20 January 2023 in Journal of Advances in Modeling Earth Systems
Recently, a growing number of studies have used machine learning (ML) models to parameterize computationally intensive subgrid‐scale processes in ocean models. Such studies typically train ML models with filtered and coarse‐grained high‐resolution data and evaluate their predictive performance offline, before implementing them in a coarse resolution model and assessing their online performance. In this work, we systematically benchmark the online performance of such models, their generalization to domains not encountered during training, and their sensitivity to data set design choices. We apply this proposed framework to compare a large number of physical and neural network (NN)‐based parameterizations. We find that the choice of filtering and coarse‐graining operator is particularly critical and this choice should be guided by the application. We also show that all of our physics‐constrained NNs are stable and perform well when implemented online, but generalize poorly to new regimes. To improve generalization and also interpretability, we propose a novel equation‐discovery approach combining linear regression and genetic programming with spatial derivatives. We find this approach performs on par with neural networks on the training domain but generalizes better beyond it. We release code and data to reproduce our results and provide the research community with easy‐to‐use resources to develop and evaluate additional parameterizations.
15 August 2020 in Journal of Climate
Precipitation extremes intensify in most regions in climate model projections. Changes in vertical velocities contribute to the changes in intensity of precipitation extremes but remain poorly understood. Here, we find that midtropospheric vertical velocities in extratropical precipitation extremes strengthen overall in simulations of twenty-first-century climate change. For each extreme event, we solve the quasigeostrophic omega equation to decompose this strengthening into different physical contributions. We first consider a dry decomposition in which latent heating is treated as an external forcing of upward motion. Much of the positive contribution to upward motion from increased latent heating is offset by negative contributions from increases in dry static stability and changes in the horizontal length scale of vertical velocities. However, taking changes in latent heating as given is a limitation when the aim is to understand changes in precipitation, since latent heating and precipitation are closely linked. Therefore, we also perform a moist decomposition of the changes in vertical velocities in which latent heating is represented through a moist static stability. In the moist decomposition, changes in moist static stability play a key role and contributions from other factors such as changes in the depth of the upward motion increase in importance. While both dry and moist decompositions are self-consistent, the moist dynamical perspective has greater potential to give insights into the causes of the dynamical contributions to changes in precipitation extremes in different regions.
20 January 2022 in Journal of Geophysical Research: Atmospheres
This study explores synoptic‐scale circulation conditions that favor vertical motion on heavy rainfall days (HRDs) over North China (NC) and analyzes the quantitative contribution to vertical motion by different forcings using the quasi‐geostrophic (QG) omega equation. On HRDs, moisture is brought from the South China Sea and East China seas by surface cyclonic vortex activity. When South Asian High is located anomalously northward, NC is on the south side of the upper‐level jet (ULJ) stream entrance region. Diagnosis of the QG omega equation shows that vertical motion on HRDs is caused by dynamic forcing, diabatic forcing, and topographic forcing. The greatest contribution to vertical motion is from diabatic forcing, while topographic contributes to vertical motion at low levels. The dynamic forcing, which is manifested by the Q‐vector, indicates that both the ULJ and baroclinic trough generate the vertical motion. Furthermore, the effect of the baroclinic trough is larger than that of the ULJ. On medium rainfall days, the contributions from dynamic forcing, diabatic forcing, and topographic forcing are clearly weaker than on HRDs, with reductions of 62.3%, 71.3%, and 73.8%, respectively.
1 March 2021 in Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
Projections of precipitation extremes in simulations with global climate models are very uncertain in the tropics, in part because of the use of parameterizations of deep convection and model deficiencies in simulating convective organization. Here, we analyse precipitation extremes in high-resolution simulations that are run without a convective parameterization on a quasi-global aquaplanet. The frequency distributions of precipitation rates and precipitation cluster sizes in the tropics of a control simulation are similar to the observed distributions. In response to climate warming, 3 h precipitation extremes increase at rates of up to in the tropics because of a combination of positive thermodynamic and dynamic contributions. The dynamic contribution at different latitudes is connected to the vertical structure of warming using a moist static stability. When the precipitation rates are first averaged to a daily timescale and coarse-grained to a typical global climate-model resolution prior to calculating the precipitation extremes, the response of the precipitation extremes to warming becomes more similar to what was found previously in coarse-resolution aquaplanet studies. However, the simulations studied here do not exhibit the high rates of increase of tropical precipitation extremes found in projections with some global climate models. This article is part of a discussion meeting issue ‘Intensification of short-duration rainfall extremes and implications for flash flood risks’.
3 June 2021 in Geophysical Research Letters
The 2019–2020 Australian wildfire crisis broke the historical bushfire record and heavily contaminated the continental and offshore atmosphere. This study found that lightning strokes increase considerably, by 73% over land and 270% over ocean, during the wildfire season. Thermodynamic parameters support a weaker forcing, unfavorable for frequent lightning activity over ocean. Clear augmentation of smaller cloud ice particles is identified with aerosol, while cloud liquid water path changes are feeble over ocean. Added aerosol invigorates positive intra‐cloud (IC) strokes and negative cloud‐to‐ground (CG) strokes in moist oceanic convection and facilitates a noticeable positive correlation between precipitation and lightning strokes. Rainfall events accompanied by lightning increase by 240% with added aerosol. Aerosol advected from land to ocean can lead to a larger hydrometeor concentration and smaller‐size ice crystals above the freezing level and thereby, invigorate convective strength systematically to stimulate more frequent and more robust mixed‐phase development, energizing the lightning discharge process.