Fusion of Orbital and Surface Data for Air Temperature Estimation in Belém Using Machine Learning
Authors:
- Leonardo de O. TamasauskasLTLeonardo de O. Tamasauskas
- Waldemiro J. A. G. NegreirosWaldemiro J. A. G. Negreiros
- Pedro H. Do V. GuimarãesPGPedro H. Do V. Guimarães
- Williane G. S. PereiraWPWilliane G. S. Pereira
- Jean A. C. DiasJDJean A. C. Dias
- Gabriel B. CostaGabriel B. Costa
- Marcos C. da R. SeruffoMSMarcos C. da R. Seruffo
Open Access
- 19 July 2026
- conference paper
- Published by Sociedade Brasileira de Computação
- p. 137-145
Abstract
This study evaluates fusion strategies between ERA5-Land orbital data and INMET surface observations to estimate daily air temperature in Belém, using data from 1994 to 2024. Different Random Forest models were compared, along with SARIMAX, LSTM, and stacking (SARIMAX and RF). The RF model with multivariate fusion achieved the best performance (RMSE = 0.5157 °C; R2 = 0.708; Skill = 6.19%), reducing bias relative to the orbital baseline, while chronological validation confirmed the temporal stability of RF-based models. The results indicate that, under low equatorial thermal variability, tree-based models are the most robust for satellite–surface fusion and thermal downscaling.Keywords
This publication has 27 references indexed in Scilit:
- Advances in multi-source data fusion for precipitation estimation: remote sensing and machine learning perspectivesEarth-Science Reviews, 2025
- Unequal distributions of crowdsourced weather data in England and WalesNature Communications, 2024
- Patterns of tropical forest understory temperaturesNature Communications, 2024
- Statistical Analysis with Missing Data, Third Edition2019
- Ensemble learning: A surveyWIREs Data Mining and Knowledge Discovery, 2018
- WorldClim 2: new 1‐km spatial resolution climate surfaces for global land areasInternational Journal of Climatology, 2017
- Retrieval of subpixel Tamarix canopy cover from Landsat data along the Forgotten River using linear and nonlinear spectral mixture modelsRemote Sensing of Environment, 2010
- Reexamination of instrument change effects in the U.S. Historical Climatology NetworkGeophysical Research Letters, 2006
- Stacked generalizationNeural Networks, 1992
- River flow forecasting through conceptual models part I — A discussion of principlesJournal of Hydrology, 1970