The use of machine learning (ML) for the online correction of coarse-resolution atmospheric models has proven effective in reducing biases in near-surface temperature and precipitation rate. However, this often introduces biases in the upper atmosphere and improvements are not always reliable across ML-corrective models trained with different random seeds. Furthermore, ML corrections can feed back on the baseline physics of the atmospheric model and produce profiles that are outside the distribution of samples used in training, leading to low confidence in the predicted corrections. This study introduces the use of a novelty detector to mask the predicted corrections when the atmospheric state is deemed out-of-sample. The novelty detector is trained on profiles of temperature and specific humidity in a semi-supervised fashion using samples from the coarsened reference fine-resolution simulation. Offline, the novelty detector determines more columns to be out-of-sample in simulations which are known, using simple metrics like mean bias, to drift further from the reference simulation. Without novelty detection, corrective ML leads to the development of undesirably large climate biases for some ML random seeds but not others. Novelty detection deems about 21% of columns to be novelties in year-long simulations. The spread in the root mean square error (RMSE) of time-mean spatial patterns of surface temperature and precipitation rate across a random seed ensemble is sharply reduced when using novelty detection. In particular, the random seed with the worst RMSE is improved by up to 60% (depending on the variable) while the best seed maintains its low RMSE.

We present a machine learning based emulator of a microphysics scheme for condensation and precipitation processes (Zhao-Carr) used operationally in a global atmospheric forecast model (FV3GFS). Our tailored emulator architecture achieves high skill (≥94%) in predicting condensate and precipitation amounts and maintains low global-average bias (≤4%) for 1 year of continuous simulation when replacing the Fortran scheme. The stability and success of this emulator stems from key design decisions. By separating the emulation of condensation and precipitation processes, we can better enforce physical priors such as mass conservation and locality of condensation, and the vertical dependence of precipitation falling downward, using specific network architectures. An activity classifier for condensation imitates the discrete-continuous nature of the Fortran microphysics outputs (i.e., tendencies are identically zero where the scheme is inactive, and condensate is zero where clouds are fully evaporated). A temperature-scaled conditional loss function ensures accurate condensate adjustments for a high dynamic range of cloud types (e.g., cold, low-condensate cirrus clouds or warm, condensate-rich clouds). Despite excellent overall performance, the emulator exhibits some deficiencies in the uppermost model levels, leading to biases in the stratosphere. The emulator also has short episodic skill dropouts in isolated grid columns and is computationally slower than the original Fortran scheme. Nonetheless, our challenges and strategies should be applicable to the emulation of other microphysical schemes. More broadly, our work demonstrates that with suitable physically motivated architectural choices, ML techniques can accurately emulate complex human-designed parameterizations of fast physical processes central to weather and climate models.

Accurate precipitation simulations for various climate scenarios are critical for understanding and predicting the impacts of climate change. This study employs a Cycle-generative adversarial network (CycleGAN) to improve global 3-hour-average precipitation fields predicted by a coarse grid (200~km) atmospheric model across a range of climates, morphing them to match their statistical properties with reference fine-grid (25~km) simulations. We evaluate its performance on both the target climates and an independent ramped-SST simulation. The translated precipitation fields remove most of the biases simulated by the coarse-grid model in the mean precipitation climatology, the cumulative distribution function of 3-hourly precipitation, and the diurnal cycle of precipitation over land. These results highlight the potential of CycleGAN as a powerful tool for bias correction in climate change simulations, paving the way for more reliable predictions of precipitation patterns across a wide range of climates.

