Tomorrows Children’s Fund

This study assesses the performance of a regional WRF configuration used operationally by SIMEPAR (SIWRF) and a set of MPAS experiments at 5-km grid spacing over southern Brazil to predict four recent severe weather events. MPAS is tested using a 200-5 km global variable-resolution (VR) mesh with two physics suites—Mesoscale Reference (MR) and Convection-Permitting (CP)—and two sources of initial conditions — NCEP-GFS and ECMWF-IFS forecasts. Skill is evaluated with respect to 48-h accumulated precipitation (correlation, bias, RMSE), hourly precipitation (fractions skill score; FSS, and QPF measures), radar reflectivity (FSS), near-surface variables (MAE, Taylor diagrams), and vertical soundings (bias, RMSE). Across events, MPAS generally outperforms SIWRF for precipitation, with CP physics suite often reducing positive bias relative to MR, while SIWRF tends to underestimate totals during the heaviest events. Notably, MR-IFS achieves the highest correlation for the event dominated by a long-lived, organized convective system, whereas CP-IFS provides the most balanced metrics in two other cases, including the coastal event, consistent with improved placement and intensity of convection. SIWRF exhibits the lowest skill for radar reflectivity, while MPAS shows event-dependent gains, particularly for MR-IFS and CP-GFS. Analysis of convective parameters reveal that MR configurations produce higher median CAPE and 0–3-km SRH than CP and SIWRF, implying more favorable storm environments. These results highlight event-dependent sensitivity to physics and initial conditions and suggest that tailoring MPAS configuration to event features (e.g., favoring CP physics for coastal/diurnally forced convection and MR physics for strongly forced organized systems) may improve forecasts, but this inference is preliminary and requires broader testing (larger case set, data assimilation, ensemble and operational-runtime evaluation) before operational adoption. • First comparison of MPAS and operational WRF for severe weather in southern Brazil. • MPAS with 200-5 km global VR mesh often outperformed 5-km regional WRF. • MPAS precipitation skill and biases strongly depend on physical options and IC source. • MPAS forecasts improved when initialized with ECMWF-IFS ICs rather than NCEP-GFS.

Abstract The Tomorrow.io microwave sounder (TMS) builds on a heritage of passive microwave sounders (PMWSs) that have well‐established value in numerical weather prediction (NWP). The TMS constellation of CubeSats uses a combination of sun‐synchronous and inclined orbits to fill gaps in current public PMWS data. This work establishes a data assimilation methodology for the TMS and demonstrates its real global weather forecast impacts with observing system experiments. Our assimilation methodology for the TMS encompasses all‐sky error modeling and quality control, and novel variational bias correction predictors. The all‐sky error model is based on techniques in earlier literature, and a new cloud impact filter enables more effective assimilation of surface‐sensitive TMS channels. TMSs in inclined orbits are exposed to a transient thermal environment that causes periodic biases in the current level 1 calibrated brightness temperatures. In response, we develop variational bias correction predictors related to the unique night–day cycle of each satellite. The new predictors partially mitigate the transient instrument biases and improve verification scores for upper‐level temperature forecasts. However, this is a partial solution, and remains an active area of investigation. In our research‐quality NWP setting, assimilating two TMS instruments in January 2025 achieves 6‐hr forecast accuracy improvements similar to two advanced technology microwave sounder instruments for water vapor, and 50% of two advanced technology microwave sounders for tropospheric temperature. Statistically significant forecast accuracy improvements from the TMS persist for water vapor up to 3 days, and for temperature and winds up to 2 days. The TMS provides complementary information content to public PMWSs. Additional efforts at operational forecasting centers are ongoing to conduct more complete investigations.

Supplementary figures and domains Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Recent advances in generative models have significantly facilitated the development of personalized content creation. Given a small set of images containing a user-specific concept, personalized image generation allows the user to create images that incorporate that concept while adhering to provided text descriptions. The technologies used for personalization have evolved alongside the development of generative models, with their distinct and interrelated components. In this survey, we present a comprehensive review of generalized personalized image generation across various generative models, including traditional GANs, contemporary text-to-image diffusion models, and emerging multi-modal autoregressive (AR) models. We first define a unified framework that standardizes the personalization process across different generative models, encompassing three key components: inversion spaces, inversion methods, and personalization schemes. This unified framework offers a structured approach to dissecting and comparing personalization techniques across different generative architectures. Building upon our framework, we provide an in-depth analysis of personalization techniques within each generative model, highlighting their unique contributions and innovations. Through comparative analysis, we elucidate the current landscape of personalized image generation, identifying commonalities and distinguishing features of existing methods. Finally, we discuss open challenges in the field and propose potential directions for future research. We keep a bibliography of related works at https://github.com/csyxwei/Awesome-Personalized-Image-Generation.

In addition to the unprecedented ability in imaginary creation, large text-to-image models are expected to take customized concepts in image generation. Existing works generally learn such concepts in an optimization-based manner, yet bringing excessive computation or memory burden. In this paper, we instead propose a learning-based encoder, which consists of a global and a local mapping networks for fast and accurate customized text-to-image generation. In specific, the global mapping network projects the hierarchical features of a given image into multiple "new" words in the textual word embedding space, i.e., one primary word for well-editable concept and other auxiliary words to exclude irrelevant disturbances (e.g., background). In the meantime, a local mapping network injects the encoded patch features into cross attention layers to provide omitted details, without sacrificing the editability of primary concepts. We compare our method with existing optimization-based approaches on a variety of user-defined concepts, and demonstrate that our method enables high-fidelity inversion and more robust editability with a significantly faster encoding process. Our code is publicly available at https://github.com/csyxwei/ELITE.

.

Precipitation type is a key parameter used for better retrieval of precipitation characteristics as well as to understand the cloud–convection–precipitation coupling processes. Ice crystals and water droplets inherently exhibit different characteristics in different precipitation regimes (e.g., convection, stratiform), which reflect on satellite remote sensing measurements that help us distinguish them. The Global Precipitation Measurement (GPM) Core Observatory’s microwave imager (GMI) and dual-frequency precipitation radar (DPR) together provide ample information on global precipitation characteristics. As an active sensor, the DPR provides an accurate precipitation type assignment, while passive sensors such as the GMI are traditionally only used for empirical understanding of precipitation regimes. Using collocated precipitation type flags from the DPR as the “truth”, this paper employs machine learning (ML) models to train and test the predictability and accuracy of using passive GMI-only observations together with ancillary information from a reanalysis and GMI surface emissivity retrieval products. Out of six ML models, four simple ones (support vector machine, neural network, random forest, and gradient boosting) and the 1-D convolutional neural network (CNN) model are identified to produce 90–94% prediction accuracy globally for five types of precipitation (convective, stratiform, mixture, no precipitation, and other precipitation), which is much more robust than previous similar effort. One novelty of this work is to introduce data augmentation (subsampling and bootstrapping) to handle extremely unbalanced samples in each category. A careful evaluation of the impact matrices demonstrates that the polarization difference (PD), brightness temperature (Tc) and surface emissivity at high-frequency channels dominate the decision process, which is consistent with the physical understanding of polarized microwave radiative transfer over different surface types, as well as in snow and liquid clouds with different microphysical properties. Furthermore, the view-angle dependency artifact that the DPR’s precipitation flag bears with does not propagate into the conical-viewing GMI retrievals. This work provides a new and promising way for future physics-based ML retrieval algorithm development.