Extractive operations along the mineral-rich coasts of Ghana over so many years have begun costing the lives of indigenous people living near the affected areas. Assessing the environmental impact of these activities over the 67 years of its perpetuation, underpins all local and global climate efforts, sustainability, and prevention of its adverse effects. This paper presents VitSegh24, an image segmentation model based on the Segformer pretrained model. The model is fine-tuned on high-resolution publicly available satellite masks data covering features like waste rock dumps, pits, water ponds, tailings dams and heap leach pads all identifiable within 238,533 square kilometers land cover of the country. We propose this model and future automation of the inference pipeline as a key tool that can be scaled over the entire continent of Africa, towards climate action and responsible land use. The code and dataset is provided at https://github.com/armtos-np/vitsegh