Stuttgart in-depth analysis. This work presents AQTA, an interactive system-user-system interface to involve and support users for environmental time series data, for dynamic future predictions and detailed patterns analysis. AQTA combines visual representations of multiple time series (i.e.i.e., historical, present) with future prediction information generated by deep machine learning models. AQAT helps analysts engage in a back-and-forth dialogue with the designed models such as Long Short-Term Memory(LSTM), Random Forest (RF), Multi-Convolutional Neural Network (MCNN), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN). These models can be dynamically selected in real-time (on the fly) and the analysts can compare the results visually in different conditions.