Papers for

emergency management teams

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Italian deep learning model improves short range rainfall forecasts

IRENE: A Convolutional GRU Ensemble Model for Radar Precipitation Nowcasting over Italy

Abstract: We present IRENE (Italian Radar Ensemble Nowcasting Experiment), a deep learning model for probabilistic short-range precipitation nowcasting over the Italian domain at \SI{1}{km} spatial and 5 min temporal resolution. IRENE adopts an encoder--forecaster architecture built on multi-scale Convolutional Gated Recurrent Units (ConvGRUs), trained on the national radar composite produced by the Italian Civil Protection Department (DPC). An importance-sampling scheme focuses training on precipitation-relevant events, while the almost-fair Continuous Ranked Probability Score (afCRPS) is adopted as the primary probabilistic loss function. Two additional training configurations are proposed: an adversarial (GAN) variant, IRENE-GAN, designed to improve the spatial sharpness of the generated forecasts, and a spectrally constrained variant, IRENE-GAN-RAPSD, in which the adversarial objective is complemented by an explicit penalty on the radially averaged power spectral density. The three configurations are evaluated against the stochastic extrapolation method STEPS and the pre-trained deep learning model DGMR. All IRENE configurations attain a lower Continuous Ranked Probability Score than both benchmarks at every lead time and rank histograms closer to uniformity, indicating better probabilistic skill and ensemble calibration. In terms of ensemble-mean mean absolute error the advantage is confined to the first 90 min, beyond which the strongly damped DGMR fields and, to a lesser extent, STEPS become competitive. Spectral analysis shows that the adversarial training removes the progressive loss of small-scale variance exhibited by IRENE, at the cost of an excess of fine-scale power at long lead times that the spectral penalty only partially controls.

Tue 15 SeptMachine Learning
The gist
Predicting where and when rain will fall in the next few minutes is important but difficult. The authors created IRENE, a deep learning model that uses radar data to make short-term rain forecasts across Italy at high detail. They trained it to focus on real rain events and compared three versions, including one using a technique to make sharper images. All versions performed better than previous methods at predicting the chance of rain and how confident the model’s guesses were. However, longer-term forecasts lost some small details unless special training was used.
Open 2609.17175v1

Weather agent improves alerts with smart learning and reasoning

Agent-Based ML-LLM Fusion with Self-Optimizing Prompts for Plateau Weather Alerts

Abstract: To address insufficient contextualization, weak generalization, and poor scenario adaptation in tourism meteorological services, we propose SmartWeatherAgent--a unified three-stage architecture integrating intent recognition, hazard prediction, and reasoning-enhanced generation. The system fuses rule-based methods with large language models to parse queries at multiple granularities and employs a LightGBM model enriched with highland-specific features (e.g., wind speed abruptness rate), achieving an F1-Macro score of 0.605 with 1.60 ms latency on high-wind, precipitation, and low-temperature events. A 12-round micro-step prompt self-optimization loop boosts the composite warning quality score S_final from 4.2 (B01) to 8.9 (B12, +112%). Key improvements include a sharp rise in B08 from data source citation (6.5 -> 8.5), sustained high performance in B10 via physical mechanism explanation, and a peak scientific rigor score of 9.2 in B12 through explicit uncertainty statements. The system autonomously generates structured warnings that integrate causal mechanisms, spatiotemporal evolution, quantitative evidence, regulatory references, and confidence statements--enhancing professional depth, logical rigor, and scientific soundness, and advancing meteorological services toward proactive perception, explainable decision-making, and intelligent agency.

Wed 9 SeptArtificial IntelligenceMachine Learning
The gist
Weather alerts for tourists often struggle to give clear, reliable, and specific information for mountain areas. The authors built a system called SmartWeatherAgent that understands questions better, predicts hazards more accurately, and creates detailed warnings with reasons and evidence. It combines rules with smart language models and a special machine learning method trained on mountain weather features. The system also improves itself step-by-step, making alerts more trustworthy and easier to understand.
Open 2609.10135v1