Aerosol-cloud-precipitation interactions are a leading source of uncertainty in estimating climate sensitivity. Remote marine boundary layers where accumulation mode (~100-400 nm diameter) aerosol concentrations are relatively low are very susceptible to aerosol changes. These regions also experience heightened Aitken mode aerosol (~10-100 nm) concentrations associated with ocean biology. Aitken aerosols may significantly influence cloud properties and evolution by replenishing cloud condensation nuclei and droplet number lost through precipitation (i.e., Aitken buffering). We use a large-eddy simulation with an Aitken-mode enabled microphysics scheme to examine the role of Aitken buffering in a mid-latitude decoupled boundary layer cloud regime observed on July 15, 2017 during the ACE-ENA flight campaign: cumulus rising into stratocumulus under elevated Aitken concentrations (~100-200 mg-1). In situ measurements are used to constrain and evaluate this case study. Our simulation accurately captures observed aerosol-cloud-precipitation interactions and reveals time-evolving processes driving regime development and evolution. Aitken activation into the accumulation mode occurs primarily in the cumulus layer, providing a reservoir for turbulence and convection to carry accumulation aerosols into the drizzling stratocumulus layer above. Thus, the cloud regime is buffered against precipitation removal, reducing cloud break-up and associated increases in heterogeneity. We examine cloud evolution sensitivity to initial aerosol conditions. With halved accumulation number, Aitken aerosols restore accumulation concentrations, maintain droplet number similar to original values, and prevent cloud break-up. Without Aitken aerosols, precipitation-driven cloud break-up occurs rapidly. In this regime, mesoscale and synoptic-scale uplift enhance cloud condensate and brightness, but Aitken buffering sustains brighter, more homogeneous clouds for longer.

Coarse-grid weather and climate models rely particularly on parameterizations of cloud fields, and coarse-grained cloud fields from a fine-grid reference model are a natural target for a machine-learned parameterization. We machine-learn the coarsened-fine cloud properties as a function of coarse-grid model state in each grid cell of NOAA’s FV3GFS global atmosphere model with 200 km grid spacing, trained using a 3 km fine-grid reference simulation with a modified version of FV3GFS. The ML outputs are coarsened fine fractional cloud cover and liquid and ice cloud condensate mixing ratios, and the inputs are coarse model temperature, pressure, relative humidity, and ice cloud condensate. The predicted fields are skillful and unbiased, but somewhat under-dispersed, resulting in too many partially-cloudy model columns. When the predicted fields are applied diagnostically (offline) in FV3GFS’s radiation scheme, they lead to small biases in global-mean top-of-atmosphere (TOA) and surface radiative fluxes. An unbiased global mean TOA net radiative flux is obtained by setting to zero any predicted cloud with grid cell mean cloud fraction less than a threshold of 6.5%; this does not significantly degrade the ML prediction of cloud properties. The diagnostic, ML-derived radiative fluxes are  far more accurate than those obtained with the existing cloud parameterization in the nudged coarse-grid model, as they leverage the accuracy of the fine-grid reference simulation’s cloud properties. 

Can the current successes of global machine learning-based weather simulators be generalized beyond two-week forecasts to stable and accurate multiyear runs? The recently developed AI2 Climate Emulator (ACE) suggests this is feasible, based upon 10-year simulations trained on a realistic global atmosphere model using a grid spacing of approximately 110~km and forced by a repeating annual cycle of sea-surface temperature. Here we show that ACE, without modification, can be trained to emulate another major atmospheric model, EAMv2, run at a comparable grid spacing for at least ten years with similarly small climate biases. ACE accurately reproduces EAMv2’s frequency distribution of daily-mean precipitation, its time-mean spatial pattern of precipitation, and its space-time structure of tropical precipitation, including the Madden-Julian Oscillation. Moreover, ACE’s climate biases with respect to EAMv2 are substantially smaller than EAMv2’s own biases compared to the observed historical average surface precipitation rate and top-of-atmosphere radiative fluxes.

A document by Jacqueline M Nugent. Click on the document to view its contents.

Cirrus dominate the longwave radiative budget of the tropics. For the first time, we quantify the variability in cirrus properties and longwave cloud radiative effects (CREs) that arises from differences in microphysics within nudged global storm-resolving simulations from a single model. Nudging allows us to compute radiative biases precisely using coincident satellite measurements and to fix the large-scale dynamics across our set of simulations and isolate the influence of microphysics. We run five-day simulations with four commonly-used microphysics schemes of varying complexity (SAM1MOM, Thompson, M2005 and P3) and find that the tropical average longwave CRE varies over 20 W m$^{-2}$ between schemes. P3 best reproduces observed longwave CRE. M2005 and P3 simulate cirrus with realistic frozen water path but unrealistically high ice crystal number concentrations which commonly hit limiters and lack the variability and dependence on frozen water content seen in aircraft observations. Thompson and SAM1MOM have too little cirrus.

