Rising summer temperatures in Greenland have accelerated the formation of supraglacial lakes. Since these lakes play a significant role in ice sheet dynamics and bed lubrication, their continuous monitoring in a warming Arctic is becoming essential. The 31st ACM SIGSPATIAL competition (GISCUP 2023) aims to automate the detection of these lakes using satellite imagery. In this paper, we present two solutions to this problem based on image segmentation techniques: a DeepLabv3+ model that ranked first, and a U-Net-based approach that ranked fourth. We provide details about our implementations and explain the rationale behind our choices and the challenges we faced. Our results contribute to the understanding of supraglacial lake fluctuations and offer a valuable tool for ongoing environmental monitoring.