Long-term climate change reshapes crop growing conditions and threatens global food security. Using 1991-2020 observations from 154 Chinese agro-meteorological stations, this study applies spatiotemporal panel regression to quantify maize's nonlinear yield responses to growing-season mean temperature, cumulative precipitation and total solar radiation. During 1991-2020, maize-planting regions experienced rising temperature and precipitation alongside declining solar radiation. Under prevailing local agronomic management, the sample-wide statistical climate conditions corresponding to peak observed maize yields are 24.3 °C mean temperature, 770 mm cumulative precipitation, and 3210 MJ/m2 solar radiation. Simulations under an exploratory compound-stress scenario (+1.5 °C warming, 20 % precipitation reduction, 10 % solar-radiation reduction) yield a national mean maize yield decline of 2.9 %, featuring strong regional divergence: Northwest China gains 6.4 % yield, whereas Northeast China suffers a 9.7 % loss. We develop a region-calibrated nonlinear statistical model to fill the gap of China-specific maize yield-loss functions for Integrated Assessment Models. We identify a baseline-dependent climatic-adaptation mechanism: yield outcomes are determined by the coupling between local baseline climate and climate perturbations, where regions near statistical climate optima show higher resilience and cold-marginal zones may benefit from moderate warming. As a research-oriented prototype rather than an operational forecasting tool, this study provides region-specific parameters for impact modelling and evidence for targeted agricultural-climate adaptation policy-making.
Southeast Asia is central to global rice security; however, rising heat extremes increasingly threaten the stability of major rice-production systems. Here, we combined a high-resolution downscaled six-model Coupled Model Intercomparison Project Phase 6 (CMIP6) ensemble at 8-km resolution with CD35, defined as the maximum consecutive duration of days with daily maximum temperature (T max) at or above 35 °C within rice-growth phases, normalized by phase duration, to quantify sustained heat exposure across the principal rice-growing systems of Thailand, Vietnam, and Indonesia under SSP1-2.6, SSP2-4.5, and SSP5-8.5. Risk was evaluated for the historical baseline (1995–2014), near future (2031–2050), and far future (2080–2099) using country-specific crop calendars and phenological stages. Projected heat exposure increased across all three countries, with the strongest intensification during the midseason. Thailand showed the earliest and most severe shift, reaching a high to extreme CD35-based risk across most seasons by 2031–2050 under SSP2-4.5 and becoming uniformly extreme by 2080–2099 under SSP5-8.5. By the late century under SSP5-8.5, midseason CD35 reached up to 60% in the Mekong Delta, 40%–55% in northeastern Thailand, and 50%–65% in the major rice-growing areas of Java, indicating that threshold-exceeding heat is becoming concentrated in the core production systems of the region. Together, these findings identify midseason as the critical window for future rice heat stress and provide a scalable framework to support climate services, adaptation planning, and prioritization of heat-risk management under future warming.
Urban heat adaptation requires spatially explicit information on whether ecological cooling capacity is aligned with populations experiencing greater heat-related need. However, green-space assessments often examine thermal performance, vulnerable populations, and accessibility separately, limiting their value as integrated climate-service diagnostics. This study develops an exposure–vulnerability–accessibility (E–V–A) framework that links heterogeneous green-space cooling supply, heat-related cooling need, and potential accessibility, and demonstrates it in Wuhan's main urban area. Summer heat exposure intensified from 2016 to 2024. The LST contrast between green spaces and their surroundings flattened after approximately 120 m and approached background conditions at 330–360 m; 300 m was therefore used as a representative cooling-performance reference rather than a service radius. Cooling accessibility was highly uneven, with lower values in dense central communities. Twenty-seven communities exhibited high heat exposure, high demographic vulnerability, and low cooling accessibility; 26 of these were classified into the highest diagnostic priority tier (P1). Overall, the framework provides a potentially transferable community-scale diagnostic for identifying spatial mismatches between ecological cooling capacity and heat-related need and for guiding subsequent local adaptation appraisal.Practical Implications.The E–V–A framework translates community-scale heat exposure, demographic vulnerability, green-space cooling performance, and cooling accessibility into differentiated diagnostic priorities. Its practical value lies in moving beyond green-space quantity alone: communities with similar levels of green-space provision may face very different cooling-access conditions because of differences in thermal performance, demographic vulnerability, spatial separation, and competition among potential beneficiaries. The P1–P4 classification therefore helps identify where further local assessment is most warranted, while the underlying E–V–A configuration indicates which dimension or combination of dimensions contributes to a community's disadvantage.Candidate adaptation options can be considered in relation to the underlying E–V–A configuration. For P1 communities, where high exposure, high vulnerability, and low accessibility converge, further local assessment may consider combinations of stronger local cooling provision and age-sensitive heat-health support. For P2 communities, assessment can be differentiated according to the dominant two-dimensional configuration: HE–HV communities may warrant consideration of heat-mitigation and health-protection