The purpose of this work is analyzing the temperature measurements from the UltraFast Thermometer 2.0 (UFT-2) during the campaign ACORES 2017. Several types of anomalies in the readings were investigated and labelled using an app written for that purpose. Several machine learning algorithms were tested on the task of automatically detecting said anomalies, using the data from UFT-2 and readings from other devices, including the liquid water content and wind speed.