The frequency and severity of severe weather events are increasing, posing a growing threat to the community and economy. Enhanced weather monitoring and forecasting enabled by IoT and deep learning may offer more complete and practical insights into weather trends and patterns, thereby facilitating improved decision-making across multiple sectors. Utilising Internet of Things (IoT) devices and deep learning techniques, this study proposes an improved system for weather monitoring and forecasting. The proposed system gathers meteorological data from various IoT devices, such as temperature sensors, humidity sensors, and barometers, and predicts weather patterns using a deep learning model. The deep learning model is capable of predicting weather conditions such as temperatures, humidity, pressure, and precipitation because it has been trained using a large dataset of historical weather data. The proposed system's performance is compared to extant weather tracking and forecasting systems with its constraints and potential improvements are discussed. Results indicate that the accuracy and validation accuracy of both VGG19 and VGG16 increase with each epoch, while their loss and validation loss decrease. VGG19 outperforms VGG16 in terms of accuracy and loss, but for certain epochs, VGG16 performs better in validation accuracy. Every epoch, InceptionResNetV2 outperforms CNN in terms of accuracy and loss. VGG19 and VGG16 exhibit indications of overfitting, which may impact their performance.