We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Twelve Times Faster yet Accurate: A New State‐Of‐The‐Art in Radiation Schemes via Performance and Spectral Optimization.
- Authors
Ukkonen, Peter; Hogan, Robin J.
- Abstract
Radiation schemes are critical components of Earth system models that need to be both efficient and accurate. Despite the use of approximations such as 1D radiative transfer, radiation can account for a large share of the runtime of expensive climate simulations. Here we seek a new state‐of‐the‐art in speed and accuracy by combining code optimization with improved algorithms. To fully benefit from new spectrally reduced gas optics schemes, we restructure code to avoid short vectorized loops where possible by collapsing the spectral and vertical dimensions. Our main focus is the ecRad radiation scheme, where this requires batching of adjacent cloudy layers, trading some simplicity for improved vectorization and instruction‐level parallelism. When combined with common optimization techniques for serial code and porting widely used two‐stream kernels fully to single precision, we find that ecRad with the TripleClouds solver becomes 12 times faster than the operational radiation scheme in ECMWF's Integrated Forecast System (IFS) cycle 47r3, which uses a less accurate gas optics model (RRMTG) and a more noisy solver (McICA). After applying the spectral reduction and extensive optimizations to the more sophisticated SPARTACUS solver, we find that it's 2.5 times faster than IFS cy47r3 radiation, making cloud 3D radiative effects affordable to compute in large‐scale models. The code optimization itself gave a threefold speedup for both solvers. While SPARTACUS is still under development, preliminary experiments show slightly improved medium‐range forecasts of 2‐m temperature in the tropics, and in year‐long coupled atmosphere‐ocean simulations the 3D effects warm the surface substantially. Plain Language Summary: A crucial step in simulating weather and climate is calculating how atmospheric radiation (shortwave radiation from the sun and terrestrial longwave radiation) interacts with the Earth's atmosphere and surface. The complexity of the underlying physics has necessitated making approximations in how radiative transfer is treated, such as assuming that radiation can only enter or leave a cloud through its top or base, thereby ignoring 3D effects. Even so, radiative transfer has historically been one of the computationally most demanding steps in making weather and climate simulations. Here we show that a state‐of‐the‐art radiation code can be sped up threefold by using code optimization techniques that seek to maximize performance on modern processors. Combining this with a recent innovation that reduces the number of spectral computations required for accurate solutions, an order‐of‐magnitude increase in speed is obtained compared to the existing radiation scheme in a global weather model. Crucially, these improvements also make a radiation scheme that accounts for cloud 3D radiative effects fast enough to be used operationally. When included in global simulations, these 3D effects act to warm the lower atmosphere substantially. Key Points: Restructuring and serial code optimization is performed for the ECMWF radiation schemeWhen combined with spectrally reduced gas optics, a speed‐up of 12 is obtained compared to the operational schemeCloud 3D radiative effects become affordable to compute in dynamical models
- Subjects
ATMOSPHERIC radiation; GLOBAL radiation; TERRESTRIAL radiation; RADIATIVE transfer; RADIATION; RADIATIVE transfer equation; CLIMATE sensitivity
- Publication
Journal of Advances in Modeling Earth Systems, 2024, Vol 16, Issue 1, p1
- ISSN
1942-2466
- Publication type
Article
- DOI
10.1029/2023MS003932