S2S prediction. Preliminary studies and analysis have also shown that LS4P models are unable to preserve the initialized LST anomalies in producing the observed anomalies largely for two main reasons: (i) inadequacies in the land models arising from total soil depths which are too shallow and the use of simplified parameterizations, which both tend to limit the soil memory; (ii) reanalysis data, which are used for initial conditions, have large discrepancies from the observed mean state and anomalies of LST over the Tibetan Plateau. Innovative approaches have been developed to largely overcome these problems.

This study presents the dependency of the simulation results from a global atmospheric numerical model on machines with different hardware and software systems. The global model program (GMP) of the Global/Regional Integrated Model system (GRIMs) is tested on 10 different computer systems having different central processing unit (CPU) architectures or compilers. There exist differences in the results for different compilers, parallel libraries, and optimization levels, primarily a result of the treatment of rounding errors by the different software systems. The system dependency, which is the standard deviation of the 500-hPa geopotential height averaged over the globe, increases with time. However, its fractional tendency, which is the change of the standard deviation relative to the value itself, remains nearly zero with time. In a seasonal prediction framework, the ensemble spread due to the differences in software system is comparable to the ensemble spread due to the differences in initial conditions that is used for the traditional ensemble forecasting.

General Circulation Models (GCMs) have for decades exhibited difficulties in modelling the diurnal cycle of precipitation (DCP). This issue can be related to inappropriate representation of the processes controlling sub‐diurnal phenomena like convection. In this study, 11 single‐column versions of GCMs are used to investigate the interactions between convection and environmental conditions, processes that control nocturnal convections, and the transition from shallow to deep convection on a diurnal time‐scale. Long‐term simulations are performed over two continental land sites: the Southern Great Plains (SGP) in the USA for 12 summer months from 2004 to 2015 and the Manacapuru site at the central Amazon (MAO) in Brazil for two full years from 2014 to 2015. The analysis is done on two regimes: afternoon convective regime and nocturnal precipitation regime. Most models produce afternoon precipitation too early, likely due to the missing transition of shallow‐to‐deep convection in these models. At SGP, the unified convection schemes better simulate the onset time of precipitation. At MAO, models produce the heating peak in a much lower level compared with observation, indicating too shallow convection in the models. For nocturnal precipitation, models that produce most of nocturnal precipitation all allow convection to be triggered above the boundary layer. This indicates the importance of model capability to detect elevated convection for simulating nocturnal precipitation. Sensitivity studies indicate that (a) nudging environmental variables towards observations has a minor impact on DCP, (b) unified treatment of shallow and deep convection and the capability to capture mid‐level convection can help models better capture DCP, and (c) the interactions of the atmosphere with other components in the climate system (e.g. land) are also important for DCP simulations in coupled models. These results provide long‐term statistical insights on which physical processes are essential in climate models to simulate DCP.

The parametrization of orographic drag processes is a major source of circulation uncertainty in models. The COnstraining ORographic Drag Effects (COORDE) project makes a coordinated effort to narrow this uncertainty by bringing together the modeling community to: explore the variety of orographic drag parametrizations employed in current operational models; assess the resolution sensitivity of resolved and parametrized orographic drag across models; and to validate the parametrized orographic drag in low‐resolution simulations using explicitly resolved orographic drag from high‐resolution simulations. Eleven models from eight major modeling centers are used to estimate resolved orographic drag from high‐resolution (km‐scale) simulations and parametrized orographic drag from low‐resolution simulations, typically used for seasonal forecasting (∼40 km) and climate projections (∼100 km). In most models, at both seasonal and climate resolutions, the total (resolved plus parametrized) orographic gravity wave drag over land is shown to be underestimated by a considerable amount (up to 50%) over the Northern and Southern Hemisphere and by more than 60% over the Middle East region, with respect to the resolved gravity wave drag estimated from km‐scale simulations. The km‐scale simulations also provide evidence that the parametrized surface stress and the parametrized low‐level orographic drag throughout the troposphere are overestimated in most models over the Middle East region, particularly at climate resolutions. Through this process‐based evaluation, COORDE provides model developers new valuable information on the current representation of orographic drag at seasonal and climate resolutions and the vertical partitioning of orographic low‐level and gravity wave drag.