Monitoramento climático acessível na amazônia: projeto e análise de desempenho de uma estação meteorológica de baixo custo

Abstract

This study aimed to develop, implement, and evaluate the technical and economic performance of a Low-Cost Weather Station (LCWS) based on the Raspberry Pi platform for climate monitoring in the Amazon region, specifically in the municipality of Belém–PA, Brazil. The research addresses the need to expand access to meteorological monitoring systems, considering that professional weather stations involve high acquisition and maintenance costs, limiting their deployment in regions with budget constraints. The developed prototype integrated temperature and humidity sensors (AM2301 and DHT11), a tipping-bucket rain gauge, and an anemometer, along with a power supply system based on 18650 lithium-ion batteries and photovoltaic generation. Unlike the previous model built with NodeMCU, this study adopted the Raspberry Pi 3B to provide greater processing capacity, programming flexibility, and improved data management. An ADS1115 analog-to-digital converter was incorporated to enable analog signal acquisition. The methodology included study area characterization, structural assembly in accordance with World Meteorological Organization recommendations, Python-based algorithm implementation, and comparative data collection alongside a reference station. Statistical analysis considered indicators such as Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), in addition to operational and energy performance evaluation. The results indicated technical feasibility and significant cost reduction compared to conventional commercial weather stations. However, operational limitations were identified in some sensors, and the adoption of the Raspberry Pi led to a considerable increase in energy demand. The study concludes that the proposed LCWS demonstrates strong potential for decentralized environmental monitoring, although structural improvements are recommended for future versions.