IoT Based Weather Station Prototype with CRHUDA Algorithm and Regression for Rain Prediction

Fast and accurate rainfall predictions are needed to improve preparedness for climate change. This work applies Internet of Things (IoT) technology in real-time weather monitoring prototype through the development of an Automatic Weather System (AWS). This system utilizes Arduino Uno, ESP8266, DHT-22, and BMP180 to measure temperature, humidity, and air pressure, and integrates IoT infrastructure for more accurate data collection. The collected data is used in the CRHUDA (Crossing Humidity, Dewpoint, and Atmospheric pressure) model to determine the starting point for rain formation based on the intersection of humidity and air pressure. This model is strengthened by a multiple linear regression algorithm which is refined with the Kalman Filter for more accurate rainfall estimates. The results show that the CRHUDA model and linear regression technique are able to predict rainfall patterns effectively, although the accuracy of the regression model is lower in mild weather conditions.