Reanalysis Data Validation as a Tool to Support Agrometeorological Studies and SDG Monitoring in the Amazon: A Case Study of the 2023–2024 El Niño Event

This is the authors' abstract. We don't add key points for this paper.

Reliable climate information is essential for assessing extreme events and supporting environmental management in the Amazon, where sparse meteorological networks constrain climate monitoring. This study evaluated ERA5 and CFSR reanalysis datasets for estimating mean air temperature in Boa Vista, Belém, and Manaus during the 2023–2024 El Niño event. Reanalysis estimates were compared with Brazilian National Institute of Meteorology (INMET) observations using R2, MAE, RMSE, MBE, Taylor diagrams, and ANOVA followed by Tukey’s HSD test. ERA5 consistently outperformed CFSR, with higher R2 values (Boa Vista: 0.702 vs. 0.396; Belém: 0.606 vs. 0.336; Manaus: 0.787 vs. 0.339) and lower RMSE (1.06 vs. 2.40 °C; 0.59 vs. 1.40 °C; 0.94 vs. 2.15 °C, respectively). Substantial temperature differences were observed among the historical period (1990–2022), 2023, and 2024 (p < 0.001). These results demonstrate ERA5’s superior ability to represent air temperature variability during the 2023–2024 El Niño in the Brazilian Amazon. Validated ERA5 data can strengthen climate monitoring and support agricultural planning, drought assessment, climate adaptation, and environmental management, contributing to evidence-based actions related to SDGs 2, 11, 13, and 15.