Santos (2026) Escalating thermal extremes and climate risk in a Brazilian semi-arid region: insights from ETCCDI indices and CMIP6 model projections

Identification

• Journal: Natural Hazards
• Year: 2026
• Date: 2026-01-01
• Authors: Daris Correia dos Santos
• DOI: 10.1007/s11069-025-07814-y

Research Groups

Department of Civil and Environmental Engineering, Federal University of Rio Grande do Norte, Natal, Brazil

Short Summary

This study evaluates the performance of 15 CMIP6 climate models in simulating thermal extremes in Rio Grande do Norte (RN), Brazil, identifying the MRI-ESM2-0 model as most suitable. Projections indicate a significant intensification of heat extremes across all future emission scenarios, with annual maximum daily maximum temperature (TXx) exceeding 42 °C and Warm Spell Duration Index (WSDI) surpassing 250 days annually under the high-emission SSP5-8.5 scenario.

Objective

• To evaluate the performance of 15 CMIP6 climate models in simulating air temperature extremes in Rio Grande do Norte (RN), Brazil, during the historical period (1850–2014).
• To identify the most suitable CMIP6 model for future climate projections in the region.
• To project future thermal extremes using eight ETCCDI indices under four Shared Socioeconomic Pathway (SSP) scenarios (2015–2100).
• To provide critical insights for climate-sensitive sectors and inform evidence-based policy planning for climate adaptation and mitigation in the Brazilian semi-arid region.

Study Configuration

• Spatial Scale: State of Rio Grande do Norte (RN), Northeast region of Brazil, located between latitudes 4°49′ S and 6°59′ S and longitudes 34°58′ W and 38°36′ W.
• Temporal Scale:
  • Historical period: 1850–2014
  • Future projections: 2015–2100

Methodology and Data

• Models used:
  • 15 Coupled Model Intercomparison Project Phase 6 (CMIP6) Global Climate Models (GCMs) were evaluated: MIROC-ES2L, HadGEM3-GC31-LL, ACCESS-CM2, CanESM5, MPI-ESM1-2-LR, MPI-ESM1-2-HR, FGOALS-g3, KACE-1-0-G, GFDL-ESM4, INM-CM4-8, EC-Earth3, BCC-CSM2-MR, MIROC6, NESM3, and MRI-ESM2-0.
  • MRI-ESM2-0 was selected as the best-performing model for future projections.
  • Eight extreme temperature indices recommended by the Expert Team on Climate Change Detection and Indices (ETCCDI) were analyzed: TXx (annual maximum value of daily maximum temperature), TXn (annual maximum value of daily minimum temperature), TNx (annual minimum value of daily maximum temperature), TNn (annual minimum value of daily minimum temperature), TX90p (percentage of days when daily maximum temperature is above the 90th percentile), TN90p (percentage of days when daily minimum temperature is above the 90th percentile), DTR (Diurnal Temperature Range), and WSDI (Warm Spell Duration Index).
  • Four Shared Socioeconomic Pathway (SSP) scenarios were used for future projections: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5.
• Data sources:
  • ERA5-Land reanalysis data (Copernicus Climate Change Service, 2023) served as the reference for model performance assessment.
  • Observational air temperature data from meteorological stations operated by the Brazilian National Institute of Meteorology (INMET) and the Agricultural Research Corporation of Rio Grande do Norte (EMPARN) for the period 2002–2014 were used for model validation.
  • Daily maximum and minimum temperature series from the 15 CMIP6 GCMs (Copernicus Climate Change Service, 2023).
  • Bias correction applied to CMIP6 simulations using the ISIMIP3b protocol with the WFDE5 observational dataset derived from ERA5.
  • Model performance assessed using Taylor diagrams, Pearson correlation, root mean square error (RMSE), and bias.
  • Temporal trends analyzed using the Mann-Kendall test and Sen’s slope estimator.
  • Probability density distributions analyzed using Kernel Density Estimation (KDE).

Main Results

• Model Validation: The MRI-ESM2-0 model demonstrated the best performance in simulating air temperature over RN, showing a standard deviation close to observed values, a Pearson correlation coefficient of approximately 0.5, and an RMSE of 5.13 °C.
• Historical Trends (1850–2014): Statistically significant positive trends (p < 0.05) were observed for TXx, TNn, TNx, TX90p, and TN90p, indicating an intensification of extreme heat and a reduction of cold extremes. TXx increased by an average of 0.0039 °C per year, totaling +0.64 °C over 164 years.
• Future Projections (2015–2100):
  • Overall Warming: A continuous and statistically significant intensification of thermal extremes is projected across all SSP scenarios.
  • High-Emission Scenario (SSP5-8.5): Under this scenario, the annual maximum daily maximum temperature (TXx) is projected to exceed 42 °C, and the Warm Spell Duration Index (WSDI) is projected to surpass 250 days per year, with a total variation of up to 277.78 days by 2100. The frequencies of warm days (TX90p) and warm nights (TN90p) are projected to approach 90%. The total accumulated increase in TXx by 2100 is projected to be +4.52 °C.
  • Spatial Variability: Western and Central mesoregions are projected to experience the most extreme daytime heat, while coastal areas will face increased nocturnal heat stress.
  • Nighttime Warming: The Diurnal Temperature Range (DTR) shows a slight decreasing trend across all scenarios, indicating that minimum temperatures are increasing at a faster rate than maximum temperatures. TN90p is projected to rise to over 80% in SSP2-4.5, SSP3-7.0, and SSP5-8.5 scenarios.
  • Low-Emission Scenario (SSP1-2.6): Even under this most optimistic scenario, warming trends remain statistically significant, with TXx increasing by 0.0093 °C per year and TN90p by 0.4800 % per year. WSDI tends to stabilize at intermediate levels (approximately 100–200 days per year).
• Probability Density Distributions: Future scenarios show shifts towards higher values, reduced symmetry, and narrowing of distributions for TXx and TXn, indicating lower interannual variability and increased recurrence of intense extremes. TNn and TNx distributions exhibit flattening and longer tails, reinforcing persistent nighttime warming. TX90p and TN90p distributions become more concentrated at higher values, suggesting a new climatic regime where hot days and nights become the norm.

Contributions

• Provides a robust validation of 15 CMIP6 models for simulating temperature extremes in a semi-arid tropical region, identifying MRI-ESM2-0 as the most suitable for regional applications in Rio Grande do Norte.
• Offers comprehensive projections of eight ETCCDI temperature extreme indices for the state of RN under four SSP scenarios, covering both historical (1850–2014) and future (2015–2100) periods.
• Quantifies the significant intensification and persistence of thermal extremes, highlighting that future heatwaves could last over 250 consecutive days per year under high-emission scenarios, a stark contrast to historical patterns.
• Identifies specific regional vulnerabilities, showing that Western and Central RN will face the most extreme daytime heat, while coastal areas will experience stronger nocturnal warming.
• Emphasizes that warming trends and their impacts are statistically significant even under low-emission scenarios, underscoring the urgency and irreversibility of regional climate change and the necessity for immediate adaptation strategies.
• Delivers critical, evidence-based insights for climate-sensitive sectors such as public health, agriculture, and water resource management, supporting strategic planning for climate adaptation and mitigation policies.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Citation

@article{Santos2026Escalating,
  author = {Santos, Daris Correia dos},
  title = {Escalating thermal extremes and climate risk in a Brazilian semi-arid region: insights from ETCCDI indices and CMIP6 model projections},
  journal = {Natural Hazards},
  year = {2026},
  doi = {10.1007/s11069-025-07814-y},
  url = {https://doi.org/10.1007/s11069-025-07814-y}
}