Climate processes

Abstract:

<jats:p>Abstract. The rapidly growing volume of weather and climate data, both from models and observations, is increasing the pressure on data centers, restricting scientific analysis, and data distribution. For example, kilometre-scale climate models can generate petabytes of data per simulated month, making it generally infeasible to store all output. To address this challenge, numerous novel compression techniques have been proposed to ease data storage requirements. However, there exist no well-defined benchmarks for rigorously evaluating and comparing the performance of these compressors, including their impact on the data's properties. The lack of benchmarks makes it difficult to design and standardize compressors for weather and climate data, and for scientists to trust that compression errors have no significant impact on their analysis. Here, we address this gap by presenting ClimateBenchPress, a benchmark suite for lossy compression of climate data, which defines both data sets and evaluation techniques. The benchmark covers climate variables following various statistical distributions at medium to very high resolution in time and space, from both numerical models and satellite observations. To ensure a fair comparison between different compressors, each variable comes with a set of maximum error bound checks that the lossy compressors need to pass. By evaluating an initial set of baseline compressors on the benchmark, we gather practical insights for effective application of lossy compression. Our benchmark is open source and extensible: users can easily add new compressors, data sources, and evaluation metrics depending on their own specific use cases.</jats:p>

Abstract:

Abstract In May 2025, nearly 700 participants from all around the world coalesced at 10 regional nodes and a few satellite nodes to take part in a global hackathon of kilometer-scale (horizontal grid spacing < 10 km) regional and global Earth system models. Exciting science is emerging from these efforts, ranging across novel model analysis, new ways of integrating with satellite data, and emulation with machine learning. New technologies were trialed that enable the community to work in new and complementary ways to democratize access to global information at a local scale from a set of the world’s highest-resolution climate models. The hackathon demonstrated how exascale data can be organized to be accessible to anyone. Fundamentally, the community could apply these techniques and technologies to move toward more participative models for coproduction and delivery of diverse sources of climate information for climate scientists and citizens alike. Significance Statement This report documents the results of a global hackathon for kilometer-scale atmospheric models held in May 2025. Nearly 700 participants at 10 nodes around the world were able to produce more than a 1000 plots from the most advanced global and regional atmospheric models. New tools were used to access and analyze these large datasets. Success was enabled by 1) using data at the right scale, 2) the ability to access only needed data, and 3) a common open-source analysis platform to analyze data without copying it. The community could apply these techniques and technologies widely to move toward more participative models for the delivery of climate information.

Abstract:

Our limited understanding of clouds is the dominant source of uncertainty in future climate predictions. Understanding how changing atmospheric environmental conditions constrain cloud organisational patterns and their radiative effects is key to understanding their impact on future climate change. We present CloudFlow, a flow matching model that is able to generate high-resolution cloud structures conditioned on coarse-scale atmospheric conditions. Our model generates realistic cloud structures that match the spectra and distributions of the original high-resolution scenes. CloudFlow introduces a new modeling regime to study how atmospheric environmental conditions impact cloud morphologies which will contribute to an improved understanding of cloud feedbacks, the cloud response to a changing climate and its effect on climate itself.

Abstract:

Deep convective clouds substantially modify the balance of shortwave and longwave radiative energy at the top of the atmosphere. Although in the present-day these effects approximately balance out, projected changes in deep convective clouds could alter the future top-of-atmosphere energy balance. Past studies have found relationships between convection and anvil clouds, but our understanding of how convection typically controls the properties and evolution of anvil clouds that determine anvil radiative effects remains incomplete, limiting our ability to explain or justify projected changes in cloud optical properties. This manuscript presents a new method to track the lifecycle of deep convective clouds and their convective cores in three-dimensional space in km-scale global climate models. An analysis of how convective organisation, intensity and area relate to anvil properties in the ICOsahedral Non-hydrostatic (ICON) model is then presented. Approximately 1000 deep convective clouds are tracked over one simulation week in the tropical Amazon region. We find that while both updraft intensity and area correspond to larger anvils, the correlation between convective area and anvil size is stronger than that between anvil size and updraft intensity. Updraft intensity was associated with a 4-fold increase in anvil extent when convective cores were larger, compared to when they were in the bottom 50th size percentile. This result could not be explained by associated changes in peak convective mass flux or organisation. These results indicate how changes in the frequency or typical size of convective updrafts may link to changes in anvil development, extent and, ultimately, radiative effects.