Physics-informed machine learning (ML) applied to geophysical simulation is developing explosively. Recently, graph neural net and vision transformer architectures have shown 1-7 day global weather forecast skill superior to any conventional model with integration times over 1000 times faster, but longer simulations rapidly degrade. ML that achieves high skill in both weather and climate applications is a tougher goal. This Commentary was inspired by \citeA{ArcomanoEtAl2023}, who show impressive progress toward that goal using hybrid ML, combining reservoir computing to a coarse-grid climate model and coupling to a separate data-driven reservoir computing model that interactively predicts sea-surface temperature. This opens new horizons; where will the next ML breakthrough come from, and is conventional climate modeling about to be disrupted?

Intense convection (updrafts exceeding 10 m∙s-1) plays an essential role in severe weather and Earth’s energy balance. Despite its importance, how the global pattern of intense convection changes in response to warmed climates remains unclear, as simulations from traditional climate models are too coarse to simulate intense convection. Here we take advantage of a kilometer-scale global storm resolving model and conduct year-long simulations of a control run, forced by analyzed sea surface temperature (SST), and one with a 4-K increase in SST for comparison. Comparisons show that the increased SST enhances the frequency of intense convection globally with large spatial and seasonal variations. Increases in the intense convection frequency do not necessarily reflect increases in convective available potential energy (CAPE). Results are also compared with traditional climate model projections. Changes in the spatial pattern of intense convection are associated with changes in planetary circulation.

A new Aitken mode aerosol scheme is developed for a large eddy simulation (LES) model in order to better investigate cloud-aerosol interactions in the marine boundary layer and to study the Aitken buffering hypothesis of McCoy et al. (2021). This scheme extends the single-mode two-moment prognostic aerosol scheme of Berner et al. (2013). Nine prognostic variables represent accumulation and Aitken log-normal aerosol modes in air and droplets as well as 3 gas species. Scavenging of interstitial aerosol by cloud and rain drops and coagulation of dry aerosol are treated using the scheme described in B13. The scheme includes a simple chemistry model with gas phase H2SO4, SO2, and DMS as prognostic variables to capture basic influences of sulfur chemistry on the model aerosols. Primary nucleation of H2SO4 aerosol particles from gas-phase H2SO4 is neglected. A deep, precipitating stratocumulus case (VOCALS RF06) is used to test the new scheme. The presence of the Aitken mode aerosol increases the cloud droplet concentration through activation of the larger Aitken particles and delays the creation of an ultraclean, strongly precipitating cumulus state. Scavenging of dry accumulation and Aitken particles by cloud and precipitation droplets accelerates the collapse. Increasing either the above-inversion Aitken concentration or the surface Aitken flux increases the Aitken population in the boundary layer and prevents the transition to an ultraclean state.

We present the characteristics of rotating convective updrafts in the 2021 version of GFDL’s Experimental System for High-resolution prediction on Earth-to-Local Domains (X-SHiELD), a kilometer scale global storm resolving model (GSRM). Rotation is quantified using 2–5 km Updraft Helicity (UH) in a year-long integration forced by analyzed SSTs. Updrafts with UH magnitudes above 50 m2/s2; are common over the mid-latitude continents, especially in the warm seasons where they are associated with severe weather but are also common over most tropical ocean basins. In nearly all areas cyclonically rotating convection dominates, with larger UH values increasingly preferring cyclonic rotation. The ratio of cyclonic to anticyclonic updrafts is largest in the subtropical and mid-latitude oceans and is slightly lower over mid-latitude continents. The ratio of cyclonic to anticyclonic updrafts can be substantively explained by the mean storm-relative helicity (SRH) in convective regions, indicating the importance for environmental controls on the sense of storm rotation, although internal storm dynamics also plays a role in the generation of anticyclonic updrafts.