Design and Development of a Self-Sustaining High Temporal Resolution Weather Station with Integrated Forecasting

Abstract

This study presents the design and preliminary validation of a self-sustaining automatic weather station (AWS) that uniquely combines high temporal-resolution data acquisition with embedded short-term forecasting capability. Unlike conventional low-cost AWS units that primarily function as data loggers, the proposed system integrates lightweight neural network and stochastic modeling routines for real-time prediction of atmospheric variables. The compact station incorporates sensors for air temperature, soil temperature, relative humidity, atmospheric pressure, solar radiation, and precipitation, all interfaced with an ATMEGA328microcontroller and a 24-bit analog-to-digital converter for enhanced measurement accuracy. Continuous off-grid operation is achieved through a solar rechargeable 12 V battery, while dual data handling- local microSD logging and GSM transmission at 10-minute intervals with redundant fail-safe storage ensures reliability in remote deployments. Simulated outputs for temperature, humidity, pressure, and solar radiation demonstrate realistic diurnal variability, confirming the system’s temporal sensitivity. The novelty of this work lies in embedding forecasting functionality directly within a low-cost, modular AWS platform, something not achieved in previous studies that either provide simple logging capability or rely on infrastructure-heavy systems for prediction. By integrating machine learning–based short-term forecasting with dual-mode redundancy in a solar-powered, field-deployable unit, the system addresses the unique challenges of data-sparse and resource-constrained regions such as the Niger Delta. This makes it a scalable solution for microclimatic monitoring, agricultural decision support, and localized early warning applications. Future efforts will focus on field deployment, long term performance evaluation, remote web interfacing, and integration with larger meteorological networks.