Kairui Feng

Artificial‐intelligence weather prediction models have recently surpassed numerical models in large‐scale skill, but they still systematically underestimate typhoon intensity due to their reliance on coarse‐resolution training data from ERA5. To overcome this limitation, we constructed a bespoke 9‐km high‐resolution typhoon reanalysis (HiRes) by dynamically downscaling ERA5 with the operational Shanghai Typhoon Model (SHTM). The evaluation shows that HiRes has a clear advantage over ERA5 in aspects such as surface wind and land precipitation. HiRes inherently combines a large‐scale deterministic background with small‐scale residual corrections, motivating a learning framework that mirrors this structure. We therefore introduce the Intelligent Shanghai Typhoon Model (ISTM), a regression‐residual diffusion emulator that first predicts the conditional mean fields and then stochastically reconstructs the residuals with a conditional diffusion model. Trained on HiRes, ISTM substantially improves wind and precipitation intensity and organization compared with ERA5 and deterministic U‐Net baselines. After fine‐tuning, ISTM can downscale AIFS forecast fields to produce SHTM‐consistent, physically coherent fields; while maintaining track errors comparable to AIFS, it reduces the 72‐hr mean typhoon intensity forecast error by about 39.5% relative to AIFS.

The chance constrained optimal power flow (CC-OPF) essentially finds the low-cost generation dispatch scheme ensuring operational constraints are met with a specified probability, termed the security level. While the security level is a crucial input parameter, how it shapes the CC-OPF feasibility boundary has not been revealed. Changing the security level from a parameter to a decision variable, this paper proposes a security maximization approach based on the chance constrained DC-OPF model, termed inverse CC-OPF (ICC-OPF), that seeks the maximum security level achievable by the system. To efficiently solve the ICC-OPF, we design a Newton-Raphson-like iteration algorithm leveraging the duality-based sensitivity analysis of an associated surrogate problem. Numerical experiments validate the proposed approach, revealing complex feasibility boundaries for security levels that underscore the importance of coordinating security levels across multiple chance constraints.

The emerging tropical cyclone (TC)-blackout-heatwave compound risk under climate change is not well understood. In this study, we employ projections of TCs, sea level rise, and heatwaves, in conjunction with power system resilience modeling, to evaluate historical and future TC-blackout-heatwave compound risk in Louisiana, US. We find that the return period for a compound event comparable to Hurricane Ida (2021), with approximately 35 million customer hours of simultaneous power outage and heatwave exposure in Louisiana, is around 278 years in the historical climate of 1980–2005. Under the SSP5-8.5 emissions scenario, this return period is projected to decrease to 16.2 years by 2070–2100, a ~17 times reduction. Under the SSP2-4.5 scenario, it decreases to 23.1 years, representing a ~12 times reduction. Heatwave intensification is the primary driver of this increased risk, reducing the return period by approximately 5 times under SSP5-8.5 and 3 times under SSP2-4.5. Increased TC activity is the second driver, reducing the return period by 40% and 34% under the respective scenarios. These findings enhance our understanding of compound climate hazards and inform climate adaptation strategies.

Conventional computational models of climate adaptation frameworks inadequately consider decision-makers’ capacity to learn, update, and improve decisions. Here, we investigate the potential of reinforcement learning (RL), a machine learning technique that efficaciously acquires knowledge from the environment and systematically optimizes dynamic decisions, in modeling and informing adaptive climate decision-making. We consider coastal flood risk mitigations for Manhattan, New York City, USA (NYC), illustrating the benefit of continuously incorporating observations of sea-level rise into systematic designs of adaptive strategies. We find that when designing adaptive seawalls to protect NYC, the RL-derived strategy significantly reduces the expected net cost by 6 to 36% under the moderate emissions scenario SSP2-4.5 (9 to 77% under the high emissions scenario SSP5-8.5), compared to conventional methods. When considering multiple adaptive policies, including accomodation and retreat as well as protection, the RL approach leads to a further 5% (15%) cost reduction, showing RL’s flexibility in coordinatively addressing complex policy design problems. RL also outperforms conventional methods in controlling tail risk (i.e., low probability, high impact outcomes) and in avoiding losses induced by misinformation about the climate state (e.g., deep uncertainty), demonstrating the importance of systematic learning and updating in addressing extremes and uncertainties related to climate adaptation.

The emerging tropical cyclone (TC)-blackout-heatwave compound hazard under climate change are not well understood. In this study, we employ future projections of TCs, sea levels, and heatwaves, in conjunction with power system resilience modeling, to evaluate historical and future TC-blackout-heatwave compound hazard risks in Louisiana, US. We find that the return period for a compound hazard event comparable to Hurricane Ida (2021), with approximately 35 million customer hours of simultaneous power outage and heatwave exposure in Louisiana, is around 278 years in the historical climate (1980-2005). Under the emissions scenario SSP5 8.5 (SSP2 4.5), this return period may decrease by a factor of ~17×(10x) to 16.2 (28.4) years in the future climate (2070-2100). The significant increase in risk can be primarily attributed to projected escalations in heatwaves, which result in an approximate 5(2)-fold decrease in compound hazard return period, and in TC activity, which cause an estimated 2(1)-fold decrease in the return period. The findings contribute to our knowledge of and adaptation to compound climate hazards.

City-scale traffic data, such as traffic flow, speed, and density on every road segment, are the foundation of modern urban research. However, accessing such data on a city scale is challenging due to the limited number of sensors and privacy concerns. Consequently, most of the existing traffic datasets are typically limited to small, specific urban areas with incomplete data types, hindering the research in urban studies, such as transportation, environment, and energy fields. It still lacks a city-scale traffic dataset with comprehensive data types and satisfactory quality that can be publicly available across cities. To address this issue, we propose a unified approach for producing city-scale traffic data using the classic traffic assignment model in transportation studies. Specifically, the inputs of our approach are sourced from open public databases, including road networks, traffic demand, and travel time. Then the approach outputs comprehensive and validated citywide traffic data on the entire road network. In this study, we apply the proposed approach to 20 cities in the United States, achieving an average correlation coefficient of 0.79 in average travel time and an average relative error of 5.16% and 10.47% in average travel speed when compared with the real-world data.