پیش‌نگری و تحلیل تغییرات امواج گرمایی خاورمیانه بر پایه شبیه‌سازی‌های چندمدلی CMIP6 در دوره 2026-2055

Extended Abstract

Introduction

The increase in greenhouse gas concentrations in recent decades, alongside the intensification of global warming, has led to more frequent, more intense, and longer-lasting extreme atmospheric events—especially heatwaves. Rising mean air temperatures and a shift in the temperature distribution toward higher values have caused warm-season thermal thresholds to be exceeded more quickly and for longer periods. As a result, episodes of extreme heat not only occur more often but also last longer and are associated with higher peak temperatures, placing greater stress on human health and infrastructure. The Middle East is considered one of the most vulnerable regions to heat stress due to the dominance of arid and semi-arid climates, limited water resources, and the high sensitivity of natural systems. At the same time, population growth, rapid urbanization, and the expansion of the urban heat island effect can further exacerbate heatwave impacts in many cities across the region. Therefore, monitoring and projecting changes in heatwaves is essential for risk management, adaptation planning, and reducing socio-economic consequences in the Middle East.



Materials and Methods

In this study, the Middle East was selected as the target region and, to capture spatial heterogeneity, was subdivided into three sub-regions (northern, central, and southern) based on geographic–climatic characteristics and temperature gradients. To characterize past climate conditions and project future heatwave changes, daily maximum temperature from ERA5 for the reference period 1985–2014 was combined with 0.25° downscaled CMIP6 projections from the NEX-GDDP dataset. Simulations from three selected CMIP6 models were analyzed for the future period 2026–2055 under three emissions pathways_SSP1-2.6 (low), SSP2-4.5 (intermediate), and SSP5-8.5 (high)_to span a range of plausible futures. Before index calculation, all datasets underwent preprocessing, including quality control, handling of missing values, and calendar harmonization (removal of leap days) to ensure consistency across products. Model skill for the reference period was assessed by benchmarking model outputs against ERA5 using the root-mean-square error (RMSE), mean bias, and the Pearson correlation coefficient (R), providing a basis for uncertainty assessment and confidence in future projections.

Heatwave intensity was quantified following Russo et al. (2014), where a heatwave is defined as at least three consecutive days with daily maximum temperature exceeding a day-specific 90th-percentile threshold computed from the 1985–2014 baseline using a centered 31-day moving window; the annual heatwave intensity index was then derived as the maximum intensity among all events in each year.