measures; HE–LA communities may warrant assessment of exposure reduction together with improvements in the amount, thermal effectiveness, or spatial availability of nearby cooling supply; and HV–LA communities may warrant consideration of more proximate and age-accessible cooling opportunities together with targeted support for older residents. Where local assessment confirms physical access barriers, pedestrian connectivity or entrance improvements may also be considered. P3 communities may warrant preventive and lower-intensity measures, including neighborhood greening, shaded resting spaces, heat-risk communication, and periodic reassessment, whereas P4 communities generally warrant routine monitoring and maintenance rather than immediate detailed intervention appraisal.The framework is intended to support urban planning and green-space management agencies, district and community administrations, public health agencies, and climate-service providers. Planning agencies can use the diagnostic priority maps to identify locations where further assessment of green-space investment, park upgrading, neighborhood-scale cooling measures, or heat-health responses may be warranted. Public health agencies can use the exposure, vulnerability, and priority layers to identify communities that may require closer attention in heat-warning, temporary cooling, or outreach programs for older residents, while local administrations can use E–V–A configurations to identify the dominant dimension or combination of dimensions associated with each community's disadvantage. Overall, the framework provides a practical basis for differentiated adaptation planning by identifying both where further assessment is most warranted and which E–V–A dimensions should be considered in that assessment. The assessment can be updated as new summer thermal imagery, demographic data, or green-space inventories become available, with an annual update providing a practical cycle where sufficiently current data are available. Uncertainty should be communicated through input-data timing and provenance, modeling assumptions, and sensitivity to alternative parameter settings, with communities whose classifications are more sensitive to alternative parameter settings identified for additional local assessment.
Energy-system studies often rely on a small number of weather years to test system performance under extreme conditions. Translating observational and climate-projection datasets into traceable and model-ready stress-test scenarios is an important interface between climate information and energy-system modelling. However, the selection of heatwave stress-test years is difficult to reproduce because the link between climate data, heatwave definition, spatial aggregation and scenario year is not always documented. This study presents a transparent workflow for selecting heatwave years from gridded observations and climate-model projections. The workflow uses the Heat Wave Magnitude Index daily (HWMId) as an event-based metric that combines heatwave duration and locally normalized temperature anomaly and ranks years according to spatially aggregated heatwave severity. Applied to a Germany–France case study, the workflow identifies 2003 as the strongest completed historical heatwave year and shows that the June 2026 event, evaluated using ERA5 and NOAA CORe data through 1 July 2026, approaches but remains slightly below the 2003 benchmark. Future stress-test years vary across CORDEX-CMIP5 and CORDEX-CMIP6 projection chains, confirming that selected years should be reported together with the climate data product, model chain, spatial domain and ranking criterion defining them. Sensitivity analyses across temperature thresholds, reanalysis datasets, regional definitions, electricity-generation-capacity and population weightings show robust stress-test years per dataset and model chain, although their ordering can change. The workflow provides energy-system researchers and practitioners with a reproducible way to translate climate data into traceable heatwave stress-test scenarios while retaining information on the data product, climate-model chain, spatial domain and selection criterion.
Seasonal forecasting is a key tool for managing climate risks in arid regions, yet its performance over the Arabian Peninsula (AP) remains insufficiently characterized. In this study, we developed and evaluated a multi-model seasonal forecasting system for the AP, comparing simulations from the Copernicus Climate Change Service (C3S), the North American Multimodel Ensemble (NMME), and a combined ensemble (CMME). Forecasts were post-processed using two alternative methods, the extended logistic regression (ELR) and the counting approach, and verified using probabilistic skill metrics. Results show that no single system or method consistently outperforms the others. The counting method generally yields higher skill than the ELR across most regions, but also produces some areas with negative skill that are not observed in the ELR results. The C3S and CMME ensembles show comparable performance, while the NMME exhibits localized skill in predicting certain extreme events. A sensitivity study indicates that the spatial resolution of seasonal forecast models plays a significant role in improving retrospective skill. Analysis of heavy rainfall episodes suggests that seasonal forecasts can potentially provide early indications of extreme precipitation in the region, though predictability varies spatially and seasonally. These findings underscore the need to communicate forecast skill alongside probabilistic outcomes to ensure informed use by decision-makers. The proposed forecasting framework will form the basis of an operational seasonal prediction system for the AP, supporting climate-resilient planning in sectors such as water management, agriculture, and infrastructure. Future developments will focus on downscaling approaches to enhance local-scale prediction of extreme rainfall events.
