Use of reanalysis data of particulate matter for hospital admission prediction

• Grant number: 25/12252-1
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: December 01, 2025
• End date: November 30, 2026
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Guilherme Palermo Coelho
• Grantee: Laura Rodrigues Russo
• Host Institution: Faculdade de Tecnologia (FT). Universidade Estadual de Campinas (UNICAMP). Limeira , SP, Brazil

Abstract

Particulate matter (PM), especially inhalable particles such as PM10 (diameter less than 10 µm) and PM2.5 (diameter less than 2.5 µm), is among the main atmospheric pollutants associated with the worsening of respiratory diseases. However, in Brazil, PM monitoring remains limited, with monitoring networks concentrated in only a few states, making air quality supervision at the national level challenging. Given this limitation, this project proposes the use of PM reanalysis data from MERRA-2, wich provides continuous spatial coverage, combined with hospitalizations data for respiratory diseases from the TABNET platform, to develop predictive models based on machine learning. The goal is to predict hospital admissions related to PM in selected Brazilian capitals, chosen based on the consistency of available data. To this end, the following activities are planned: (i) an in-depth analysis of reanalysis processes, particularly MERRA-2, related to PM and meteorological variables; (ii) selection of capitals with reliable records of hospital admissions; (iii) literature review and exploratory data analysis to investigate associations between PM and respiratory diseases; and (iv) development of models to predict hospital admissions based on historical PM data obtained from MERRA-2, with accuracy evaluated using statistical metrics. It is expected that the resulting models can be applied in regions with limited air quality monitoring.