Results and Discussion

The findings indicate that the performance of climate models in reproducing maximum temperature over the Middle East is significantly season-dependent. In summer, the models achieve the lowest systematic and random errors alongside the highest correlation with reanalysis data, demonstrating strong skill in representing both absolute temperatures and their variability during the hottest season. In contrast, winter exhibits the weakest performance: higher bias and overall error combined with reduced correlation point to limitations in capturing wintertime temperature variability. Spring shows an improvement relative to winter, although correlations remain only moderate, whereas autumn—characterized by very high correlation and comparatively moderate overall error—provides robust reliability for temperature-based analyses. Overall, these results confirm that model outputs are more dependable for temperature applications in summer and autumn, while winter requires greater emphasis on seasonal bias correction and targeted calibration. The annual spatial evaluation further shows that model performance is acceptable and relatively stable across much of the Middle East, with errors increasing mainly in areas of complex topography and along land–sea transition zones. This is physically plausible, as simulations in such regions are more sensitive to local processes and land–atmosphere interactions and depend more strongly on parameterization quality. Although the bias pattern is relatively coherent at the regional scale, it appears amenable to correction: a tendency toward underestimation is evident in parts of the Arabian Peninsula, while overestimation occurs in portions of the northwest. Taken together, these results suggest that the selected models are sufficiently capable for regional analyses and future-change assessments, provided that spatial and seasonal biases are explicitly considered in interpretation. For future projections, the overall signal points to a pronounced increase in summertime heat stress during the projected period relative to the reference climate. Heatwave frequency and the number of very hot days rise across most of the domain, with the strongest amplification over interior and desert regions—particularly the Arabian Peninsula, Iraq, and large parts of Iran. A key implication is that, even under optimistic emissions pathways, a substantial intensification of summer heatwaves is largely unavoidable and may have serious consequences for public health, energy demand, and water resources. Under the intermediate pathway, increases in both the frequency and spatial extent of heatwave occurrence become more persistent across many mid- and higher-latitude parts of the region, revealing clearer spatial contrasts. This indicates that regional responses to future warming are not uniform and are strongly shaped by local climate conditions and geographic setting. One of the most important findings is the emergence of nonlinear behavior in heatwave metrics under the high-emissions pathway. While a monotonic increase in heatwave frequency might be expected with stronger warming, the results show that in some areas—particularly toward the end of the projection period—frequency can decline relative to the intermediate pathway, even though overall heat stress remains above the baseline. This suggests that, under intense warming, the dominant change mechanism may shift from “more events” toward event merging and longer-lasting heatwaves. In other words, rather than producing a greater number of discrete episodes, a hotter climate may favor more persistent, multi-week heatwave conditions—reducing the count of events while increasing their duration and intensity. The spatiotemporal analyses further indicate that heatwave changes vary with time and scenario in both latitude and longitude, and do not organize into a single, uniform gradient across the study domain. Under the optimistic pathway, interannual variability and episodic spikes are evident, but a consistent, region-wide upward trend is not dominant. Under the intermediate pathway, increases become more coherent and sustained, with larger portions of the region—especially mid- and higher-latitude areas—remaining at elevated heatwave-day levels for longer periods. Under the high-emissions pathway, the pattern becomes more unstable and distinctly nonlinear: intervals of sharp increases alternate with periods of relative decline or large variability, and hotspots of thermal stress may emerge in different longitudes at different times. This behavior highlights the importance of accounting for interannual variability and regional processes in heatwave risk assessment. At the decadal scale, heatwave intensity generally increases across all scenarios, with scenario divergence becoming more evident from the middle decades onward. In some cases, the intermediate pathway exhibits a larger upward shift in the median and spread of intensity, whereas the high-emissions pathway displays nonlinear responses. Nevertheless, heatwave duration intensifies most strongly under the high-emissions pathway and can accelerate toward the end of the period. This is particularly important for risk management, as prolonged heatwaves—even if fewer in number—can impose greater cumulative stress on public health, labor productivity, electricity demand, and water-system reliability. Sub-regional analyses reveal clear contrasts among the northern, central, and southern sectors. The northern and central sectors are more sensitive in terms of intensity increases, with stronger intensification under higher-emissions pathways. The southern sector, despite being warmer in absolute terms, shows smaller relative increases in intensity—potentially reflecting proximity to already-high thermal thresholds and physical–statistical constraints on relative growth. In terms of duration or heatwave days, all three sectors experience substantial increases, but the central sector exhibits the largest jump; under the high-emissions pathway, heatwaves may shift from multi-day events to multi-week phenomena. The convergence of scenarios early in the period and their divergence later further indicate that the influence of emissions pathways grows over time, and that mitigation and adaptation choices can meaningfully shape future risk. Overall, the Middle East is projected to face a significant rise in summertime heat stress during 2026–2055. Importantly, this increase is not limited to a higher number of events; under warmer scenarios, it may involve a transformation toward longer and more intense heatwaves. Accordingly, adaptation planning should simultaneously address the three core dimensions of heatwave risk—frequency, duration, and intensity—while explicitly considering spatial differences across sub-regions. Given the weaker model performance in winter and the heightened sensitivity of complex terrain and land–sea interface areas, seasonal bias correction and uncertainty assessment should be treated as integral components of both scientific analysis and policy-relevant applications.



Conclusion

This study shows that CMIP6 model performance in estimating temperature is seasonally dependent: simulations are more reliable in summer and autumn, whereas winter exhibits the largest bias and RMSE, underscoring the need for seasonal bias correction and targeted calibration—particularly for winter conditions. Spatially, errors are concentrated mainly over mountainous terrain and along land–sea transition zones, while model performance is more stable across interior regions. The results also indicate that future warming is not confined to higher daytime maxima; the widespread rise in minimum temperatures (especially at night) emerges as a key climate-change signal across the region. Heatwave metrics further reveal that thermal risk increases under all emissions scenarios, but the magnitude and spatial pattern of change are scenario- and location-dependent and can be nonlinear. Heatwave intensity increases most strongly in the northern and central sub-regions, while persistence/number of heatwave days shows an even more pronounced rise—particularly in the central sector, where events may extend to near multi-week durations. Overall, the Middle East is likely to experience a shift in the heatwave regime toward more intense and longer-lasting events; consequently, adaptation strategies should address both “intensity” and “persistence” simultaneously and explicitly incorporate sub-regional differences in planning for health, energy, water, and critical infrastructure.