Contrails are usually formed by aviation vehicles and are also known as ”ice clouds”. As the aircraft is flying, water vapor condenses on dust particles, forming the contrail. Due to the nature of these clouds—containing more ice crystals compared to those naturally produced by clouds— they are very good at conserving heat within the earth. This heavily contributes to global warming. Autonomously detecting contrails is the best solution to preventing their spread. The dataset leveraged in this paper is taken from Kaggle and is originally from the GOES-16 Advanced Baseline Imager (ABI). Kaggle provides image data of different wavelengths (9), along with labeled masks for each contrail. The image data used from Kaggle has images of the same contrail taking in 9 different wavelengths. The proposed model, uses a resnest26d backbone, with a base U-Net model to provide accurate results for detecting contrails. This model provides more accurate results on segmenting images of contrails compared to other ensemble models.